I REVIEWS
Horizontal gene transfer: building the web of life
Shannon M. Soucy', Jinling Huang2 andJohann Peter Cogarten'5
Abstract | Horizontal gene transfer (HGT) is the sharing of genetic material between organisms that are not in a parent-offspring relationship. HGT is a widely recognized mechanism for adaptation in bacteria and archaea. Microbial antibiotic resistance and pathogenicity are often associated with HGT, but the scope of HGT extends far beyond disease-causing organisms. In this Review, we describe how HGT has shaped the web of life using examples of HGT among prokaryotes, between prokaryotes and eukaryotes, and even between multicellular eukaryotes. We discuss replacement and additive HGT, the proposed mechanisms of HGT, selective forces that influence HGT, and the evolutionary impact of HGT on ancestral populations and existing populations such as the human microbiome.
HGT has long been recognized as an important force in the evolution of bacteria and archaea. However, the exchange of genetic information between prokaryotic symbionts and their eukaryotic hosts, and even between eukaryotes, signifies that HGT in eukaryotes occurs more frequently than previously thought7'8. Often these transfers involve gene donations to unicellular eukaryotes9 and are frequently associated with bacterial endo-symbionts10 (known as endosymbiotic gene transfer (EGT) or intracellular gene transfer (IGT)). However, bacterial genes can also be transferred to multicellular eukaryotes8. Recent interest in the human microbiome has reinvigorated the search for HGTs from symbionts into the human genome. Although transfers of bacterial genes into the human germ line1112 have not be confirmed, evidence is accumulating of HGT from bacteria to human somatic cells13. These findings demonstrate the enduring influence of HGT on the evolution of all parts of the web of life, eukaryotes included.
In this Review, we present an overview of how HGT has contributed to innovation throughout the web of life by providing novel combinations of gene sequences for selection to act upon, thus shaping the evolution of species ranging from single-celled microorganisms to multicellular eukaryotes. Advances in the understanding of mechanisms of HGT, methods of identifying HGT events and the growth of genome databases have facilitated these insights.
Mechanisms of HGT
The three most recognized mechanisms of HGT in prokaryotes are conjugation, transformation and
Selfish genetic element
A gene or group of genes that enhance their own transmission and reproductive success without making a positive contribution to the host's fitness.
'Department of Molecular and Cell Biology, University of Connecticut, 91 North Eagleville Road, Starrs, Connecticut 06269-3125, USA.
2Department of Biology, East Carolina University, Greenville, North Carolina 27858, USA. 3Institute for Systems Genomics, University of Connecticut, Connecticut 06269-3125, USA.
Correspondence to J. P. G. e-mail: gogarten&uconn.edu
doi:10.1058/nrg5962
Horizontal gene transfer (HGT) was first described in microorganisms in the late 1940s1, and around 20 years later it was speculated to have a role in the adaptation of multicellular eukaryotes — specifically plants2. Since then, methods to detect HGT have improved, and these have revealed the surprising extent and relevance of HGT to the variation of viral, prokaryotic and eukaryotic gene content. Many apparent gene duplications, for example, are now known to be the result of HGT, not autochthonous gene duplication, resulting in a 'web of life' rather than in a steadily bifurcating tree3'4.
For a transferred gene to survive in the recipient lineage for long periods of time, the gene usually needs to provide a selective advantage either to itself (in the case of a selfish genetic element) or to the recipient, and research on HGT initially focused on such genes. However, it is now known that many of the genes that have been identified as transferred through comparative genomics between close relatives have neutral or nearly neutral effects in the recipient in both prokaryotic and eukaryotic organisms5. One rule for transferred genes seems to be 'first do no harm' — genes that are successfully integrated into a recipient are often expressed at low levels and encode functions at the periphery of metabolism6. These neutral acquisitions, however, can later provide novel combinations of genetic material for selection to act on — in some cases, the transferred material becomes domesticated over time and produces a beneficial phenotype. In other cases, when the imported genes remain neutral and there is no obvious benefit associated with their retention, the genes are likely to be lost over time.
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics
REVIEWS
Microbiome
Following a definition ascribed to Joshua Lederberg this term is most often used to denote the collective genome of the indigenous microorganisms of a multicellular or unicellular host. However, the term has also been used by Lederberg and others to signify an ecological community of commensal, symbiotic and pathogenic microorganisms.
Phylogenetic conflict
Differences between the evolutionary history of a species and the evolutionary history of its genes are embodied by discrepancies in branching order between the species and the gene tree.
transduction (FIG. 1). Conjugation requires physical contact between a donor and a recipient cell via a conjugation pilus, through which genetic material is transferred. Conjugation is canonically restricted to bacterial cells as the donor and recipient, however, Agrobacterium spp. is an exception and uses its conjugation machinery for HGT into plant cells1415. Transformation is the uptake of exogenous DNA from the environment and has been reported in both archaea and bacteria1617. Transduction is the delivery of genetic material through phage preda-tion owing to the integration of exogenous host genetic material into a phage genome, and this phenomenon has been observed in both bacteria and archaea. There are two types of transduction: generalized, in which a random piece of the host DNA is incorporated during cell lysis; and specialized, in which a prophage imprecisely excises itself from a host genome and incorporates some of the flanking host DNAs.
Other mechanisms of gene transfer, such as gene transfer agents (GTAs) and cell fusion, have more recently been described. GTAs are gene delivery systems that are integrated into a host chromosome and are sometimes under host regulatory control. GTAs carry small random pieces of host genome in capsids for delivery to nearby hosts. GTAs are found in both bacteria and archaea. The GTA-encoding genes do not provide an obvious benefit to the host, which donates its DNA to others, nor is the benefit to the GTA-encoding genes obvious, because the GTA does not preferentially transfer the GTA-encoding genes. The question of how these genes remain under selection for function remains enigmatic18. One study found that GTAs from Rhodobacter capsulatus were able to transfer antibiotic resistance to bacteria from different phyla; however, other studies have shown that not all bacteria, including those with the genes encoding GTAs, are able to receive gene donations via GTAs18. GTAs have evolved from prophages that have lost the ability to target their own DNA for packaging18. Most GTAs cannot package a long enough segment of DNA to transfer all the genes that are necessary to produce GTAs — that is, in contrast to phages, GTAs cannot transfer all of the genes that encode them to a new host. This is an important distinction from transduction.
Cell fusion has been observed in both Euryarchaeota (Haloferax spp.) and Crenarchaeota (Sulfolobus spp.)19'20. Experimentally, cell fusion has been observed on solid media where Haloferax volcanii forms aggregates and cells become physically joined by several small bridges of fused cell membrane21. Bidirectional gene transfer that is mediated through cell fusion has also been observed between different Haloferax species22. The bidirectional-ity of this method of gene exchange means that it is more similar to sexual reproduction in eukaryotes than it is to conjugation in prokaryotes.
