This page has been archived and is no longer updated
Endosymbiotic gene transfer: Organelle genomes forge eukaryotic chromosomes
Keywords
Keywords for this Article
Add keywords to your Content
Save
|
Cancel
Share
|
Cancel
Revoke
|
Cancel
Rate & Certify
Rate Me...
Rate Me
!
Comment
Save
|
Cancel
Flag Inappropriate
The Content is
Objectionable
Explicit
Offensive
Inaccurate
Comment
Flag Content
|
Cancel
Delete Content
Reason
Delete
|
Cancel
Close
Full Screen
"NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 123 REVIEWS Mitochondria and plastids were once free-living prokaryotes. They have retained the bulk of their prokaryotic biochemistry but harbour only a rem- nant of the eubacterial genome that their respective ancestors possessed. Over time, chloroplast (the photo- synthetic plastid) and mitochondrial genomes have shrunk by orders of magnitude from the size of fully- fledged eubacterial genomes to approximately the size of plasmids. Concurrently, eukaryotic nuclear genomes have been the recipients of mitochondrial (mt) and chloroplast (cp) DNA donations and have expanded, often to enormous size and complexity. Signs of such chromosome rebuilding through endosymbiotic gene transfer are unmistakable in sequenced genomes. Studies make it clear that, during the roughly two billion years since eukaryotes arose, many genes have relocated from the ancestral organellar genomes to the nucleus. Many of these genes have become func- tionally competent nuclear copies that now drive the biogenesis of mitochondria and chloroplasts, but some others have evolved to control further essential cellular processes. As remodelled nuclear copies of organelle genes usurped the functions of those located in the organelle, biochemical pathways were transferred wholesale from the organelles to the cytosol and the mitochondrial and plastid genomes were reduced in size. More than 20 years have elapsed since it first became apparent that mitochondrial- and chloroplast-DNA sequences are also present in the nuclear genomes of most eukaryotic species. Genome sequencing projects have now uncovered abundant organelle-to-nucleus transfers. Rates of organelle DNA transfer to the nucleus are now measurable in the laboratory, and the results of the first studies show that it occurs at staggeringly high frequencies. Controversy surrounds the impact of endosymbiotic gene transfer on eukaryote genome evo- lution, on transgenic crop technology and on natural variation within species. Here, we review comparisons of nuclear, organellar, CYANOBACTERIAL and ?-PROTEOBACTERIAL genomes that address endosymbiotic gene transfer. We discuss direct observations of organelle-to-nucleus gene transfer in the laboratory, their evolutionary implica- tions and their consequences for organelle-based trans- gene containment strategies in genetically modified (GM) crops. Organelle genomes ? prokaryotic remnants That eukaryotic organelles contain genes with a non- Mendelian mode of inheritance was inferred at the beginning of the 1900s (REF. 1) (BOX 1), as was an endosym- biotic origin for organelles 2 .However, it was not until the 1970s that the notion that organelles originated from endosymbiotic prokaryotes gained some acceptance 3 and, later still, that sequence comparisons unequivocally ENDOSYMBIOTIC GENE TRANSFER: ORGANELLE GENOMES FORGE EUKARYOTIC CHROMOSOMES Jeremy N. Timmis*, Michael A. Ayliffe ? , Chun Y. Huang* and William Martin � Genome sequences reveal that a deluge of DNA from organelles has constantly been bombarding the nucleus since the origin of organelles. Recent experiments have shown that DNA is transferred from organelles to the nucleus at frequencies that were previously unimaginable. Endosymbiotic gene transfer is a ubiquitous, continuing and natural process that pervades nuclear DNA dynamics. This relentless influx of organelle DNA has abolished organelle autonomy and increased nuclear complexity. CYANOBACTERIA The group of pigmented, photosynthetic bacteria that contains the endosymbiont ancestors of chloroplasts. ?-PROTEOBACTERIA A subgroup of gram-negative bacteria, often called the purple bacteria, that are thought to be the endosymbiont ancestors of mitochondria. *School of Molecular and Biomedical Science, The University of Adelaide, South Australia 5005, Australia. ? CSIRO Plant Industry, GPO Box 1600, Australian Capital Territory 2601, Australia. � Institute of Botany III, University of D�sseldorf, D�sseldorf 40225, Germany. Correspondence to J.N.T. e-mail: jeremy.timmis@ adelaide.edu.au doi:10.1038/nrg1271 DISOMIC The condition in which there are two sets of similar (homologous) chromosomes, such that there are two alleles for each gene locus. These homologous chromosomes pair at meiosis and their segregation and transmission results in Mendelian inheritance. HAPLOID The condition in which there is only a single chromosome, or set of chromosomes, such that all loci are represented by only a single allele. CYTOPLASMIC ORGANELLES Here,confined to mean mitochondria and plastids. PROMISCUOUS DNA DNA that is present in more than one genetic compartment of the eukaryotic cell. 124 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS organelles themselves are biochemically diverse. Some mitochondria consume oxygen, some produce hydrogen (hydrogenosomes) 13 ,some make ATP with or without oxygen 17 and some extremely reduced types (mitosomes) make no ATP at all, but make iron-sulphur clusters for the cell instead 18 . Plastids are an even more diverse assemblage of organelles 16 . The closest cousins of mitochondrial and chloro- plast genomes are free-living ?-proteobacteria and cyanobacteria respectively, but which lineages among those groups gave rise to present-day organelles remains unresolved 10,19,20 .The modern ?-proteobac- terium Mesorhizobium loti harbours 7 Mb of DNA that encodes more than 6,700 proteins and its relative Bradyrhizobium japonicum contains a 9.1 Mb genome and more than 8,300 proteins. The cyanobacterium Nostoc PCC 7,120 has a 6.4 Mb genome, which encodes approximately 5,400 proteins, whereas that of Nostoc punctiforme is >9 Mb, coding for more than 7,200 proteins. Comparing these genome sizes with those of organelles puts the magnitude of organelle genome reduction into perspective (TABLE 1). Although microbial parasites can also have highly reduced genomes, genome reduction in parasites (loss of genes and functions) is fundamentally different from genome reduction in organelles (loss of genes, but not functions), a distinction that is too seldom underscored (BOX 3). Comparative analyses To day, we know that genes have been transferred to the nucleus from the ancestral genomes of organelles but key questions are, at present, how many and what kinds of genes were transferred, how did the transfer occur and how often did it occur. Two main categories of compara- tive analyses address these questions. First, studies that identify copies of genes that are still present in an organelle genome but that are also in other compart- ments of the same cell indicate the process of DNA movement and identify evolutionarily recent transfer events. Second, studies that analyse nuclear genomes, organelle genomes and the genomes of candidate prokaryotic ancestors to identify genes in the nucleus that are no longer present in the organelle show more ancient transfer events. Evolutionarily recent transfers shown by genome comparisons. Initial evidence that DNA could move among cell compartments came when fragments of cpDNA were found in the maize mitochondrial genome 21 .Reports of mtDNA sequences 22,23 and chloro- plast sequences 24 in nuclear DNA followed, and the term ?PROMISCUOUS DNA? was coined by J. Ellis 25 to connote DNA mobility among the genetic compartments of eukaryotic cells (FIG. 1).These findings were important not only because they provided evidence that gene transfer among cell compartments could occur, but also because they indicated that this might be a continuing process given how recently (in evolutionary terms) the DNA must have been transferred. Subsequently, complete copies of mtDNA were discovered in cat genomes and identified proteobacteria and cyanobacteria as ancestors of mitochondria and chloroplasts, respectively 4 (BOX 2). To account for the observation that many proteins that are encoded by the nuclear genome are essential to chloroplasts and mitochondria 5,6 , it was suggested that genes had been relocated from the ancestral organelles to the nucleus during evolution 7 . This hypothesis has proved to be robust, although the details of the process are more complex than initially predicted. Cytoplasmic organelles contain a minuscule set of genes compared with the nuclear genome. Both chloroplasts and mitochondria generally contain multi- ple circular haploid genomes that are present as mono- mers and multimers. The protein-coding capacity of organelle genomes varies markedly across eukaryotic lin- eages. Sequenced plastid genomes encode from 20 to 200 proteins 8,9 and mitochondrial genomes encode anything from 3 to 67 proteins 10,11 (see TABLE 1). However, some unusual mitochondrial genomes are composed of many linear chromosomes with one gene each 12 and there are even cases in which the mitochondrial genome is miss- ing altogether 13 .Similarly, circular DNA molecules with one gene each have been identified in some plastids 14 and occasionally the entire plastid has apparently disap- peared 15,16 .In addition to their genomic diversity, Box 1 | Inheritance of cytoplasmic and nuclear genes The nuclear genomes of most higher eukaryotic organisms are diploid and are characterized by DISOMIC inheritance and sexual reproduction. So, nuclear genes come in allelic pairs that are often subtly different from each other. Gametes that result from meiosis are HAPLOID and carry only a single allele (in this example, alleles A, B and C from the female parent and a, b and c from the male parent). The zygote that results from fertilization inherits one nuclear allele of each gene from each parent (that is, at the three example loci, it is Aa, Bb and Cc). By contrast, the CYTOPLASMIC ORGANELLES characteristically contain multiple, homogeneous genomes that are usually inherited from one parent only (in this example, and most commonly, the female parent). In tobacco and many other plants, the mitochondrial and chloroplast genomes are specifically degraded before fertilization (red crosses). There are many exceptions to this common inheritance pattern of genes in mitochondria and chloroplasts (for a review, see REF. 119). Female parent AA,BB,DD, diploid nucleus cp, mt, cytoplasmic genomes Diploid zygote Aa,Bb,Dd, diploid nucleus cp, mt, cytoplasmic genomes Male parent aa,bb,dd, diploid nucleus cp, mt, cytoplasmic genomes A,B,C Aa, Bb, Cc Female gamete Male gamete a,b,c Fertilization cp cp cp mt mt mt NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 125 REVIEWS divergence indicates recurrent transfer events, from ancient to contemporary. The human genome has at least 296 different numts of between 106 bp and 14,654 bp (90% of the mitochondrial genome) that cover the entire mtDNA circle 34 .Other studies tallied 612 mtDNA insertions in the human genome 35 ,a greater number because different sequence conservation criteria for identifying numts are used in different studies 36 .Older numts are more abundant in the human genome than recent INTEGRANTS, indicating that mtDNA can be ampli- fied once inserted 36,37 and many are organized as tandem repeats 35 .Barely detectable numts are present in Plasmodium 32 ,but highly conserved numts have now the term ?NUMTS? was coined to designate these nuclear stretches of mtDNA 26 .Numts have been found in the nuclear genomes of grasshoppers 27 ,primates 28,29 and shrimps 30 , and are often mistaken for bona fide mtDNA 31,32 . Eukaryotic genome sequences have more fully exposed the scale of integrated mitochondrial and cpDNA in the nuclear genome. Fragments of organelle DNA are becoming recognized as a normal attribute of nearly all eukaryotic chromosomes. For example, the yeast genome contains tracts with 80?100% similarity to mtDNA that range in size from 22 to 230 base pairs (bp) integrated at 34 sites 33 .This range of sequence ARCHAEBACTERIA An ancient group of organisms that have ribosomes and cell membranes that distinguish them from eubacteria. They sometimes show environmentally extreme ecology. NUMT An acronym to describe nuclear integrants of mitochondrial DNA. Box 2 | Endosymbiotic evolution and the tree of genomes Intracellular endosymbionts that originally descended from free-living prokaryotes have been important in the evolution of eukaryotes by giving rise to two cytoplasmic organelles. Mitochondria arose from ?-proteobacteria and chloroplasts arose from cyanobacteria. Both organelles have made substantial contributions to the complement of genes that are found in eukaryotic nuclei today. The figure shows a schematic diagram of the evolution of eukaryotes, highlighting the incorporation of mitochondria and chloroplasts into the eukaryotic lineage through endosymbiosis and the subsequent co-evolution of the nuclear and organelle genomes. The host that acquired plastids probably possessed two flagella 113 . The nature of the host cell that acquired the mitochondrion (lower right) is fiercely debated among cell evolutionists. The host is generally accepted by most to have an affinity to ARCHAEBACTERIA but beyond that, biologists cannot agree as to the nature of its intracellular organization (prokaryotic, eukaryotic or intermediate), its age, its biochemical lifestyle or how many and what kind of genes it possessed 120 . The host is usually assumed to have been unicellular and to have lacked mitochondria. PlantsEukaryotes Early diversification of algal/plant lineages and gene transfer to the host Cyanobacteria Proteobacteria Archaebacteria Origin of mitochondria The host that acquired mitochondria Ancient proteobacterium Ancient cyanobacterium Ancient protozoon Origin of plastids Early diversification of eukaryotic lineages and gene transfer to the host 126 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS genome of Arabidopsis contains a large (~620 kb) insert of mtDNA on chromosome 2 (REFS 39,40),as well as an extra ~7 kb that is distributed among 13 small integrants and 17 insertions of cpDNA (?NUPTS?: the plastid counterparts of numts), totalling 11 kb (REF. 41).In the rice genome, chromosome 10 alone been identified in genome data for Rattus norvegicus, Caenorhabditis elegans, Drosophila melanogaster, Schizosachharomyces pombe, Ciona intestinalis, Neurospora crassa and Fugu rubripes 38 . Plant genomes provide even more abundant evi- dence of endosymbiotic DNA transfer. The nuclear INTEGRANT Here, used to describe nuclear tracts of DNA that resemble plastid DNA or mitochondrial DNA. NUPT An acronym to describe nuclear integrants of plastid DNA. Table 1 | Sizes and coding content of some organelle and prokaryote genomes Genome Length [kbp] Number of protein-coding genes GenBank accession number Algae cp Porphyra purpurea 191 200 PPU38804 cp Cyanidium caldarium 165 197 AF022186 cp Guillardia theta 122 148 AF041468 cp Cyanophora paradoxa 136 136 CPU30821 cp Odontella sinensis 120 124 OSCHLPLXX cp Euglena gracilis 143 58 CLEGCGA Land plants cp Marchantia polymorpha 121 84 CHMPXX cp Chlorella vulgaris 151 78 AB001684 cp Nicotiana tabacum 156 76 CHNTXX cp Oryza sativa 134 76 X15901 cp Zea mays 140 76 ZMA86563 cp Pinus thunbergii 120 69 PINCPTRPG Non-phosynthetic plastids cp Toxoplasma gondii 35 26 U87145 cp Eimeria tenella 35 28 AY217738 cp Epifagus virginiana 70 21 EPFCPCG Cyanobacteria Synechocystis sp. 3573 3168 AB001339 Prochlorococcus marinus 1660 1884 NC_005071 Nostoc PCC 7120 6413 5368 AP003602 Nostoc punctiforme ~9000 ~7400 http://www/jgi/doe/gov Plants and algae mt Pylaiella littoralis 59 52 NC_003055 mt Marchantia polymorpha 187 41 MPOMTCG mt Laminaria digitata 38 39 AJ344328 mt Cyanidioschyzon merolae 32 34 NC_000887 mt Arabidopsis thaliana 367 31 MIATGENA mt Chondrus crispus 26 25 MTCCGNME mt Scenedesmus obliquus 43 20 NC_002254 Various protists and fungi mt Reclinomonas americana 69 67 NC_001823 mt Malawimonas jakobiformis 47 49 AF295546 mt Naegleria gruberi 50 46 NC_002573 mt Rhodomonas salina 48 44 NC_002572 mt Dictyostelium discoideum 56 40 NC_000895 mt Phytophthora infestans 38 40 NC_002387 mt Acanthamoeba castellanii 42 36 U12386 mt Cafeteria roenbergensis 43 34 NC_000946 mt Monosiga brevicollis 77 32 AF538053 mt Physarum polycephalum 63 20 AB027295 mt Harpochytrium sp 24 14 AY182006 mt Candida albicans 40 13 NC_002653 mt Cryptococcus neoformans 25 12 NC_004336 mt Plasmodium falciparum 63 NC_001677 Anaerobic mitochondria mt Hydrogenosomes* 0 0 ?-proteobacteria Caulobacter crescentus 4017 3767 AE006573 Mesorhizobium loti 7596 7281 BA000012 Bradyrhizobium japonicum ~9100 ~8300 BA000040 Yeast (nuclear) 13,469 6,327 http://www.ebi.ac.uk An excellent, up-to-date list of sequenced organelle genomes is available at http://megasun.bch.umontreal.ca/ogmp/projects/other/ all_list.html. Prokaryote data was gratefully received from http://dna-res.kazusa.or.jp and http://www.jgi.doe.gov/JGI_microbial/html. *Hydrogenosomes are anaerobic forms of mitochondria that usually lack a genome. cp, chloroplast genome; mt, mitochondrial genome. NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 127 REVIEWS unique losses that are shared by descendant lineages by a ratio of more than10 to 1. Therefore, the similarity in gene content among contemporary plastid genomes is the result of immensely convergent evolution (FIG. 2). Ancient transfers that are shown by genome comparisons. The recognition of functional gene relocations, like the identification of numts and nupts, relied on sequence similarity between nuclear and organelle genomes. The discovery of genes that were transferred to the nucleus, but are no longer present in most organelle genomes, requires a different approach. To detect such transfers, searches of nuclear genomes with sequences that are pre- sent in the organelle genome of at least one species have been undertaken. The PROTIST Reclinomonas americana has the mitochondrial genome with the largest known gene content, and homologues of nearly all these genes have been found in the nucleus of other eukaryotes 60 . Similarly, most genes that are contained in the larger chloroplast genomes can be found as transferred homo- logues among plant and algal nuclear genomes 19,59 . Comparative studies have also shown gene transfer dur- ing secondary symbiosis ? the origin of plastids from eukaryotic algae instead of from cyanobacteria ? in the evolution of unicellular eukaryotes 61?63 . A search for nuclear genes in the Arabidopsis genome that branch with cyanobacterial homologues in PHYLO- GENETIC trees (or that have homologues in cyanobacteria only) showed that approximately 1,700 of the 9,368 Arabidopsis proteins that are sufficiently conserved in sequence to allow phylogenetic analysis, come from cyanobacteria, indicating that roughly 18% of the protein-coding genes in Arabidopsis are acquisitions from the plastid 19 .Many of these acquired genes clearly control cellular systems other than chloroplast biogene- sis and many are targeted to the cytosol or the secretory pathway 64 .Among eukaryotes, 630 nuclear-encoded proteins were identified that originated from mitochon- dria 65 ,ofwhich <30% were predicted to be targeted to mitochondria in yeast and human. So, the proteins encoded by many nuclear genes that are derived from organelle DNA ultimately take on new functions in new compartments. Ta rgeting and retargeting of proteins that are encoded by transferred genes. Case studies 8,66?68 and genome- wide analyses 64,69 show that the relationship between organelle gene donations and organelle protein imports is complex and difficult to predict. This contrasts with the older idea that proteins were always targeted to the cell compartment from which the genes that encoded them originated (the PRODUCT SPECIFICITY COROLLARY 7 (see REF. 66 for a discussion) to endosymbiotic theory). In Arabidopsis,fewer than half the proteins that are identi- fiable as acquisitions from cyanobacteria are predicted to be targeted to chloroplasts. Many are targeted to the cytosol, the secretory pathway or the mitochondrion. Conversely, a similar proportion of proteins that are targeted to the plastid do not seem to be acquired from cyanobacteria 70,71 .Clearly, the products of nuclear genes that originated from endosymbionts are free to explore contains 28 cpDNA fragments >80 bp long, including two very large insertions (33 kb (REF. 42) and ~131 kb (REF. 