Circumstances that facilitate HGT in eukaryotes. The development of the nucleus sequestered genetic material in eukaryotes made gene exchange a more complicated process, although physical association over extended periods of time can facilitate HGT. Obligate endosym-biosis as a stable form of physical association often leads
to the presence of foreign genes in eukaryotic genomes, as is the case for mitochondria and plastids, which are eukaryotic organelles that evolved from bacterial endo-symbionts10, and many other endosymbionts that have donated genetic material to their host genomes23. In the absence of an endosymbiotic partner, a congruent phylogenetic signal from multiple foreign genes has also been used to infer the presence of obsolete endosymbionts in plants and other photosynthetic eukaryotes25'30. Notably, however, genes of endosymbiotic origin are either absent or not obviously enriched in several eukaryotes that harbour endosymbionts24,26, suggesting that proximity alone is not enough to ensure successful HGT.
Feeding activities are also frequently linked to gene acquisition. The mechanism of the you are what you eat' gene transfer ratchet proposed by W Ford Doolittle suggests that many protists acquire genes through phagotro-phy27. This mechanism is consistent with the findings that phagotrophic microbial eukaryotes often harbour many foreign genes28'29.
The recently proposed weak-link model suggests that weakly protected unicellular or early developmental stages, especially in oviparous species, might constitute potential entry points for foreign genes into multicellular eukaryotes8. These foreign genes could then be spread through mitosis to germline cells, and thus to offspring. This model could potentially explain the fact that genes are frequently acquired in plants and animals that have eggs associated with endosymbionts or exposed to exterior environments (for example, mosses, Drosophila spp. and nematodes)23'31'32.
One way that genes can be exchanged between related species is through introgression — that is, gene flow due to interspecies hybridization followed by repeated backcrosses to one of the parent species. This mechanism is a major concern in transgenic crops that are grown in proximity to non-domesticated relatives33. Introgression of adaptive genes is not limited to plants. For example, introgression was inferred to have introduced an allele that is important in brain development from archaic to modern humans, and this transferred allele shows signs of being under positive selection in human populations34.
Detecting HGT
Methods for detecting HGT generally rely on phylogenetic conflict, that is, conflicting branching patterns between two gene trees; usually one of these trees is considered to be an accepted species or a reference tree. Often the reference tree is assumed to represent the vertical evolution of the organisms that are being analysed; however, detecting conflict between a gene tree and the reference tree that is not due to uncertainty in phylogenetic reconstruction is sufficient to infer the transfer of either the gene or the markers used to calculate the reference tree35. Deviations from the branching pattern of the reference tree identify potential HGT events, and provide information about the organisms between which genes were exchanged. Species trees are often built using well-conserved housekeeping or informational genes, such as ribosomal proteins. These genes are
NATURE REVIEWS I GENETICS
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS
a Conjugation
b Cell fusion
c Transduction
e Transformation
d Gene transfer agents
f Intracellular or endosymbiotic gene transfer
g Introgression
Population A
Population B
Backcrossing
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics
REVIEWS
■* Figure 1 | Mechanisms of gene transfer. Each panel represents a method of gene transfer. Conjugation (part a) occurs through donor-recipient cell contact, and single-stranded DNA is transferred from the donor cell to the recipient cell. Cell fusion (part b) differs from conjugation in that DNA is exchanged bi-directionally after cell contact and bridge formation between two cells. Gene transfer mediated by phage is known as transduction (part c). In the case of generalized transduction, any piece of genomic DNA may be loaded into the phage head; a general transducing phage is shown with host DNA (red). Specialized transduction occurs when an activated prophage loads a piece of genomic DNA neighbouring the prophage genome into the phage head together with the phage DNA (not shown). Gene transfer agents (GTAs) (part d) are phages that no longer recognize their own DNA and only carry random fragments of host DNA. Like prophage, they reside in the host cell genome. During transformation (part e) DNA is taken up from the surrounding environment; in the picture the DNA is depicted as entering the cell in the double stranded form, though many DNA uptake systems degrade one of the strands upon cell entry. Intracellular or endosymbiotic gene transfer (part f) occurs when genetic material from an endosymbiont or organelle (such as a chloroplast or mitochondrion) is incorporated into the host genome, this mainly pertains to eukaryotes. Introgression (part g) occurs when a hybridization event occurs between two diverging species (orange and blue populations). Backcrosses with one of the parent populations (orange) can lead to only a small piece of the divergent genome (blue) remaining in the recipient.
transferred less frequently between divergent organisms and can thus provide a good measure of vertical ancestry. Historically, the small subunit rRNA gene (SSU rRNA) has been used to determine the prokaryotic phylogeny. This practice was suggested to be problematic because several organisms have multiple divergent rRNA oper-ons, and it was reported that homologous recombination can occur between them (see REF. 36 for a review). Multi-locus sequence analysis (MLSA) has emerged as a supplementary method for determining prokaryotic phylogeny. The aim is to minimize the phylogenetic conflict that results from the transfer of one or more of the genes by concatenating many genes. However, if the individual genes are not screened for phylogenetic conflict caused by HGT between divergent organisms, the resulting MLSA tree might not represent either a single gene tree or the organismal evolutionary history5. Careful screening of genes used in an MLSA data set for significant phylogenetic conflict, and using a large number of genes (such as the suite of 50 ribosomal proteins), can help to mitigate this problem. Generally, within a phylum, phylogenetic trees that are generated using MLSA are in good agreement with those made using SSU rRNA and also provide better resolution at the species level37'38.
Quantification of bacterial and archaeal HGT is difficult because most transfers occur between closely related organisms and are difficult to distinguish owing to the genetic similarity of the host and the recipient genomes39"41. As mentioned above, the canonical method for detecting HGT events uses phylogenetic conflict comparing the gene history to the species history. Substantial and statistically supported conflict in the branching patterns of the gene and species trees can identify possible gene donors or the gene exchange partners if the direction of transfer cannot be interpreted. Gene duplication followed by differential gene loss is an alternative to HGT5; however, the more genome
sequences become available, the more independent gene loss events need to be postulated and the less parsimonious the differential gene loss scenario becomes compared with an HGT explanation. Gene composition (codon usage and oligonucleotide composition) provides a tool to identify HGT candidates42. Composition that is different from the genome average performs especially well to identify recent transfers from distantly related donors or from phages, which have a composition that is distinct from that of the recipient43. Generally, the sets of identified HGTs using each of these methods (composition or phylogenetic based) are complementary rather than redundant44.