43)). Similarly, the draft sequence of rice chromo- some 1 shows many plastid insertions 44 . MtDNA insertions are also plentiful in the rice genome: chro- mosome 10 has 57 such segments that range from 80 to 2,552 bp (REF. 43).Whether larger nuclear genomes generally harbour more promiscuous DNA than smaller ones, such as Arabidopsis,remains to be seen. Transferred organelle DNA segments and some- times even complete organelle genomes are more or less ubiquitous as integrated constituents of eukary- otic genomes, as early studies had indicated 45,46 ,but the evolutionary consequences of such transfers have yet to be fully explored. Recurrent transfers and convergent gene losses. The dele- tion and functional replacement of mitochondrial genes by nuclear copies has effectively stopped in higher ani- mals 47 in which mitochondria encode 12 to 13 proteins, but the process is still actively continuing in higher plants, which have larger numbers of mitochondrially- encoded proteins. Thorough studies among flowering plants have uncovered many cases of transfer that result in expressed genes 48,49 .For example, the mitochondrial rps10 gene has been independently transferred to the nucleus many times 49,50 .Most mitochondrial transfer and activation events seem to involve recombination into pre-existing promoter and/or TRANSIT PEPTIDE-coding regions 8,48?57 . Similar to rps10, the chloroplast translation initiation factor 1 gene (infA) also shows striking evidence of mobility 58 .Nuclear relocations of this chloroplast gene were accompanied by MUTATIONAL DECAY and/or deletion of the corresponding chloroplast sequence. However, unlike rps10,characterization of transplanted nuclear infA genes indicated the appearance of de novo transit peptides rather than the parasitization of existing nuclear genes. Genes such as infA underscore earlier findings from genome analyses that parallel losses in indepen- dent lineages are regular occurrences in chloroplast genome evolution 59 . Plotting the process of chloro- plast genome reduction over time (gene losses from cpDNA) onto a chloroplast genome phylogeny 19 shows that parallel losses in independent lineages outnumber TRANSIT PEPTIDE A peptide sequence, often at the N-terminus of a precursor protein, that directs a gene product to its specific cellular destination. MUTATIONAL DECAY The process that describes the random changes that might occur in a DNA sequence in the absence of selection pressure. PROTIST A single-celled eukaryote. PHYLOGENETICS Reconstruction of the evolutionary relationships between sequences using any of a variety of inference methods. PRODUCT SPECIFICITY COROLLARY The situation in which the product of a gene that is donated by a cytoplasmic organelle to the nucleus is expected to be returned to that organelle. Box 3 | Genome reduction in organelles and parasites In addition to organelles, microbial parasites can also have highly reduced genomes. Examples are Rickettsia prowazeckii with ~830 protein-coding genes, Mycoplasma genitalium with ~470 and the parasitic eukaryote Encephalitozoon cuniculi with ~2,000 genes (half as many as Escherichia coli). However, the mechanism of genome reduction in parasites differs fundamentally from that in organelles. Whereas parasites simply lose the genes that they no longer need 70 ,organelles do require the products of many of the genes that they relinquish to the chromosomes of their host. So, organelle genome reduction is not simply an extension of parasite genome reduction. The nature of the two processes ? reduction through specialization to a nutrient-rich intracellular environment in the case of parasites versus reduction through export of essential genes to the host?s genetic apparatus with import of thousands of essential proteins from the cytosol in the case of organelles ? could not differ more. 128 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS mixing and matching of endosymbiotically inherited functions with newly evolved, eukaryote-specific bio- chemistry 78,79 . In summary, retargeting of proteins among organ- elles and, most notably, the cytosol is a highly dynamic and influential process in eukaryotic evolution. When genes are donated from organelles to the nucleus, there is no homing device that automatically re-routes the protein product back to the donor organelle. Rather, chance, natural selection and lineage diversification seem to govern the intracellular targeting fate of genes that organelles donate to the chromosomes of their host. In this sense, gene donations from organelles are important starting material for the evolution of new genes that are specific to the eukaryotic lineage. Laboratory estimates of transfer frequencies Comparative genome analyses show us that gene trans- fers have occurred at different times in the past, and indicate that the process is continuing. The challenge has been to get direct empirical estimates of the fre- quency at which DNA is being transferred among cellular compartments. new patterns of compartmentalization in the cell 11 . Moreover, gene donations from organelles often lead to functional replacement of pre-existing and functionally equivalent host genes, a process known as endosymbi- otic gene replacement 69 . The number of proteins that are predicted to be imported into mitochondria varies markedly across eukaryotic groups, ranging from ~150 proteins in the parasitic fungus Encephalitozoon cuniculi to ~4,000 pro- teins in humans. Only ~50 proteins were common to the mitochondria of all non-parasitic eukaryotes 72 . Similarly, the number of nuclear-encoded proteins that are predicted to be targeted to chloroplasts differs by a factor of two between rice and Arabidopsis 73 .Such pre- dictions still have clear limitations but are improving with the accumulation of more direct experimental data for localization 73,74 . For biochemical pathways that are present in both the original host and its endosymbionts, competition can ensue 8,66 .In some cases, the pathway of the sym- biont can predominate 18,75 but hybrid pathways can develop from both host and endosymbiont sources 66,76,77 . Organelle division is a prime example of lineage-specific EPISOME A unit of genetic material that is composed of a series of genes that sometimes has an independent existence in a host cell and at other times is integrated into a chromosome of the cell, replicating itself along with the chromosome. BIOLISTIC TRANSFORMATION A commonly used transformation method in which metal beads are coated with gene contructs and shot into cells. LEAF EXPLANTS Small sterile sections of leaf or other plant tissue from which whole plants might sometimes be regenerated. UNIPARENTAL INHERITANCE The mode of inheritance that generally characterizes the genes of cytoplasmic organelles in which only one of the two sexual partners contributes to the offspring. Other Mitochondrion Proteobacterium-like endosymbiotic ancestor Cyanobacterium-like endosymbiotic ancestor Proteins Chloroplast Nucleus Organelle DNA Organelle DNA Figure 1 | Organellar DNA mobility and the genetic control of biogenesis of mitochondria and chloroplasts. The eukaryotic mitochondrion is derived from a proteobacterial endosymbiotic ancestor but most of the genes that were originally present in this ancestor?s genome have been transferred to the nucleus (thick black arrow), with only a small number being retained in the organelle (blue circle). Similarly, most of the genes from the cyanobacterial endosymbiont ancestor of the chloroplast were also transferred to the nucleus (thick black arrow). So, as a result, cytoplasmic organelles are heavily dependent on nuclear genes and import more than 90% of their proteins from the cytoplasm (white arrows). The dotted arrows indicate how DNA of mitochondrial (blue) and chloroplast (green) origin is still being transferred to the nucleus. Chloroplast and nuclear sequences are also found in the mitochondrial genome but little or no promiscuous DNA is located in the chloroplast. NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 129 REVIEWS Mitochondrion-to-nucleus transfers. In yeast, a recombi- nant plasmid, which was introduced into a genome- lacking mitochondrion, was shown to relocate to the nucleus as an EPISOME (that is, not recombined into nuclear DNA) at a frequency of 2 x 10 ?5 per cell per generation 80 .A lower frequency (5 x 10 ?6 per cell per gen- eration) of episomal relocation was observed when the plasmid was integrated into the mitochondrial chromo- some 81 .In these experiments, the released DNA was epi- somal, indicating that release of DNA from the yeast mitochondrion is frequent, but integration might be rare in yeast nuclei because of their characteristically high level of reliance on homologous recombination for DNA incorporation. Newer work indicates that mtDNA escape in yeast occurs through an intracellular mechanism that depends on the composition of the growth medium and the genetic state of the mitochondrial genome, and is independent of an RNA intermediate 82 . Chloroplast-to-nucleus transfers in higher plants. Only more recently has it been possible to quantify the process of chloroplast-to-nucleus DNA transfer. To determine the frequency of plastid DNA transfer and integrative recombination into the higher plant nuclear genome, the plastome of tobacco was transformed with a neomycin phosphotransferase gene (neoSTLS2) that was tailored for expression only in the nuclear genome 83 (BOX 4).In 16 out of ~250,000 seedlings, the neoSTLS2 marker had been integrated into a nuclear chromosome, each time in a different location, which equates to a chloroplast- to-nucleus DNA transfer frequency of one in 16,000 gametes tested. The diversity of insertion locations indicates that the marker might be transposed during meiotic or postmeiotic events during male gamete for- mation because the extreme alternative explanation for these integrations ? a single transfer event that is subse- quently amplified by somatic cell division ? would lead to the same integration site being found in all plants with chloroplast integrants 83 .In agreement with the DNA integrations induced by BIOLISTIC TRANSFORMATION, these transfers show no particular preference for recombina- tion sites in either the nuclear or plastid genomes. Using a similar experimental strategy with a trans- gene