The comparison of genomes from closely related organisms has identified large variation in gene content within a single species, especially in prokaryotic species. This variation in genome content reflects the ongoing process of gene gain and loss. Pan-genomes have been useful for studying the evolution of gene content in both prokaryotic species and genera. The pan-genome is defined as the set of all genes present in a taxon; the accessory genome contains genes that are present in only one or a few members of the taxon; and the core genome is the set of genes present in every member of the taxon. Each individual genome thus represents a sample from the pan-genome (BOX 1). An analysis of 61 Escherichia coli genomes revealed that only 6% of gene families were present in all genomes45. Pan-genomes were originally developed to explore the fluidity of prokaryotic genomes46; however, because HGT is more frequent between close relatives, the pan-genome also represents the set of genes that is potentially available via HGT to any member of the group. The eukaryotic pan-genome has been less extensively studied than the prokaryotic pan-genome, possibly because the impact of HGT is less well understood and the genomes are much larger. However, the pan-genome of Emiliania huxleyi, a globally distributed haptophyte phytoplankton species, has been studied. Although the accessory genome accounts for approximately one-third of genes present in the reference genome E. huxleyi CCMP1516, much of the variation in the pan-genome is related to intron tandem repeats and exon swapping, rather than HGT47. These data suggest that HGTs may be less frequent or that transferred genes may be less likely to persist in eukaryotes.
HGT in evolution
Mobile selfish genetic elements promote HGT. HGT enables innovations that evolved in one group of organisms to be shared across the web of life. Many HGTs provide a selective advantage to the recipient but, as described above, some transferred genes seem to be initially neutral or nearly neutral to the recipient. HGT of self-splicing selfish genetic elements such as introns and inteins provide examples of nearly neutral mobile genetic elements. Although the self-splicing activity minimizes the cost to the host organism, the additional DNA, RNA and protein synthesis associated with the selfish genetic element provide an additional burden to the host48. These elements persist because their success in invading new hosts compensates for the fitness cost to the host. Once
NATURE REVIEWS I GENETICS
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS
Genome streamlining
The reduction of genome size through relaxed selection and eventual loss of loci that are superfluous to the niche occupied by the organism.
Mobilome
The aggregate of mobile genetic elements in a genome, population or environment of interest.
Genome architecture imparting sequences
Strand-biased sequence motifs that are enriched towards the termini of replication; thought to direct proteins towards the termini.
established, these elements can provide material for variation, increased complexity and innovations. For example, in Saccharomyces cerevisiae the HO endonuclease, which evolved from an intein, functions as a mating-type switch cleaving at the MAT locus. Split inteins have become an integral part of synthesizing the DNA polymerase in marine picocyanobacteria. The group 2 introns evolved into spliceosomal introns, which now enable alternative splicing and fine-tuned regulation in most eukaryotes (see REF. 4 for a review). Thus, HGT disseminates beneficial, neutral and nearly neutral genes; subsequent selection can act on the variations that occur in the transferred genes, leading in some cases to their integration into cellular regulatory and metabolic networks.
Selfish genetic elements are commonly involved in promoting HGT and genome rearrangements, as well as facilitating the acquisition of genes that provide a selective advantage for recipients49. One example is the localization of antibiotic resistance genes in compound selfish elements such as plasmids, integrative conjuga-tive elements (ICEs) and even group 2 introns50. These compound structures can contain a large repertoire of genes with unrelated functions. Compound selfish elements are often associated with toxin resistance genes, metabolic genes, virulence factors and a wide range of secreted factors50. The acquisition of a useful gene repertoire could offset the cost of maintaining and transferring a large selfish element such as a conjugal plasmid. The traits carried on compound mobile elements can be used as a gene reservoir in times of adversity50'51. Genome streamlining is common in prokaryotic populations, and thus the mobility of adaptive genes associated with the mobilome becomes an important evolutionary strategy. Studies of the mobilome in different populations might provide information about the selective pressures (FIG. 2) that act on these populations and that influence gene distribution via HGT.
Selfish genetic elements are common in large multicellular eukaryotic genomes. Long terminal repeats
often flank selfish elements and have been frequently co-opted to either increase or decrease gene expression in different tissues52. Syncytin genes that have a key role in trophoblast cell fusion during placenta development were repeatedly derived from retroviral envelope protein genes52'53. In organisms with distinct somatic and germline cells, phenotypic ingenuity often depends on the result of changes in the copy number or expression of a gene, which are often the result of selfish element dynamics in the germ line54. These changes can lead to divergence among or within species.
Biased gene transfer and highways of HGT. Successful HGTs frequently occur between closely related organisms55, and the compositional similarity between the donor and the recipient genomes promotes homologous recombination that leads to homologous replacement with divergent alleles from close relatives. Additionally, the similarity between genome architecture imparting sequences in closely related organisms (same species or genera) leads to streamlined integration of the imported material56. In an analysis of 21 haloarchaeal genomes, over 90% of the HGTs identified through phylogenetic conflict were integrated into the recipient genome through homologous recombination39. The frequency of successful HGTs between pairs of Haloarchaea was shown to decrease exponentially with the phylogenetic distance (FIG. 3), probably due to the reduced efficiency of homologous recombination between genetically divergent organisms.
It was long thought that orthologous replacement through homologous recombination would be limited to the exchange of very similar gene sequences; however, the discovery of divergent isofunctional genes (known as homeoalleles) that can replace a divergent homologue in the recipient genome illustrated that homologous replacement can occur through homologous recombination in the conserved region flanking the divergent homeoalleles40. Divergent homeoalleles
Box 11 Pan-genome
This depiction (see the figure) of the pan-genome and core genome is based on Strain-specific
Edward's Venn cogwheel104, and was designed by O. Zhaxybayeva, Dartmouth College, USA. The pan-genome of a group refers to the sum of all the genes that are present in members of the group. Pan-genomes comprise the core genome, which comprises the genes found in all members of a group of interest, and the accessory genome — genes that are present in only one or a few members of the group. The concept of a pan-genome has led to the idea that steps in metabolic pathways may be distributed over several individuals within a community. The Black Queen hypothesis105 suggests that the combination of leaky functions — genes that produce a product that is shared with others in the community — combined with a selection for small genomes, will lead to a situation in which leaky functions are encoded in the genomes of only a fraction of community members that produce this function as a common good. The pan-genomes of many taxa seem to be open (that is, of an unlimited size)106-108, although the combination of limited population size and limited time of divergence from a common ancestor certainly limits the numbers of genes actually present in a given taxon. Estimated pan-genome sizes taking population size and divergence time into consideration can be large; for example, the Prochlorococcus pan-genome has been estimated to contain approximately 58,000 genes109, whereas the individual genomes of the members of this genus encode only about 2,000 genes each.
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics
REVIEWS
Ecotypes
Genetically distinct subsets of organisms within a population or species, usually genetic differences correspond to niche adaptation.
Holobiont
A multicellular or unicellular host and its collective symbionts.
Competition between holobionts (host plus symbionts) and between microbial communities (consisting of multiple species in a syntrophic relationship)
'Competition between groups
(groups that adapt or evolve faster outcompete
Rther groups)
Competition between individuals Igenes in organisms with higher fitness Increase in frequency in the population)
Gene-level selection (selfish genes that cooperate to construct a fit organism; parasitic genetic elements that may have a negative impact on host fitness)
Figure 2 | Nested levels of selection on gene content.