in a different plastomic location, the frequency of chloroplast-to-nucleus transposition was estimated in tobacco somatic cells 84 .Leaf tissue from transplastomic tobacco that contained an intron-less neo gene was cul- tured on medium that contained high concentrations of kanamycin (100?400 mg/L). Twelve highly resistant plants were regenerated, 11 of which showed Mendelian inheritance of the antibiotic-resistant phenotype. After a courageous approximation of the number of regenerat- able cells that were present in LEAF EXPLANTS,a chloroplast- to-nucleus transposition frequency of one event in ~five million somatic cells was estimated 84 .Taken at face value, the frequency in somatic cells is ~300 times lower than that in male gametes of the same species 83 .The pro- grammed degeneration of plastids that occurs during pollen-grain development ? the process that underpins UNIPARENTAL INHERITANCE of plastid genes (FIG. 1) ? might explain this difference. After the chloroplast genomes are Evolutionary time 201 141 125 181 59 79 92 100 87 73 80 81 80 80 136 3,200 ~2,000 ? >7,000 Porphyra Guillardia Odontella Cyanidium Euglena Chlorella Nephroselmis Mesostigma Marchantia Pinus Flowering plants Cyanophora Synechocystis Other cyanobacteria Lineage diversification Chloroplast genome reduction and gene transfer to the nucleus Photosynthetic electron transport, respiration, ATPase Translation Other Functional categories Genome reduction Figure 2 | Reduction of the chloroplast genome over time. We know that plastids originated more than 1.2 billion years ago, because fossil red algae of that age have been found 121 . The ancestor of plastids was a free-living cyanobacterium and therefore must have possessed several thousand genes as did its contemporaries. Subsequent to the invention of a plastid protein import apparatus (a prerequisite for relocating genes that encode proteins required by the organelle to the nucleus), plastids relinquished most of their genes to the genome of their host cell. This gene relocation process occurred massively at the onset of endosymbiosis and continued in parallel during algal diversification, yet the same core set of genes (for photosynthesis and translation) has been retained in all lineages. The size of the bars shown indicates the genome sizes of chloroplasts from a diversity of plant lineages, from red algae (Porphyra) to angiosperms (flowering plants) and Cyanophora (belonging to the most ancient lineage of photosynthetic eukaryotes), and their free-living cyanobacterial relatives (cyanobacteria). The reduction in chloroplast genome size has been mapped onto a phylogenetic tree of the relationships among these genomes. Numbers at the end of branches indicate the number of genes that are present in the respective genome. These genes are divided into three functional categories that are represented by the three different colours making up the bars. Data from REF. 19. 130 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS somatic events 84 occurs early and enters the germline, we would expect, given a sufficient sample size, to find whole flowering branches that clonally transmit de novo nupts to pollen nuclei. Chloroplast-to-nucleus transfers in algae. Chloroplast- to-nucleus transfer has also been investigated in the unicellular alga, Chlamydomonas reinhardtii 71 .Some 13 degraded, fragmented DNA could become more avail- able for transfer. In yeast, an enhanced rate of mtDNA escape to the nucleus has been linked to increased degra- dation of abnormal mitochondria 85 .Similarly, in animals ? which have a sequestered germline ? the transfer of mitochondrial sequences to the nucleus might be most likely during sperm mitochondrial degeneration follow- ing fertilization 37 .Assuming that a proportion of the TRANSPLASTOME The condition of a plastid genome after the insertion of non-native genes. Box 4 | Design of experiments that showed DNA transfer from organelles to nucleus in real time A construct that consists of chloroplast sequences (C and D) that flank two selectable marker genes is inserted into the chloroplast genome through homologous recombination, thereby transforming the native plastome into a TRANSPLASTOME (a). In the experiments of Huang et al. 83 , the flanking chloroplast sequences were in the inverted repeats of the tobacco plastome. One of the selectable genes (aadA in the case illustrated) is designed for exclusive expression in the chloroplast and incorporation of this marker confers spectinomycin resistance. The other gene, a neomycin phosphotransferase gene neoSTLS2 (that encodes NPTII and incorporates a nuclear intron; here neo), is tailored, by virtue of a nuclear-specific promoter and the presence of a nuclear intron in the reading frame, for expression only when it is transposed to the nucleus. Continuous selection of growing leaf cells on spectinomycin medium allows transformed plastomes to be selected and eventually the transplastome entirely replaces the native chloroplast genome, such that all copies of the chloroplast genome contain the two selectable marker genes (b). Selection of cells or progeny seedlings on kanamycin medium allows the detection of the rare cases in which the neo gene has changed its location, such that strong expression is promoted from the nuclear environment (c). The progeny of self-fertilized transplastomic plants were not screened directly 83 .Rather, to eliminate low-level expression of neoSTLS2 from the chloroplast genome, transplastomic plants were crossed with wild-type female plants such that, because of strict maternal inheritance of tobacco plastids (BOX 1),progeny that contained only wild- type chloroplasts were produced. Therefore, chloroplast-to-nucleus transposition must have occurred at some stage during the life cycle of the male parent of the seedlings that were screened on kanamycin plates. The observation that 1 in 16,000 male tobacco gametes contained a newly integrated segment of chloroplast DNA (REF. 83) was unpredictably high, but it must be an underestimate of the true chloroplast-to- nucleus transposition frequency. In this experiment, the detection strategy enabled the identification only of those events that resulted in an entire, expressed neoSTLS2 gene in the nucleus. Other regions of the tobacco plastome that integrated in the nucleus without this selectable marker necessarily remained undetected. A similar strategy was used by Stegmann et al. 84 and by Lister et al. 85 . a A B C DE F G H I J Transplastome Insertion by homologous recombination b Selection for spectinomyacin resistance Chloroplast c Selection for kanamyacin resistance ABC DEFGH Transposition Chloroplasts Nucleus aadA (inactive) neo (inactive) aadA (active) neo (active) A B C DE F G H IJ C D Transplastome NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 131 REVIEWS are present in the mtDNA of some species are sought among many other species to monitor loss of the gene in the mtDNA, accompanied by the occurrence of a functional copy in the nuclear DNA (reviewed in REFS 48,49). As mitochondrial protein-coding genes of flowering plants often have introns and RNA EDITING 48 , and because the nuclear copies of mtDNA so detected lack the organelle-specific introns and the edited sites, the inference has been that cDNA intermediates of edited and spliced mRNAs are directly involved in the physical transfer process 48?51 . However, there are alternative interpretations for observations that underpin the cDNA intermediate view 52 .When cDNA intermediates of spliced and edited higher-plant mitochondrial transcripts occur, they should arise in the mitochondrion and so be more likely to recombine with mtDNA ? thereby erasing edited sites and introns in the mitochondrial gene ? than with nuclear DNA 52 .In this way, cDNA-mediated dynamics of intron loss and edited sites in higher-plant mtDNA can mimic potential cDNA intermediates in sequence comparisons, even if bulk mtDNA only were physically being transferred for recombination in the nucleus 52 .Adding to this is the complex nature of flow- ering plant mtDNA genomes, which are a hetero- geneous mixture of DNA molecules that are smaller than the ?master copy? 48 . Although the possibility that cDNA intermediates might be involved in the transfer of genes from mito- chondria to the nucleus in flowering plants cannot be excluded, as is often suggested 48?51,68 , there are alternative interpretations of the same data 52 . On a broader scale, evidence that implicates cDNA intermediates in mtDNA transfers in eukaryotic groups other than flowering plants is so far lacking (yeast mutants lacking mitochon- drial RNA polymerase transfer mtDNA efficiently 53 ), as is evidence that implicates cDNA in the transfer of chloroplast genes. The view of bulk DNA transfer finds support from genome comparisons 33?44 and from experi- mental organelle-to-nucleus studies in yeast and higher plants 80?85 .In the early evolution of mitochondria and chloroplasts, in which the brunt of endosymbiotic gene transfer occurred 10,11,19 (FIG. 2), cDNA intermediates would have been unnecessary in our view, because edit- ing and introns in free-living ?