Each coloured box represents a different level of selection that can act on gene content.
of aminoacyl tRNA synthetases (aaRSs) provide an example of gene transfer that would go undetected by phylogenetic and compositional HGT detection methods. For many aaRSs, divergent forms evolved early in bacterial and archaeal evolution, and thus the diversity among aaRSs is easy to detect. The two or three forms with the same amino acid specificity frequently replace one another among both archaeal and bacterial species; however, because the transfers occur between related species, the gene tree of each type of aaRS remains in good agreement with the species tree40. Only the patchy distribution of each type reveals gene transfers and losses. Surprisingly, replacement with the divergent form was found to sometimes occur through homologous recombination in the more conserved flanking regions40.
The frequency and bias of HGT makes it difficult to understand how adaptations might be maintained in ecological niches that are in close physical proximity41. At least during the initial divergence of ecotypes, genes are transferred between organisms that are adapted to different niches. It is possible that the higher frequency of within-ecotype HGT than between-ecotype HGT maintains ecotype adaptation. However, genes that adapt an organism to a particular niche are also transferred between niche boundaries57, and such HGTs might help recipients to integrate into a new ecological niche (FIG. 4).
HGT enables key metabolic innovations. The enormous pan-genome size of many microbial species illustrates the importance of additive gene transfer, which is the process of the integration of novel genetic material into a genome. Integration into the genome can occur by nonhomologous recombination or through homologous recombination involving the genes neighbouring the transferred gene (for example, see REF. 41). An additive transfer from a close relative of a gene that has an ortho-logue in the recipient genome leads to two similar copies
being present in the recipient genome, an outcome that is similar to a gene duplication4. The methylaspartate cycle, for example, combines genes from several bacterial metabolic pathways that were transferred to the haloarchaeal ancestor from different bacterial donors and incorporated into a novel pathway for carbon assimilation58. Other examples of HGT contributing to the assembly or extension of metabolic pathways are acetoclastic methanogenesis in Methanosarcina spp. and the assembly of two photosystems functioning in series in oxygen-producing photosynthesis (see discussion in REF. 4 for details). In addition to frequently exchanging genes within and between genera, Haloarchaea also exchange genes with bacteria39'59. Haloarchaea are aerobic heterotrophs, although they evolved from metha-nogens — an anaerobic chemolithotrophic lineage. More than 1,000 genes were identified as imports from bacteria into Haloarchaea, including those for carbon assimilation, respiratory chain complexes, membrane transporters and cofactor biosynthesis59. The influx of these bacterial genes allowed the haloarchaeal ancestor to move into an aerobic environment. Similarly, the influx of bacterial genes to the ancestors of 12 other major archaeal clades is thought to have provided the key innovations to the origin of these groups60. Debate continues about whether the transferred genes originated from one or a few donors over a short period of time, or whether these transfers involved diverse bacterial donors112113. The limited distribution of these genes within single groups of archaea indicates that 'highways' of gene sharing between archaea and bacteria have promoted archaeal diversity.
HGT and the evolution of the holobiont
Many organisms rely on a complex network of symbionts for functions ranging from defence and immunity to metabolism. The symbiotic communities that are associated with larger macro-organisms provide an initial interface with the environment, thus new properties and physiological responses often occur through HGT involving these communities. The holobiont61 is used as a collective term for the host and its associated microbiota. For many multicellular eukaryotes, the number of genes in the microbiome62 (genes that are present in the microbiota) dwarfs the number of genes in the nuclear genome of the host and provides an important source of genetic diversity.
The composition of human gut microbiota is affected by the diet and ecology of the human host, and by competition between members of the microbiota62. For example, bacteria in the gut of Japanese people can break down polysaccharides from the cell walls of seaweeds that are commonly present in the Japanese diet. The genes encoding the polysaccharide-digesting enzymes were transferred from parasites of marine algae to the gut bacteria63'64. This HGT has enabled Japanese people to use carbohydrates from algal cell walls as a nutrient source, whereas other populations cannot. It is tempting to interpret this as selection acting on the holobiont; however, it is more likely to reflect gut bacteria evolving to fill an available ecological niche (FIG. 2).
NATURE REVIEWS I GENETICS
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS
10-
Cl
8
o o o o
o o
o o
^3 O O   O o
oo o v° °
o o o
§<§*
o cP,
i o o o o o
oo
.0000
oo
O O
O O
0 0.1 0.2 0.3 0.4 0.5
Evolutionary distance between lineages (substitutions per site)
Figure 3 | HGT is more frequent between closely related species. The frequency of horizontal gene transfer (HGT) events in haloarchaea is plotted against evolutionary distance. Gene transfers were detected through phylogenetic conflict between the gene's phylogeny and the reference phylogeny calculated from ribosomal proteins. HGTs between terminal edges of the reference phylogeny are shown in black and those between internal edges are shown in red. Similar inverse log-linear relationships between recombination rate and divergence were also observed for bacterial genera. Reprinted from Williams, D., Gogarten, J. P. and Papke, R. T. Quantifying homologous replacement of loci between haloarchaeal species. Genome Biol. Evol. (2012). 4,1223-1244 by permission of Oxford University Press.
The results of recent research on the human microbi-ome have demonstrated the importance of the microbiota in nutrient acquisition and immune defence in humans. In an analysis that investigated recent gene transfers in the human microbiome, HGT was shown to be 25-fold more frequent between pairs of human-associated organisms than between pairs of organisms in different hosts or in aquatic or terrestrial environments49. Moreover, HGT between pairs of human-associated organisms isolated from the same body site are 50-fold more likely to exchange genes than pairs from other environments49. The surprising extent of gene transfer in human microbiota compared with other environments could indicate that environmental fluctuations that promote frequent adaptive changes are more prevalent in holobiont ecology, especially in the human holobiont. Notably, however, quantification of HGT is difficult, and sampling bias between environments (in that particular study, for example, 53% of the samples were of the human holobiont and the remaining 47% were split between aquatic, terrestrial and other host-associated environments49) could falsely inflate the rate of detection of HGT in well-sampled environments (humans) compared with that in environments with less available data.
HGT in eukaryotic evolution
Although still fragmented, the available data indicate that HGT is widespread in all major eukaryotic groups and has been ongoing throughout evolutionary time7'8'65. As stated above, the sequestration of genetic material to the nucleus requires distinct mechanisms for HGT in eukaryotes. Nevertheless, HGT is important in conferring beneficial phenotypes that may lead to the origin of major lineages. Furthermore, changes brought about by HGT may prompt the adaptive radiation of other groups through organismal interactions and genetic integration in a co-evolving web of life.