-proteobacteria and cyanobacteria are extremely rare at best. Where does transferred DNA integrate? There is no evidence from genomes that organelle DNA is inte- grated into preferred chromosomal regions or sequence contexts. In human nuclear DNA, 98% of all numts are integrated into sequences that are not anno- tated as potential genes and the remaining 2% are in introns 37 .Integration of mtDNA has been implicated in human somatic mutations leading to neoplasms 86 , but only one case of a numt causing heritable muta- tion has been reported 87 .The time and place of this insertion coincided with the Chernobyl nuclear reac- tor accident 87 , indicating that radiation-induced mtDNA fragmentation (or recombination involving pre-existing numts) might have been involved. billion haploid cells were screened, but no chloroplast- to-nucleus DNA transfer was detected. This finding implies a transfer of at least six orders of magnitude lower than in higher plants, which is consistent with the paucity of integrated cpDNA in the nuclear genome of C. reinhardtii 71 .The presence of only one chloroplast in this alga might explain this difference: if chloroplast lysis and subsequent genome degradation is required for endosymbiotic gene transfer (see above), then organisms with only one essential chloroplast would not survive. However, specific degradation of cpDNA in C. reinhardtii does occur during zygote formation in the MT ? STRAIN at fertilization so it would be interesting to determine whether gene transfer can occur after mating. Mechanisms, constraints and consequences The above findings provide a glimpse of how often DNA is transferred among cellular compartments ? but what has been learned about the mechanisms and constraints that govern the process? As in pre-genomic sequencing studies 45,46 , most numts and nupts had >95% nucleotide identity to the homologous organelle genes, probably because most of these sequences are rapidly eliminated or perhaps because regions with lower homology were not sought. The lack of divergence in most known numts and nupts indicates that they must have been transferred to the nucleus recently. Also, in all these examples, there has been no evidence of preferential transfer of a particular region or type of organelle sequence ? for example, coding versus non-coding regions, introns in the case of the nuclear copy of mtDNA in Arabidopsis,structural RNA genes or promoters. How is organelle DNA transferred? How do organelle genes physically get into the nucleus to recombine with chromosomal DNA? There are two opposing and hotly debated views on this topic that are based on different observations and that can be termed ?bulk DNA? and ?cDNA intermediates?. The ?bulk DNA? view, based on early comparative studies 53 ,yeast genetics 53 and genome sequence com- parisons 52 , argues that recombination between escaped organelle DNA molecules and nuclear chromosomes is the mechanism of gene transfer followed by further recombination 52 .It is founded on evidence from exper- imental transfer in yeast 53 and on the observation that in comparisons between organelle DNA and nuclear DNA of the same species (numts and nupts), intergenic spacers and other non-coding regions of organelle DNA are found in nuclear-transferred copies as often as are coding sequences 26,34,39?43 .When whole organelle DNA molecules that are >100 kb long are found recombined into eukaryotic chromosomes 39?43 that contain the organelle introns, the tRNAs and hundreds of kb of organelle non-coding regions 39,40 ,bulk DNA transfer seems likely 52 . The view of ?cDNA intermediates? holds that cDNAs of organelle mRNAs are the vehicle of gene transfer to the nucleus 48?51 .It is based on taxon sam- pling studies in flowering plants, in which genes that MT ? STRAIN One of the two mating types (the other is mt + ) of Chlamydomonas reinhardtii;one of each is required to form a zygote. RNA EDITING Changes in the RNA sequence after transcription is completed. Examples include modification of C to U or of A to I by deamination, or insertion and/or deletion of particular bases. 132 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS How important is the continuing contribution of organellar DNA to genetic variation in the nuclear genome? Unlike transposable elements, nuclear-encoded organelle sequences cannot undergo autonomous trans- position, which probably explains their modest coloniza- tion of the nuclear genome compared with that achieved by bona fide transposons. Nonetheless, if the transfer rates that were calculated recently are typical, one in every few thousand plants we see in nature possesses a fresh piece of cpDNA in its nucleus that it acquired only one generation ago 83 ,a frequency that compares with the nuclear mutation rate 84,98 , in which case nearly every plant would have its own individual nupt content. GM crops and gene transfer In addition to its pivotal role in eukaryotic genome evolution, gene transfer from organelles to the nucleus is also relevant in the debate that concerns the containment of GM crops. The chloroplast genetic compartment is the focus of one important strategy for transgene containment in GM crops 99 .The high estimates for chloroplast-to-nucleus transfer rates 83,84 cast doubt on claims that chloroplast transgenes can be reliably contained in most potential crop plants through strict maternal inheritance. This has evoked a heated debate 100,101 on the issue of whether crop biotech- nology strategies that focus on cpDNA should recognize a measurable potential for escape of transgenes in pollen nuclei. Of course, the experimental strategies that are applied to detect high-frequency transfer used genes that carry their own nuclear promoter so that any integrant that carries the whole gene can readily be detected. By con- trast, genes that are tailored for expression in the chloro- plast would need to acquire a nuclear promoter through recombination, analogous to the PROMOTER-TRAPPING tech- nique. The functional transfer and expression of an organelle-specific transgene should therefore be orders of magnitude less frequent than the primary event of DNA relocation. The introduction of organelle-specific introns into chloroplast transgenes might further aid the prevention of their functional transfer to the nucleus. The chloroplast-specific marker genes that are used to select for plastome transformation were nearly always transferred to the nucleus along with the selec- table marker but not expressed from that new location 83,84 .Therefore, it should be possible to equip genes that are designed for high activity in the chloro- plast with tight control of expression that would essen- tially preclude any nuclear function. Nuclear-specific suicide cassettes that are introduced adjacent to or within cpDNA transgenes might also aid containment strategies. Endosymbiotic gene transfer: bigger questions Is endosymbiotic gene transfer a quirk of evolution that affects only the tips in the tree of life, or is it the mecha- nism that forged eukaryotes out of prokaryotes? The answer to this question hinges on the issue of how many genes eukaryotes acquired from their mitochon- drial and chloroplast endosymbionts 102,103 and what Evidence has been found for a clustering of genes that are involved in mitochondrial nucleic-acid processing in Arabidopsis 88 ,but it is unlikely that the clustering reflects a relic of ancestral mitochondrial genome organization. From the standpoint of transcription, evidence has been found for important groups of nuclear-encoded genes that respond to the physiological state of the plas- tid in Arabidopsis, but it is unclear whether this reflects a relic of ancestral plastid gene regulation 89 . Why do some genes remain in organelles? Given the con- tinual ingress of organelle DNA into the nucleus, why should there be any genes left in organelles? The main, and hotly debated, theories on this issue fall largely into two groups: ?hydrophobicity? 49 and ?redox control? 90 .The former view holds that hydrophobic proteins are poorly imported by organelles, and so must be encoded in organelle genomes 49 .It accounts for organelle encoding of some membrane-integral proteins in chloroplasts and mitochondria, but not for all (for example, light- harvesting proteins or mitochondrial importers). The ?redox control? view holds that individual organelles need to control the expression of genes that encode components of their electron-transport chain so that they can be synthesized when they are needed to main- tain redox balance, thereby avoiding the production of highly toxic reactive oxygen species 90 .Both sides can draw on good experimental evidence in their favour 91?93 , adding to the controversy. However, many mitochondria have a modified genetic code, extensive RNA editing or both. In such cases, the mitochondrial genes are locked in place 93 ,as are genes for such proteins that have evolved organelle- specific assembly or translation mechanisms 94 .The types of gene that plastids (FIG. 2) and mitochondria of diverse lineages most tenaciously retain fit well with the redox-control hypothesis 90 , as do the findings that hydrogenosomes (anaerobic mitochondria) lack elec- tron transport and a genome, despite possessing many hydrophobic proteins in their membranes 13 .Clearly, many factors can influence the evolutionary trends of organelle genome persistence 49,54,90?96 , and continued debate on this important issue is assured. Transfers and natural variation. The observation that 1 in 16,000 male gametes in tobacco contains a brand new plastid DNA integrant raises the question as to why plant nuclear genomes are not overflowing with such sequences. A process of sequence elimination must counterbalance insertion. Genome sequencing has shown that although organelle sequences that are pre- sent in the nuclear genome vary in age, most are very similar to their organelle counterpart (often having >95% sequence identity). If promiscuous sequence elimination were a slow process, we would expect that most of these sequences would be highly divergent. These data indicate a high turnover of organelle sequences in the nucleus that is similar to that observed for other non-essential, intergenic sequences 97 ,but with recombination events occasionally leading to selection fixation as functional genes 98 . PROMOTER TRAP A genetic engineering technique that involves randomly inserting into the genome constructs that encode an easily detectable marker, such as GFP, but contain no promoter sequences. Marker expression is only detected when the construct lands near an endogenous genomic promoter. NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 133 REVIEWS could have obtained their eubacterial genes 111,112 . However, new data show that Giardia is not primi- tive 108,113 , and that it has mitochondria 18 too, so a pos- sible mitochondrial origin of eubacterial-like genes applies to the Giardia genome as well. On balance, the evidence from the increasing num- ber of eukaryotic genome sequences might tend to favour the ?mitochondria? end of this range of views. Specifically, there is abundant evidence for integrated fragments of organelle DNA but no reports of an inte- grated bacterial chromosome segment have emerged from genome sequences 98 , not even Drosophila,which harbours symbiotic ?