HGT in the origin ofplastids and Plantae. The plant lineage is ripe with examples of HGTs that have conferred novel functions (FIG. 5). Plastids, the hallmark of photosynthetic eukaryotes, are derived from cyanobacterial endosymbionts in a eukaryotic host. With the only exception of chromatophores in amoeboid Paulinella spp., the well-founded belief is that all other photosynthetic eukaryotes trace their plastids to a single cyanobacterial endosymbiosis66. The transformation of a free-living cyanobacterium into a permanent organelle required both genetic and metabolic integration between the two partners. Several analyses identified 20-50 genes from chlamydiae, a group of obligate intracellular bacteria, in various photosynthetic eukaryotes30,67,68. These findings led to the suggestion that cyanobacterial and chlamydial endosymbionts coexisted in an early eukaryotic host cell, and that this tripartite relationship was responsible for the transformation of cyanobacterial endosymbionts into modern-day plastids30,67,69,70. Although it has been argued that these chlamydiae-related genes could have resulted from phylogenetic artefacts or could have existed in the cyanobacterial progenitor of plastids71"73, some of these genes are only adaptive in parasitic or heterotrophic bacteria and are not found in extant cyanobacteria, suggesting that chlamydial involvement in plastid establishment is plausible30,67,68,74. Non-cyanobacterial prokaryotes other than chlamydiae also contributed genes for plastid genesis and functionality69,75"77.
The establishment of cyanobacterial endosymbionts or plastids triggered the origin of Plantae: red algae, glaucophytes and green plants. Recent investigations have indicated that all three of these lineages have been affected by HGT during their evolution69,78"80. The glaucophyte Cyanophora paradoxa acquired more than 400 genes from bacteria69. In red algae, HGTs contributed to at least 5% of protein-coding genes in Galdieria sulphuraria and many others in Porphyridium purpureum7S'S0. Evidence of HGT has also been found in green algae79 and land plants81,82 (see below). For example, the moss Physcomitrella patens acquired genes from various sources, including fungi, bacteria, viruses and aquatic animals32,83,84. In most of these cases, acquired genes expanded the metabolic capabilities of recipients and had a key role in their adaptation to new environments, such as those with high salinity or acidity, extreme temperatures, or toxic substances.
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics
REVIEWS
HGT between plants and other eukaryotes. The origin of plastids and Plantae also spawned the emergence of other photo synthetic eukaryotes through secondary or higher-level endosymbioses. In addition, Plantae, which are rich in complex carbohydrates, generated new niches and resources for other organisms to exploit. Particularly, plant cell walls are the most abundant bio-mass on earth. Both the prevalence and novelty of this insoluble stored energy enhanced adaptive pressure to take advantage of novel resources free of competition. To effectively utilize plant biomass, other organisms often share genes or metabolic capabilities. For example, numerous soil bacteria reside in the rhizosphere and rely on root exudes as their primary nutrient source. An increase in exude production leads to active bacterial growth and thus more frequent plasmid transfer among rhizobacteria85. Choanoflagellates and rotifers, both of which live in aquatic environments, acquired numerous genes from plants and miscellaneous algae86'87, frequently related to complex carbohydrate degradation28. In rumen ciliates, 46 genes related to the degradation of complex carbohydrates, such as plant biomass, were acquired by HGT, many of them from the gut bacteria of ruminant animals88. Beyond choanoflagellates and rumen ciliates, the ability to degrade plant biomass has been independently acquired by many other eukaryotic groups such as oomycetes, fungi and nematodes89'90. The widespread and diverse mechanisms for degrading complex carbohydrates in plants in so many different lineages highlight the convergent evolution through HGT for adaptation.
Figure 4 | Structured exchange community. Prokaryotic members of two distinct niches are shown as green and red circles (A and B); grey circles (K and J) are related species occupying different niches. Genes that enable the adaptation of their hosts to these niches are mostly exchanged between members of the same niche (green and red arrows), but they might also be shared with recent niche invaders (blue circle; C), accelerating the adaptation of the invader to a new habitat. Adapted with permission from REF. 57, (AAAS).
Lepidopterans are the largest group of plant-feeding insects, and their diversification coincided with the emergence of flowering plants. In an analysis of HGT in lepidopteran insects, most of the acquired genes were shown to be distributed in multiple lepidopteran groups and related to nutritional metabolism and detoxification91. The production of toxins by plants and the corresponding genes for detoxification in lepidopterans, and other phytophagous arthropods, exemplifies a genetic arms race' fuelled by HGT. Many plants can produce cyanogenic glucosides, which can be converted to highly toxic hydrogen cyanide as a defence against herbivores. Conversely, phytophagous arthropods not only sequester hydrogen cyanide as a defence against their own predators, but also counteract cyanide poisoning through detoxification genes that were originally recruited from bacteria92.
HGT between multicellular eukaryotes. Many cases of HGT were reported between parasitic plants and their hosts93"96. In almost all of these cases, the direction of HGT is consistent with the direction of nutrient transfer from the host to the parasitic plant. HGT also occurs between multicellular eukaryotes with less obvious physical associations. For example, the moss P. patens acquired an actinoporin gene that is involved in desiccation resistance from metazoans83. Alloteropsis grasses switched to C4 photosynthesis at least four times in the past 10 million years through the acquisition of genes from other C4 grasses97. A photoreceptor gene was transferred from hornworts to ferns, allowing modern ferns to thrive in low-light conditions under the canopy98. Sturgeons, lampreys, which have been known to feed on sturgeons, and paddle fishes all share a transposable element, probably the result of HGT mediated by the exchange of fluids during lamprey feeding99. The sporadic distribution of type II antifreeze protein (AFP) genes in herring, smelt and sea raven was also mediated by HGT, allowing these fish to adapt to icy water31.
For a long time, mitochondria were considered uni-parentally inherited and subject to Muller's ratchet100. For many groups of organisms, this assumption seems to be correct101; however, plant, algal and fungal mitochondrial genomes are known to be dynamic and promiscuous, varying greatly among species in structure and gene content102. The transfer of mitochondrial genes between plant species can be massive and widespread. In an extreme case, Amborella trichopoda, a basal flowering plant, acquired at least four whole mitochondrial genomes from mosses and green algae, as well as many mitochondrial and, to a lesser degree, plastidal fragments from other flowering plants103. This example of HGT is not known to be associated with an adaptive benefit and is instead an important example of neutral or nearly neutral gene transfer in eukaryotes.
The mode of HGT between multicellular eukaryotes remains controversial. Are individual genes transferred, or are the transfers the consequence of between-species hybridization followed by backcrosses to one of
NATURE REVIEWS I GENETICS
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS
Eukaryotes
Land plants
Bacteria
Function: C4 photosynthesis
Unknown function Mitochondrial HCT
Function:
Phototropic response
Function:
• DNA replication and repair
• Pathogen and abiotic stress resistance
Function:
• DNA damage repair
• Vascular development
• Plant defence
• Stress tolerance
• Biosynthesis of starch, polyamines and hormones
• Cellulose degradation
• Pollen and seed germination
• Nutrient transport
Function:
• Plastid biogenesis
• Starch metabolism
• Alcohol fermentation
Function:
• Detoxification
• Environment adaptation
Figure 5 | HGT to the plant lineage. Arrows are coloured based on the origin of the gene transferred. Lines at the tips of the arrows indicate the gain of function for the plant lineage that acquired the genetic material. HGT, horizontal gene transfer. Figure modified from REF. 32, Nature Publishing Croup.
the parents7? In many instances, such as the transfer of AFP genes from herring to smelt31, donor and recipient diverged more than 200 million years ago, making hybridization an unlikely scenario. The conservation of introns between donor and recipient argues against independent transfers from bacterial symbionts. Sperm-mediated gene transfer between fish is one possible scenario31. In the moss P. patens, eggs and embryos that are exposed to bacteria and fungi in the environment might have facilitated gene acquisition. The large-scale acquisitions of mitochondrial genes in Amborella trichopoda probably occurred through mitochondrial genome fusion mediated by regenerated meristems from wounded areas.
Perspective
In this Review, we have discussed examples that illustrate how HGT shapes gene content in bacteria, archaea and unicellular eukaryotes (see Supplementary information SI (table)). Even in multicellular eukaryotes, HGT
from symbionts and between mitochondria occurs frequently and can have an important impact on gene content. Currently, we have a good understanding of the mechanisms by which prokaryotes exchange genes, including through GTAs and cell fusion in archaea; however, the mechanisms by which multicellular eukaryotes exchange genes with one another and with prokaryotes are less clear. The weak-link model, sperm-mediated gene transfer and introgression are possible gene transfer pathways, but more work is needed to explore the specific mechanisms involved. Importantly, comparisons between closely related strains will lead to a more accurate characterization of HGTs. Improvements in HGT detection based on the growing collection of sequence data will result in a more realistic estimation of HGT rates. However, accounting for false negatives and various types of transfer over different phylogenetic distances remains a challenge. Nevertheless, the surprising density of the web of life woven through genetic exchange is becoming visible.
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics
REVIEWS
1. Tatum, E. L. & Lederberg, J. Gene recombination in the bacterium Escherichia coii. J. Bacterioi. 53, 673-684(1947).
2. Went, F. W. Parallel evolution. Taxon 20, 197-226 (1971).
3. Treangen, T. J. & Rocha, E. P. C. Horizontal transfer, not duplication, drives the expansion of protein families in prokaryotes. PLoS Genet. 7, el 001284 (2011).
4. Swithers, K. S., Soucy, S. M. & Gogarten, J. P. The role of reticulate evolution in creating innovation and complexity. Int. J. Evoi. Biol. 2012, 418964 (2012).
5. Gogarten, J. P. & Townsend, J. P. Horizontal gene transfer, genome innovation and evolution. Nat. Rev. Microbiol. 3, 679-687 (2005).
6. Park, C. & Zhang, J. High expression hampers horizontal gene transfer. Genome Biol. Evol. 4, 523-532 (2012).
This paper examines the impact of expression level on the transferability of a gene in both environmental and laboratory populations of E. coli.
7. Boto, L. Horizontal gene transfer in the acquisition of novel traits by metazoans. Proc. Biol. Sci. 281, 20132450 (2014).
8. Huang, J. Horizontal gene transfer in eukaryotes: the weak-link model. Bioessays 35, 868-875 (201 3). This letter proposes a model for ongoing HGT in eukaryotes involving unicellular and early developmental stages to overcome the barrier of genome sequestration in eukaryotes.
9. Andersson, J. O. Gene transfer and diversification of microbial eukaryotes. Annu. Rev. Microbiol. 63, 177-193(2009).
10. Timmis, J. N., Ayliffe, M. A., Huang, C. Y. & Martin, W. Endosymbiotic gene transfer: organelle genomes forge eukaryotic chromosomes. Nat. Rev. Genet. 5, 123-135(2004).
11. Koonin, E. V. Darwinian evolution in the light of genomics. Nucleic Acids Res. 37, 1011-1034(2009).
1 2.  Crisp, A., Boschetti, C., Perry, M., Tunnacliffe, A. & Micklem, G. Expression of multiple horizontally acquired genes is a hallmark of both vertebrate and invertebrate genomes. Genome Biol. 16, 50 (201 5).
13. Riley, D. R. etal. Bacteria-human somatic cell lateral gene transfer is enriched in cancer samples. PLoS Comput. Biol. 9, el003107 (2013).
14. Norman, A., Hansen, L. H. & Sorensen, S. J. Conjugative plasmids: vessels of the communal gene pool. Phil. Trans. R. Soc. B 364, 2275-2289 (2009).
1 5.  Kyndt, T. et al. The genome of cultivated sweet potato contains Agrobacterium T-DNAs with expressed genes: an example of a naturally transgenic food crop. Proc. Natl Acad. Sci. USA 112, 201419685 (2015).
1 6. Johnston, C., Martin, B., Fichant, G., Polard, P. & Claverys, J.-P. Bacterial transformation: distribution, shared mechanisms and divergent control. Nat. Rev. Microbiol. 12, 181-196(2014).
1 7.  Chimileski, S., Dolas, K., Naor, A., Gophna, U. & Papke, R. T. Extracellular DNA metabolism in Haloferax voicanii. Front. Microbiol. 5, 57 (2014).
18. Lang, A. S., Zhaxybayeva, O. & Beatty, J. T. Gene transfer agents: phage-like elements of genetic exchange. Nat. Rev. Microbiol. 10, 472-482 (2012).
19. Naor, A. & Gophna, U. Cell fusion and hybrids in Archaea: prospects for genome shuffling and accelerated strain development for biotechnology. Bioengineered 4, 1 26-1 29 (201 3).
20. Schleper, C, Holz, I., Janekovic, D., Murphy, J. & Zillig, W. A multicopy plasmid of the extremely thermophilic archaeon Sulfolobus effects its transfer to recipients by mating. J. Bacterioi. 1 77, 441 7-4426 (1995).
21. Mevarech, M. & Werczberger, R. Genetic transfer in Halobacterium voicanii. J. Bacterioi. 162, 461-462 (1985).
22. Naor, A., Lapierre, P., Mevarech, M., Papke, R. T. & Gophna, U. Low species barriers in halophilic archaea and the formation of recombinant hybrids. Curr. Biol. 22, 1444-1448 (2012).
23. Dunning Hotopp, J. C. et al. Widespread lateral gene transfer from intracellular bacteria to multicellular eukaryotes. Science 317,1 753-1756 (2007).
This paper describes HGT between Wolbachia spp., an intracellular bacterial symbiont, and its multicellular eukaryotic insect hosts.
24. Nikoh, N. et al. Bacterial genes in the aphid genome: absence of functional gene transfer from Buchnera to its host. PLoS Genet. 6, el000827 (2010).
25. Moustafa, A. et al. Genomic footprints of a cryptic plastid endosymbiosis in diatoms. Science 324, 1724-1726(2009).
26. Chapman, J. A. et al. The dynamic genome of Hydra. Nature 464, 592-596 (2010).
27. 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).
28. Yue, J., Sun, G., Hu, X. & Huang, J. The scale and evolutionary significance of horizontal gene transfer in the choanoflagellate Monosiga brevicollis. BMC Genomics 14, 729 (2013).