-proteobacteria 20 (a possible transfer in a different insect 114 aside). So, organelle-to- nucleus transfer is widespread, continuing, real and abundant, but outright prokaryote-to-nucleus transfer seems to be rare, as a simple calculation illustrates. The sum of ~800 individual insertions of organelle DNA in yeast, human and Arabidopsis genomes, as well as chromosome 1 and 10 of rice, totals ~1.2 Mb of DNA, but there is no reported evidence at all for integrated chromosome fragments from free-living prokaryotes in those genomes. Accordingly, the contribution of LGT from prokaryotes to eukaryotic DNA in recent evolutionary history can be inferred from the small sample, which is, at most, 1/800th that of the contri- bution from organelles, and, at most, 1/12,000th that of organelles in terms of length of integrated DNA fragments that are more than 100 bp long. In terms of its directly observable impact on those eukaryotic genomes today, endosymbiotic gene transfer tends to outpace LGT. Looking into the depths of eukaryotic evolution- ary history when organelle genomes were still as big as their prokaryotic ancestors, and when the host genome lacked everything that it later acquired from organelles, the downpour of DNA from organelles must have decisively shaped the eukaryotic genome. After all, at the time when ancestral mitochondria first took up residence in their host, there were nei- ther transit peptides nor was there a protein import machinery 115 to target the protein products of trans- ferred genes back to mitochondria. Accordingly, early transfers from the primitive mitochondrion (a fully- fledged eubacterium) would have enriched the archaebacterial-derived chromosomes of the host with a whole genome?s worth of eubacterial genes, over and over again. However, expressed products could have been targeted only to the host?s cytosol and plasma membrane and not to the organelle 116 . Only after protein import into mitochondria had evolved (and later, independently for chloroplasts 117 ), could the process of organelle genome reduction begin. In its youth, endosymbiotic gene transfer was a powerful and chimaera-generating mechanism of natural variation that is truly unique to the eukary- otic lineage. Indeed, a look into prokaryotic chromo- somes shows that they possess nearly all the attributes of eukaryotic chromosomes 118 ;what is unique in eukaryotes is that there is more than one genetic compartment 1,2 . kinds of genes those free-living prokaryotes possessed. To day, thinking on the relatedness of prokaryotes and eukaryotes is still dominated by the ?universal? tree of rRNA, which would have us believe that eukaryotes should possess mainly, if not exclusively, genes that are archaebacterial in origin 104 .As we gather more data, this view is looking increasingly insecure 85 .For exam- ple, Rivera et al. 105 found that about two-thirds of the genes in the yeast genome that had identifiable prokaryotic homologues are more similar to eubacter- ial homologues than to archaebacterial homologues. Current views on how so many genes of eubacterial origin came to reside in the eukaryotic nuclear genome fall into a range, the extremes of which can be labelled as ?mitochondria? and ?lateral gene trans- fer? (LGT). One extreme (?mitochondria?), holds the view that all these eubacterial-like genes ultimately stem from the ancestral mitochondrial genome. At the other extreme is the view that the overall gene contribution from mitochondria is small and that various eukaryotic lineages have acquired eubacterial genes either through lineage-specific lateral transfers or from a mysterious symbiont in the ancient past. The main difficulty with the ?mitochondria? view is that most eubacterial-like genes in the eukaryotic genome do not branch specifically with ?-proteobacterial homologues in phylogenetic trees. To account for this, the ?mitochondria? view requires a few tenable corol- lary assumptions. First, gene phylogenies are imperfect (that is, gene trees produce erroneous branches for mathematical reasons) 19,106?109 .Second, sampling is incomplete (that is, more ?-proteobacterial genomes will provide a fuller picture) 20 .Third, LGT among prokaryotes in the ~two billion years since the origin of mitochondria has mixed up the chromosomes of free-living prokaryotes, thereby confounding today?s trees 110 .Fourth, when genes are transferred from organelles to the nucleus, they undergo a phase of evo- lution during which they acquire some odd mutations before they become fully functional 8 ,which can alter their position in gene trees 76 . The main difficulty with the LGT view is the lack of direct evidence from eukaryotic genomes in its favour. Initial analysis of the human genome sequence indi- cated that LGT from free-living prokaryotes to eukary- otes is widespread. However, a broader sampling of eukaryotic lineages showed that the initial evidence was far from conclusive 69 ? a salutary reminder that our sample of genomes (specifically among eukaryotes, cyanobacteria and ?-proteobacteria) is still extremely small. Nonetheless, there do seem to be some lineage- specific acquisitions in eukaryotes, as recent findings attest 103,110 .However, most of the evidence that favours LGT over eukaryotes is based on conflicting gene trees 63 , whereas theory and practice indicate that conflicting trees are to be expected even without LGT (REFS 106?109). The argument that mysterious symbionts from the ancient past donated genes to eukaryotes has recently taken a blow. Mystery symbionts helped explain how ?primitive? eukaryotes that were thought to lack mito- chondria, such as the paradigmatic Giardia intestinalis, inference into a directly observable process, opening the door to new progress in determining mechanisms and possibly even evolutionary experimentation. The future prospects seem bright for understanding the deeper evolutionary importance of endosymbiotic gene trans- fer and its role in shaping the compartmentalized chro- mosomes and heterogeneous biochemical organization of organelle-bearing (eukaryotic) cells. Conclusions Only 20 years have passed since gene transfers between eukaryotic genetic compartments became known. Individual case studies and genome sequences have marshalled overwhelming evidence for its continuous workings over evolutionary time. However, with organelle transformation now on the go, endosymbiotic gene transfer has changed from a compelling logical 134 | FEBRUARY 2004 | VOLUME 5 www.nature.com/reviews/genetics REVIEWS 1. Baur, E. Das Wesen und die Erblichkeitsverh�ltnisse der ?Varietates albomarginatae hort? von Pelargonium zonale. Z. Vererbungsl. 1, 330?351 (1909). 2. Mereschkowsky, C. �ber Natur und Ursprung der Chromatophoren im Pflanzenreiche. Biol. Centralbl. 25, 593?604 (1905). [English translation Eur. J. Phycol. 34, 287?295, 1999]. The starting point of endosymbiotic theory. Outlines the reasoning that we still use today to explain the origin of plastids from cyanobacteria, a paper that was many decades ahead of its time. 3. Margulis, L. Origin of Eukaryotic Cells 349 (Yale Univ. Press, New Haven, 1970). The rediscovery of endosymbiotic theory after Wilson?s 1928 condemnation of symbiosis as an evolutionary mechanism (see also reference 2). 4. Gray, M. W. & Doolittle, W. F. Has the endosymbiont hypothesis been proven? Microbol. Rev. 46, 1?42 (1982). 5. Bogorad, L. Evolution of organelles and eukaryotic genomes. Science 188, 891?898 (1975). 6. Ellis, R. J. Chloroplast proteins: synthesis, transport and assembly. Ann. Rev. Pl. Physiol. 32, 111?137 (1981). 7. Weeden, N. F. Genetic and biochemical implications of the endosymbiotic origin of the chloroplast. J. Mol. Evol. 17, 133?139 (1981). 8. Martin, W. & Herrmann, R. G. Gene transfer from organelles to the nucleus: how much, what happens, and why? Plant Physiol. 118, 9?17 (1998). 9. Simpson, C. L. & Stern, D. B. The treasure trove of algal chloroplast genomes. Surprises in architecture and gene content, and their functional implications. Plant Physiol. 129, 957?966 (2002). 10. Gray, M. W., Burger, G. & Lang, B. F. Mitochondrial evolution. Science 283, 1476?1481 (1999). 11. Lang, B. F., Gray, M. W. & Burger, G. Mitochondrial genome evolution and the origin of eukaryotes. Annu. Rev. Genet. 33, 351?397 (1999). 12. Burger, G., Forget, L., Zhu, Y., Gray, M. W. & Lang, B. F. Unique mitochondrial genome architecture in unicellular relatives of animals. Proc. Natl Acad. Sci. USA 100, 892?897 (2003). 13. Embley, T. M. et al. Hydrogenosomes, mitochondria and early eukaryotic evolution. IUBMB Life 55, 387?395 (2003). An incisive and up-to-date review that covers the biology and evolutionary significance of hydrogenosomes, the mitochondria that endosymbiotic theory nearly forgot. 14. Zhang, Z., Green, B. R. & Cavalier-Smith, T. Single gene circles in dinoflagellate chloroplast genomes. Nature 400, 155?159 (1999). 15. Hannaert, V. et al. Plant-like traits associated with metabolism of Trypanosoma parasites. Proc. Natl Acad. Sci. USA 100, 1067?1071 (2003). 16. Cavalier-Smith T. Membrane heredity and early chloroplast evolution. Trends Plant Sci. 5, 174?182 (2000). 17. Tielens, A. G., Rotte, C., van Hellemond, J. J. & Martin, W. Mitochondria as we don?t know them. Trends Biochem. Sci. 27, 564?572 (2002). 18. Tovar, J. et al. Mitochondrial remnant organelles of Giardia function in iron-sulphur protein maturation. Nature 426, 172?176 (2003). 19. Martin, W. et al. Evolutionary analysis of Arabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phylogeny and thousands of cyanobacterial genes in the nucleus. Proc. Natl Acad. Sci. USA 99, 12246?12251 (2002). Shows that approximately 18% of the nuclear genes in Arabidopsis come from the ancestral plastid genome. 20. Wu, M. et al. The genome sequence and evolution of the reproductive parasite Wolbachia pipientis wMel: a streamlined ?-proteobacterium massively infected with mobile genetic elements. PLoS Biology (in the press) . Incisive evolutionary insights into endosymbiont genome biology with a genome phylogeny for mitochondrial origin. 21. Stern, D. B. & Lonsdale, D. M. Mitochondrial and chloroplast genomes of maize have a 12-kilobase DNA sequence in common. Nature 299, 698?702 (1982). The paper that initiated progress; showed that DNA was able to migrate between the genetic compartments of eukaryotes. 22. Jacobs, H. T. et al. Mitochondrial DNA sequences in the nuclear genome of Strongylocentrotus purpuratus. J. Mol. Biol. 165, 609?632 (1983). 