29. Grant, J. R. & Katz, L. A. Phylogenomic study indicates widespread lateral gene transfer in entamoeba and suggests a past intimate relationship with parabasalids. Genome Biol. Evol. 6, 2350-2360
(2014) .
30. Huang, J. & Gogarten, J. P. Did an ancient chlamydial endosymbiosis facilitate the establishment of primary plastids? Genome Biol. 8, R99 (2007).
This paper discusses a complex tripartite relationship between a eukaryotic host, a cyanobacterium and a chlamydia that may have facilitated the establishment of modern plastids.
31. Graham, L. A., Li, J., Davidson, W. S. & Davies, P. L. Smelt was the likely beneficiary of an antifreeze gene laterally transferred between fishes. BMC Evol. Biol. 12, 190 (2012).
32. Yue, J., Hu, X., Sun, H., Yang, Y. & Huang, J. Widespread impact of horizontal gene transfer on plant colonization of land. Nat. Commun. 3, 1152 (2012).
33. Stewart, C. N., Halfhill, M. D. & Warwick, S. I. Transgene introgression from genetically modified crops to their wild relatives. Nat. Rev. Genet. 4, 806-817(2003).
34. Evans, P. D., Mekel-Bobrov, N., Vallender, E. J., Hudson, R. R. & Lahn, B. T. Evidence that the adaptive allele of the brain size gene microcephalin introgressed into Homo sapiens from an archaic Homo lineage. Proc. Natl Acad. Sci. USA 103, 18178-18183 (2006).
35. Williams, D. et al. A rooted net of life. Biol. Direct 6, 45 (2011).
36. Gogarten, J. P., Doolittle, W. F. & Lawrence, J. G. Prokaryotic evolution in light of gene transfer. Mol. Biol. Evol. 19, 2226-2238 (2002).
37. Colston, S. M. etal. Bioinformatic genome comparisons for taxonomic and phylogenetic assignments using Aeromonas as a test case. mBio 5 e02136-14 (2014).
38. Delamuta, J. R. M., Ribeiro, R. A., Menna, P., Bangel, E. V. & Hungria, M. Multilocus sequence analysis (MLSA) of Bradyrhizobium strains: revealing high diversity of tropical diazotrophic symbiotic bacteria. Braz. J. Microbiol. 43, 698-710 (2012).
39. Williams, D., Gogarten, J. P. & Papke, R. T. Quantifying homologous replacement of loci between haloarchaeal species. Genome Biol. Evol. 4, 1223-1244 (2012).
40. Andam, C. P. & Gogarten, J. P. Biased gene transfer in microbial evolution. Nat. Rev. Microbiol. 9, 543-555 (2011).
41. Polz, M., Aim, E. & Hanage, W. Horizontal gene transfer and the evolution of bacterial. 29, 1 70-1 75
(2015) .
This paper investigates the interplay between HGT, population structure and lineage divergence in bacteria and archaea.
42. Langille, M. G. I., Hsiao, W. W. L. & Brinkman, F. S. L. Detecting genomic islands using bioinformatics approaches. Nat. Rev. Microbiol. 8, 373-382 (2010).
43. Daubin, V. & Ochman, H. Bacterial genomes as new gene homes: the genealogy of ORFans in E. coli. Genome Res. 14, 1036-1042(2004).
44. Ragan, M. A. On surrogate methods for detecting lateral gene transfer. FEMS Microbiol. Lett. 201, 187-191 (2001).
45. Lukjancenko, O., Wassenaar, T. M. & Ussery, D. W. Comparison of 61 sequenced Escherichia coli genomes. Microb. Ecol. 60, 708-720 (2010).
46. Tettelin, H. et al. Genome analysis of multiple pathogenic isolates of Streptococcus agalactiae: implications for the microbial 'pan-genome'. Proc. Natl Acad. Sci. USA 102, 13950-13955 (2005).
47. Read, B. A. etal. Pan genome of the phytoplankton Emiliania underpins its global distribution. Nature 499, 209-213 (2013).
48. Barzel, A., Obolski, U., Gogarten, J. P., Kupiec, M. & Hadany, L. Home and away — the evolutionary dynamics of homing endonucleases. BMC Evol. Biol. 11, 324 (2011).
49. Smillie, C. S. etal. Ecology drives a global network of gene exchange connecting the human microbiome. /Vatare 480, 241-244 (2011).
This letter investigates the frequency of HGT in the human microbiome across body sites and across continents.
50. Rankin, D. J., Rocha, E. P. C. & Brown, S. P.
What traits are carried on mobile genetic elements, and why? Hered. (Edinb.). 106, 1-10(2011). This paper investigates the types of traits that are associated with compound selfish genetic elements and investigates the ecological scenarios that would select for specific types of traits.
51. Broaders, E., Gahan, C. G. M. & Marchesi, J. R. Mobile genetic elements of the human gastrointestinal tract: potential for spread of antibiotic resistance genes. Gut Microbes 4, 271-280 (2013).
52. Feschotte, C. & Gilbert, C. Endogenous viruses: insights into viral evolution and impact on host biology. Nat. Rev. Genet. 1 3, 283-296 (201 2).
53. Cornells, G. etal. Ancestral capture of syncytin-Carl, a fusogenic endogenous retroviral envelope gene involved in placentation and conserved in Carnivora. Proc. Natl Acad. Sci. USA 109, E432-E441 (2012).
54. Schaack, S., Gilbert, C. & Feschotte, C. Promiscuous DNA: horizontal transfer of transposable elements and why it matters for eukaryotic evolution. Trends Ecol. Evol. 25, 537-546 (2010).
55. Skippington, E. & Ragan, M. A. Phylogeny rather than ecology or lifestyle biases the construction of Escherichia coli-Shigella genetic exchange communities. Open Biol. 2, 120112 (2012).
56. Hendrickson, H. & Lawrence, J. G. Selection for chromosome architecture in bacteria. J. Mol. Evol. 62, 615-629 (2006).
57. Papke, R. T. & Gogarten, J. P. Ecology. How bacterial lineages emerge. Science 336, 45-46 (2012).
58. Khomyakova, M., Bukmez, O., Thomas, L. K., Erb, T. J. & Berg, I. A. A methylaspartate cycle in haloarchaea. Science 331, 334-337 (2011).
59. Nelson-Sathi, S. et al. Acquisition of 1,000 eubacterial genes physiologically transformed a methanogen at the origin of haloarchaea. Proc. Natl Acad. Sci. USA 109, 20537-20542 (2012).
60. Nelson-Sathi, S. etal. Origins of major archaeal clades correspond to gene acquisitions from bacteria. Nature 517, 77-80 (2014).
This paper suggests that acquisitions of genes from bacteria lead to the evolution of the major clades in archaea.
61. Guerrero, R., Margulis, L. & Berlanga, M. Symbiogenesis: the holobiont as a unit of evolution. Int. Microbiol. 16, 133-143 (2013).