23. Farrely, F. & Butow, R. A. Rearranged mitochondrial genes in the yeast nuclear genome. Nature 301, 296?301 (1983). 24. Timmis, J. N. & Scott, N. S. Spinach nuclear and chloroplast DNAs have homologous sequences. Nature 305, 65?67 (1983). 25. Ellis, R. J. Promiscuous DNA ? chloroplast genes inside plant mitochondria. Nature 299, 678?679 (1982). 26. Lopez, J. V., Yuhki, N., Masuda, R., Modi, W. & O?Brien, S. J. Numt, a recent transfer and tandem amplification of mitochondrial DNA to the nuclear genome of the domestic cat. J. Mol. Evol. 39, 174?190 (1994). 27. Bensasson, D., Zhang, D. X. & Hewitt, G. M. Frequent assimilation of mitochondrial DNA by grasshopper nuclear genomes. Mol. Biol. Evol. 17, 406?415 (2000). 28. Mundy, N. I., Pissinatti, A. & Woodruff, D. S. Multiple nuclear insertions of mitochondrial cytochrome b sequences in callitrichine primates. Mol. Biol. Evol. 17, 1075?1080 (2000). 29. Lu, X. M., Fu, Y. X. & Zhang, Y. P. Evolution of mitochondrial cytochrome b pseudogene in genus Nycticebus. Mol. Biol. Evol. 19, 2337?2341 (2002). 30. Williams, S. T. & Knowlton, N. Mitochondrial pseudogenes are pervasive and often insidious in the snapping shrimp genus Alpheus. Mol. Biol. Evol. 18, 1484?1493 (2001). 31. Olson, L. E. & Yoder, A. D. Using secondary structure to identify ribosomal numts: cautionary examples from the human genome. Mol. Biol. Evol. 19, 93?100 (2002). Together with reference 32, points out that numts are often mistaken for genuine mitochondrial DNA sequences. 32. Bensasson, D., Zhang, D., Hartl, D. L. & Hewitt, G. M. Mitochondrial pseudogenes: evolution?s misplaced witnesses. Trends Ecol. Evol. 16, 314?321 (2001). 33. Ricchetti, M., Fairhead, C. & Dujon, B. Mitochondrial DNA repairs double strand breaks in yeast chromosomes. Nature 402, 96?100 (1999). The initial genome-wide survey for numts; indicates a role for recombination in numt integration. 34. Mourier, T., Hansen, A. J., Willerslev, E. & Arctander, P. The human genome project reveals a continuous transfer of large mitochondrial fragments to the nucleus. Mol. Biol. Evol. 18, 1833?1837 (2001). 35. Tourmen, Y. et al. Structure and chromosomal distribution of human mitochondrial pseudogenes. Genomics 80, 71?77 (2002). 36. Hazkani-Covo, E., Sorek, R. & Graur, D. Evolutionary dynamics of large numts in the human genome: rarity of independent insertions and abundance of post-insertion duplications. J. Mol. Evol. 56, 169?174 (2003). A careful and detailed inspection of numt duplication dynamics during human and primate genome evolution. 37. Woischnik, M. & Moraes, C. T. Pattern of organisation of human mitochondrial pseudogenes in the nuclear genome. Genome Res. 12, 885?893 (2002). 38. Richly, E. & Leister, D. Numts in sequenced eukaryotic genomes. Mol. Biol. Evol. (in the press). 39. Lin, X. Y. et al. Sequence and analysis of chromosome 2 of the plant Arabidopsis thaliana. Nature 402, 761?768 (1999). Reports a nearly complete, ~270-kb copy of the 367-kb Arabidopsis mtDNA near the centromere. 40. Stupar, R. M. et al. Complex mtDNA constitutes an approximate 620-kb insertion on Arabidopsis thaliana chromosome 2: implication of potential sequencing errors caused by large-unit repeats. Proc. Natl Acad. Sci. USA 98, 5099?5103 (2001). Shows that the 270-kb copy in reference 39 is really the complete 367-kb circle. 41. The Arabidopsis Genome Initiative. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796?815 (2000). 42. Yuan, Q. et al. Genome sequencing of 239-kb region of rice chromosome 10L reveals a high frequency of gene duplication and a large chloroplast DNA insertion. Mol. Genet. Genom. 267, 713?720 (2002). First hints from genome sequences of large cpDNA chunks (>30 kb) that are integrated in nuclear chromosomes. 43. The Rice Chromosome 10 Sequencing Consortium. In-depth view of structure, activity, and evolution of rice chromosome 10. Science 300, 1566?1569 (2003). Reports a nearly complete cpDNA genome chunk (130 kb) that is integrated in the nuclear chromosome. 44. Shahmuradov, I. A., Akbarova, Y. Y., Solovyev, V. V. & Aliyev, J. A. Abundance of plastid DNA insertions in nuclear genomes of rice and Arabidopsis. Plant Mol. Biol. 52, 923?934 (2003). 45. Ayliffe, M. A. & Timmis, J. N. Tobacco nuclear DNA contains long tracts of homology to chloroplast DNA. Theor. Appl. Genet. 85, 229?238 (1992). 46. Ayliffe, M. A. & Timmis, J. N. Plastid DNA sequence homologies in the tobacco nuclear genome. Mol. Gen. Genet. 236, 105?112 (1992). 47. Boore, J. L. Animal mitochondrial genomes. Nucleic Acids Res. 27, 1767?1780 (1999). 48. Brennicke, A., Grohmann, L., Hiesel, R., Knoop, V. & Schuster, W. The mitochondrial genome on its way to the nucleus: different stages of gene transfer in higher plants. FEBS Lett. 325, 140?145 (1993). 49. Adams, K. L. & Palmer, J. D. Evolution of mitochondrial gene content: gene loss and transfer to the nucleus. Mol. Phylog. Evol. 29, 380?395 (2003). A lucid review of flowering plant mitochondrial gene migration to the nucleus. 50. Adams, K. L., Daley, D. O., Qiu, Y. L., Whelan, J. & Palmer, J. D. Repeated, recent and diverse transfers of a mitochondrial gene to the nucleus in flowering plants. Nature 408, 354?357 (2000). 51. Adams, K. L, Qiu Y.-L., Stoutemyer, M. & Palmer J. D. Punctuated evolution of mitochondrial gene content: high and variable rates of mitochondrial gene loss and transfer to the nucleus during angiosperm evolution. Proc. Natl Acad. Sci. USA 99, 9905?9912 (2002). A broad survey of mitochondrial genome reduction that is accompanied by nuclear integration events in flowering plant evolution. 52. Henze, K. & Martin, W. How are mitochondrial genes transferred to the nucleus? Trends Genet. 17, 383?387 (2001). 53. Thorsness, P. E. & Weber, E. R. Escape and migration of nucleic acids between chloroplasts, mitochondria, and the nucleus. Int. Rev. Cytol. 165, 207?234 (1996). An excellent review that covers the mechanics of gene transfer from organelles to the nucleus in the pre-genome era. 54. Blanchard, J. L. & Lynch, M. Organellar genes ? why do they end up in the nucleus? Trends Genet. 16, 315?320 (2000). 55. Figueroa, P., Gomez, I., Holuigue, L., Araya, A. & Jordana, X. Transfer of rps14 from the mitochondrion to the nucleus in maize implied integration within a gene encoding the iron-sulphur subunit of succinate dehydrogenase and expression by alternative splicing. Plant J. 18, 601?609 (1999). Fortuitous recombination in the establishment of active gene transfers. 56. Kubo, N., Harada, K., Hirai, A. & Kadowaki, K. A single nuclear transcript encoding mitochondrial RPS14 and SSDHB of rice is processed by alternative splicing: common use of the same mitochondrial targeting signal for different proteins. Proc. Natl Acad. Sci. USA 96, 9207?9211 (1999). The establishment of active gene transfers can involve the recruitment of pre-existing transit peptide regions. NATURE REVIEWS | GENETICS VOLUME 5 | FEBRUARY 2004 | 135 REVIEWS 57. Long, M., de Souza, S. J., Rosenberg, C. & Gilbert, W. Exon shuffling and the origin of the mitochondrial targeting function in plant cytochrome c1 precursor. Proc. Natl Acad. Sci. USA 93, 7727?7731 (1996). Gene transfers can involve conversion of a pre- existing nuclear coding region for a cytosolic enzyme into a transit peptide. 58. Millen, R. S. et al. Many parallel losses of infA from chloroplast DNA during angiosperm evolution with multiple independent transfers to the nucleus. Plant Cell 13, 645?658 (2001). 59. Martin, W. et al. Gene transfer to the nucleus and the evolution of chloroplasts. Nature 393, 162?165 (1998). 60. Lang, B. F. et al. An ancestral mitochondrial DNA resembling a eubacterial genome in miniature. Nature 387, 493?497 (1997). The still unsurpassed mitchondrial genome in terms of gene content and streamlined organization. 61. Henze, K. et al. A nuclear gene of eubacterial origin in Euglena gracilis reflects cryptic endosymbioses during protist evolution. Proc. Natl Acad. Sci. USA 92, 9122?9126 (1995). 62. Stibitz, T. B., Keeling, P. J., & Bhattacharya, D. Symbiotic origin of a novel actin gene in the cryptophyte Pyrenomonas helgolandii. Mol. Biol. Evol. 17, 1731?1738 (2000). 63. Archibald, J. M., Rogers, M. B., Toop, M., Isheda, K. & Keeling, P. J. Lateral gene transfer and the evolution of plastid-targeted proteins in the secondary plastid-containing alga Bigelowiella natans. Proc. Natl Acad. Sci. USA 100, 7678?7683 (2003). 64. Leister, D. Chloroplast research in the genomics age. Trends Genet. 19, 47?56 (2003). 65. Gabald�n, T. & Huynen, M. A. Reconstruction of the proto- mitochondrial metabolism. Science 301, 609 (2003). 66. Martin, W. & Schnarrenberger, C. The evolution of the Calvin cycle from prokaryotic to eukaryotic chromosomes: a case study of functional redundancy in ancient pathways through endosymbiosis. Curr. Genet. 32, 1?18 (1997). 67. Gallois, J. L. et al. The Arabidopsis chloroplast ribosomal protein L21 is encoded by a nuclear gene of mitochondrial origin. Gene 274, 179?185 (2001). 68. Adams, K. L. et al. Genes for two mitochondrial ribosomal proteins in flowering plants are derived from their chloroplast or cytosolic counterparts. Plant Cell 14, 931?943 (2002). 69. Brown, J. R. Ancient horizontal gene transfer. Nature Rev. Genet. 4, 121?132 (2003). An incisive review of gene movement across genomes, including the role of endosymbiotic transfers. 70. Gil, R. et al. The genome sequence of Blochmannia floridanus: comparative analysis of reduced genomes. Proc. Natl Acad. Sci. USA 100, 9388?9393 (2003). Underscores how reductive evolution in endosymbiotic bacteria leads to massive gene losses through biochemical parasitism of the host. 71. Lister, D. L., Bateman, J. M., Purton, S. & Howe, C. J. DNA transfer from chloroplast to nucleus is much rarer in Chlamydomonas than in tobacco. Gene 316, 33?38 (2003). 72. Richly, E., Chinnery, P. F. & Leister, D. Evolutionary diversification of mitochondrial proteomes: implications for human disease. Trends Genet. 19, 356?362 (2003). 73. Richly, E. & Leister, D. An improved prediction of chloroplast proteins reveals diversities and commonalities in the chloroplast proteomes of Arabidopsis and rice. Gene (in the press). 74. Huh, W.-K. et al. Global analysis of protein localization in budding yeast. Nature 425, 686?691 (2003). 75. Lange, B. M., Rujan, T., Martin, W. & Croteau, R. Isoprenoid biosynthesis: the evolution of two ancient and distinct pathways across genomes. Proc. Natl Acad. Sci. USA 97, 13172?13177 (2000). 76. Schnarrenberger, C. & Martin, W. Evolution of the enzymes of the citric acid cycle and the glyoxylate cycle of higher plants: a case study of endosymbiotic gene transfer. Eur. J. Biochem. 269, 868?883 (2002). 77. Aravind, L., Anantharaman, V. & Iyer, L. M. Evolutionary connections between bacterial and eukaryotic signaling systems: a genomic perspective. Curr. Opin. Microbiol. 6, 490?497 (2003). 78. Osteryoung, K. W. & Nunnari, J. The division of endosymbiotic organelles. Science 302, 1698?1704 (2003). 79. McFadden, G. I. & Ralph, S. A. Dynamin: the endosymbiosis ring of power? Proc. Natl Acad. Sci. USA 100, 3557?3559 (2003). 80. Thorsness, P. E. & Fox, T. D. Escape of DNA from the mitochondria to the nucleus in the yeast, Saccharomyces cerevisiae. Nature 346, 376?379 (1990). The initial experimental measurement of the frequency of transfer of DNA between genetic compartments. 81. Thorsness, P. E. & Fox, T. D. Nuclear mutations in Saccharomyces cerevisiae that affect the escape of DNA from mitochondria to the nucleus. Genetics 134, 21?28 (1993). 82. Shafer, K. S., Hanekamp, T., White, K. H. & Thorsness, P. E. Mechanisms of mitochondrial DNA escape to the nucleus in the yeast Saccharomyces cerevisiae. Curr. Genet. 36, 183?194 (1999). 83. Huang, C. Y., Ayliffe, M. A. & Timmis, J. N. Direct measurement of the transfer rate of chloroplast DNA into the nucleus. Nature 422, 72?76 (2003). The initial experimental measurement of the frequency of integrative transfer of DNA between chloroplast and nucleus. 84. Stegemann, S., Hartmann, S., Ruf, S. & Bock, R. High- frequency gene transfer from the chloroplast genome to the nucleus. Proc. Natl Acad. Sci. USA 100, 8828?8833 (2003). 85. Campbell, C. L. & Thorsness, P. E. Escape of mitochondrial DNA to the nucleus in yme1 yeast is mediated by vacuolar- dependent turnover of abnormal mitochondrial compartments. J. Cell Sci. 111, 2455?2464 (1998). 86. Shay, J. W. & Werbin, H. New evidence for the insertion of mitochondrial DNA into the human genome: significance for cancer and aging. Mutat. Res. 275, 227?235 (1992). 87. Turner, C. et al. Human genetic disease caused by de novo mitochondrial-nuclear DNA transfer. Hum. Genet. 112, 303?309 (2003). 88. Elo, A. et al. Nuclear genes that encode mitochondrial proteins for DNA and RNA metabolism are clustered in the Arabidopsis genome. Plant Cell 15, 1619?1631 (2003). 89. Richly, E. et al. Covariations in the nuclear chloroplast transcriptome reveal a regulatory master-switch. EMBO Rep. 4, 491?498 (2003). 90. Allen, J. F. The function of genomes in bioenergetic organelles. Phil. Trans. Roy. Soc. B 358, 19?38 (2003). A comprehensive treatment of competing views on the issue of why some genes remain within organelle genomes. 91. P�rez-Mart�nez, X. et al. Subunit II of cytochrome c oxidase in chlamydomonad algae is a heterodimer encoded by two independent nuclear genes. J. Biol. Chem. 276, 11302?11309 (2001) 92. Daley, D. O. et al. Intracellular gene transfer: reduced hydrophobicity facilitates gene transfer for subunit 2 of cytochrome c oxidase. Proc. Natl Acad. Sci. USA 99, 10510?15015 (2002). 93. Pfannschmidt, T., Nilsson, A., & Allen, J. F. Photosynthetic control of chloroplast gene expression. Nature 397, 625?628 (1999). 94. Naithani S., Saracco S. A., Butler C. A. & Fox T. D. Interactions among COX1, COX2, and COX3 mRNA- specific translational activator proteins on the inner surface of the mitochondrial inner membrane of Saccharomyces cerevisiae. Mol. Biol. Cell 14, 324?333 (2003). 95. Doolittle, W. F. You are what you eat: a gene transfer ratchet could account for bacterial genes in eukaryotic nuclear genomes. Trends Genet. 14, 307?311 (1998). 96. Race, H. L., Herrmann, R. G. & Martin, W. Why have organelles retained genomes? Trends Genet. 15, 364?370 (1999). 97. Devos, K. M., Brown, J. K. M. & Bennetzen J. L. Genome size reduction through illegitimate recombination counteracts genome expansion in Arabidopsis. Genome Res. 12, 1075?1079 (2002). 98. Martin, W. Gene transfers from organelles to the nucleus: frequent and in big chunks. Proc. Natl Acad. Sci. USA 100, 8612?8614 (2003). 99. Maliga, P. Engineering the plastid genome of higher plants. Curr. Opin. Plant Biol. 5, 164?172 (2002). 100. Daniell, H. & Parkinson, C. L. Jumping genes and containment. Nature Biotechnol. 21, 374?375 (2003). A challenge to the experimental data for gene transfer from chloroplasts by proponents of plastid transgene technology, rebutted head-to-head in reference 101. 101. Huang, C. Y., Ayliffe, M. A. & Timmis, J. N. Organelle evolution meets biotechnology. Nature Biotechnol. 21, 489?490 (2003). 102. Martin, W. & Russell, M. On the origins of cells: a hypothesis for the evolutionary transitions from abiotic geochemistry to chemoautotrophic prokaryotes, and from prokaryotes to nucleated cells. Phil. Trans. Roy. Soc. Lond. B 358, 59?85 (2003). 103. Doolittle, W. F. et al. How big is the iceberg of which organellar genes in nuclear genomes are but the tip? Philos. Trans. Roy. Soc. Lond. B 358, 39?57 (2003). 104. Woese, C., Kandler, O. & Wheelis, M. L. Towards a natural system of organisms: proposal for the domains archaea, bacteria and eucarya. Proc. Natl Acad. Sci. USA 87, 4576?4579 (1990). The current higher-level taxonomic model, with eukaryotes as sisters to archaebacteria. 105. Rivera, M. C., Jain, R., Moore, J. E. & Lake, J. A. Genomic evidence for two functionally distinct gene classes. Proc. Natl Acad. Sci. USA 95, 6239?6244 (1998). A landmark paper that uncovers more eubacterial genes than archaebacterial genes in the yeast genome. 106. Penny, D., Foulds, L. R. & Hendy, M. D. Testing the theory of evolution by comparing phylogenetic trees constructed from five different protein sequences. Nature 297, 197?200 (1982). 107. Cummings, M. P., Otto, S. P. & Wakeley, J. Sampling properties of DNA-sequence data in phylogenetic analysis. Mol. Biol. Evol. 12, 814?822 (1995). 108. Embley, T. M. & Hirt, R. P. Early branching eukaryotes? Curr. Opin. Genet. Dev. 8, 655?661 (1998). 109. Rokas, A., Williams, B. L., King, N. & Carroll, S. B. Genome- scale approaches to resolving incongruence in molecular phylogenies. Nature 425, 798?804 (2003). 110. Gogarten P. J., Doolittle W. F. & Lawrence J. G. Prokaryotic evolution in light of lateral gene transfer. Mol. Biol. Evol. 19, 2226?2238 (2002). 111. Hedges, S. B. et al. A genomic timescale for the origin of eukaryotes. BMC Evol. Biol. 1, 4 (2001). 112. Hartman, H. & Fedorov, A. The origin of the eukaryotic cell: a genomic investigation. Proc. Natl Acad. Sci. USA 99, 1420?1425 (2002). 113. Stechmann, A. & Cavalier-Smith, T. Rooting the eukaryote tree by using a derived gene fusion. Science 297, 89?91 (2002). A milestone relating to the issue of which eukaryotes might be the most ancient. 114. Kondo, N., Nikoh, N., Ijichi, N., Shimada, M. & Fukatsu, T. Genome fragment of Wolbachia endosymbiont transferred to X chromosome of host insect. Proc. Natl Acad. Sci. USA 99, 14280?14285 (2002). 115. Neupert, W. Protein import into mitochondria. Ann. Rev. Biochem. 66, 683?717 (1997). 116. Martin, W. & M�ller, M. The hydrogen hypothesis for the first eukaryote. Nature 392, 37?41 (1998). 117. Soll, J. Protein import into chloroplasts. Curr. Opin. Plant. Biol. 5, 529?535 (2002). 118. Bendich, A. J. & Drlica, K. Prokaryotic and eukaryotic chromosomes: what?s the difference? BioEssays 22, 481?486 (2000). An enlightening survey of chromosome attributes in prokaryotes and eukaryotes. 119. Birky, C. W. The inheritance of genes in mitochondria and chloroplasts: laws, mechanisms and models. Annu. Rev. Genet. 35, 125?148 (2001). 120. Martin, W., Hoffmeister, M., Rotte, C. & Henze, K. An overview of endosymbiotic models for the origins of eukaryotes, their ATP-producing organelles (mitochondria and hydrogenosomes), and their heterotrophic lifestyle. Biol. Chem. 382, 1521?1539 (2001). 121. Butterfield, N. J. Bangiomorpha pubescens n. gen., n. sp.: implications for the evolution of sex, multicellularity, and the Mesoproterozoic/ Neoproterozoic radiation of eukaryotes. Paleobiology 26, 386?404 (2000). A fossil red algae of 1. 2 billion years of age anchors plant evolution in the Precambrian age. Acknowledgements We thank M. Limpert for help in preparing the manuscript, the Australian Research Council, the Australian?German Joint Research Cooperation Scheme and the Deutsche Forschungs- gemeinschaft for financial support, and D. Leister for valuable dis- cussions and permission to modify published figures. Countless individual report on numts, nupts and eukaryotic genes that were acquired from organelles are available; we apologize to all for having to focus on selected and more recent work. Competing interests statement The authors declare that they have no competing financial interests. Online links DATABASES The following terms in this article are linked online to: TAIR: http://www.arabidopsis.org rps10 FURTHER INFORMATION CyanoBase: http://www.kazusa.or.jp/cyano/cyano.html DOE Joint Genome Institute: http://www.jgi.doe.gov/JGI_microbial/html The Institute for Genomic Research: http://www.tigr.org/tdb The Organelle Genome Megasequencing Program: http://megasun.bch.umontreal.ca/ogmpproj.html Access to this interactive links box is free online. "
Add Content to Group
|
Bookmark
|
Keywords
|
Flag Inappropriate
share
Close
Digg
Facebook
MySpace
Google+
Comments
Close
Please Post Your Comment
*
The Comment you have entered exceeds the maximum length.
Submit
|
Cancel
*
Required
Comments
Please Post Your Comment
No comments yet.
Save Note
Note
View
Public
Private
Friends & Groups
Friends
Groups
Save
|
Cancel
|
Delete
Please provide your notes.
Next
|
Prev
|
Close
|
Edit
|
Delete
Genetics
Gene Inheritance and Transmission
Gene Expression and Regulation
Nucleic Acid Structure and Function
Chromosomes and Cytogenetics
Evolutionary Genetics
Population and Quantitative Genetics
Genomics
Genes and Disease
Genetics and Society
Cell Biology
Cell Origins and Metabolism
Proteins and Gene Expression
Subcellular Compartments
Cell Communication
Cell Cycle and Cell Division
Scientific Communication
Career Planning
Loading ...
Scitable Chat
Register
|
Sign In
Visual Browse
Close
Comments
CloseComments
Please Post Your Comment