62. Ley, R. E., Peterson, D. A. & Gordon, J. I. Ecological and evolutionary forces shaping microbial diversity in the human intestine. Cell 124, 837-848 (2006).
63. Hehemann, J.-H. etal. Transfer of carbohydrate-active enzymes from marine bacteria to Japanese gut microbiota. Nature 464, 908-912 (2010).
64. Thomas, F. etal. Characterization of the first alginolytic operons in a marine bacterium: from their emergence in marine Flavobacteriia to their independent transfers to marine Proteobacteria and human gut Bacteroides. Environ. Microbiol. 14, 2379-2394 (2012).
65. Hirt, R. P., Alsmark, C. & Embley, T. M. Lateral gene transfers and the origins of the eukaryote proteome: a view from microbial parasites. Curr. Opin. Microbiol. 23, 155-162 (2015).
66. McFadden, G. I. Origin and evolution of plastids and photosynthesis in eukaryotes. Cold Spring Harb. Perspect. Biol. 6, aOl 6105 (2014).
67. Ball, S. G. etal. Metabolic effectors secreted by bacterial pathogens: essential facilitators of plastid endosymbiosis? Plant Cell 25, 7-21 (2013).
68. Moustafa, A., Reyes-Prieto, A. & Bhattacharya, D. Chlamydiae has contributed at least 55 genes to Plantae with predominantly plastid functions. PLoS ONES, e2205 (2008).
69. Price, D. C. et al. Cyanophora paradoxa genome elucidates origin of photosynthesis in algae and plants. Science 335, 843-847 (201 2).
70. Cenci, U. et al. Transition from glycogen to starch metabolism in archaeplastida. Trends PlantSci. 19, 18-28 (2014).
71. Deschamps, P. Primary endosymbiosis: have cyanobacteria and Chlamydiae ever been roommates? ActaSoc. Bot. Pol. 83, 291-302 (2014).
NATURE REVIEWS I GENETICS
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS
72. Ku, C. et al. Endosymbiotic gene transfer from prokaryotic pangenomes: inherited chimerism in eukaryotes. Proc. Natl Acad. Sei. USA http://dx.doi. org/10.1075/pnas. 1421 58511 2 (2015).
73. Domman, D., Horn, M., Embley, T. M. & Williams, T. A. Plastid establishment did not require a chlamydial partner. Nat. Commun. 6, 6421 (2015).
74. Ball, S. G. etal. Toward an understanding of the function of Chlamydiales in plastid endosymbiosis. Biochim. Biophys. Acta 1847, 495-504 (2015).
75. Huang, J. & Gogarten, J. P. Concerted gene recruitment in early plant evolution. Genome Biol. 9, R109 (2008).
76. Suzuki, K. & Miyagishima, S. Y. Eukaryotic and eubacterial contributions to the establishment of plastid proteome estimated by large-scale phylogenetic analyses. Mol. Biol. Evol. 27, 581-590 (2010).
77. Qiu, H. etal. Assessing the bacterial contribution to the plastid proteome. Trends Plant Sei. 1 8, 680-687 (2013).
78. Schonknecht, G. et al. Gene transfer from bacteria and archaea facilitated evolution of an extremophilic eukaryote. Science 339, 1 207-1 210 (201 3).
79. Blanc, G. et al. The genome of the polar eukaryotic microalga Coccomyxa subellipsoidea reveals traits of cold adaptation. Genome Biol. 13, R39 (2012).
80. Bhattacharya, D. et al. Genome of the red alga Porphyridium purpureum. Nat. Commun. 4, 1941 (2013).
81. Yue, J., Hu, X. & Huang, J. Origin of plant auxin biosynthesis. Trends Plant Sei. 19, 764-770 (2014).
82. Yang, Z. etal. Ancient horizontal transfer of transaldolase-like protein gene and its role in plant vascular development. New Phyto!. 206, 807-816 (2015).
83. Hoang, Q. T. et al. An actinoporin plays a key role in water stress in the moss Physcomitrella patens. New Phytol. 184, 502-510 (2009).
84. Maumus, F., Epert, A., Nogue, F. & Blanc, G.
Plant genomes enclose footprints of past infections by giant virus relatives. Nat. Commun. 5, 4268 (2014).
85. Molbak, L, Molin, S. & Kroer, N. Root growth and exudate production define the frequency of horizontal plasmid transfer in the rhizosphere. FEMS Microbiol. Ecol. 59, 167-176(2007).
86. Sun, G., Yang, Z., Ishwar, A. & Huang, J. Algal genes in the closest relatives of animals. Mol. Biol. Evol. 27, 2879-2889 (2010).
87. Boschetti, C. etal. Biochemical diversification through foreign gene expression in Bdelloid Rotifers. PLoS Genet. 8, el003035 (2012).
88. Ricard, G. et al. Horizontal gene transfer from bacteria to rumen ciliates indicates adaptation to their anaerobic, carbohydrates-rich environment. BMC Genomics 7, 22 (2006).
89. Paganini, J. et al. Contribution of lateral gene transfers to the genome composition and parasitic ability of root-knot nematodes. PLoS ONE 7, e50875 (2012).
90. Richards, T. A. et al. Horizontal gene transfer facilitated the evolution of plant parasitic mechanisms in the oomycetes. Proc. Natl Acad. Sei. USA 108, 15258-15263 (2011).
91. Li, Z. W., Shen, Y. H., Xiang, Z. H. & Zhang, Z. Pathogen-origin horizontally transferred genes contribute to the evolution of Lepidopteran insects. BMC Evol. Biol. 11,356(2011).
92. Wybouw, N. et al. A gene horizontally transferred from bacteria protects arthropods from host plant cyanide poisoning, elite 3, e02365 (2014).
93. Yoshida, S., Maruyama, S., Nozaki, H. & Shirasu, K. Horizontal gene transfer by the parasitic plant Striga hermonthica. Science 328, 11 28 (2010).
94. Xi, Z. et al. Horizontal transfer of expressed genes in a parasitic flowering plant. BMC Genomics 1 3, 227 (2012).
95. Zhang, Y. et al. Evolution of a horizontally acquired legume gene, albumin /, in the parasitic plant Phelipanche aegyptiaca and related species. BMC Evol. Biol. 13,48(2013).
96. Zhang, D. et al. Root parasitic plant Orobanche aegyptiaca and shoot parasitic plant Cuscuta australis obtained Brassicaceae-specificstrictosidinesynthase-like genes by horizontal gene transfer. BMC Plant Biol. 14, 19 (2014).
97. Christin, P.-A. etal. Adaptive evolution of C4 photosynthesis through recurrent lateral gene transfer. Curr. Biol. 22, 445-449 (2012).
98. Li, F.-W. et al. Horizontal transfer of an adaptive chimeric photoreceptor from bryophytes to ferns. Proc. Natl Acad. Sei.       111, 6672-6677 (2014).
99. Zhang, H.-H., Feschotte, C, Han, M.-J. & Zhang, Z. Recurrent horizontal transfers of Chapaev transposons in diverse invertebrate and vertebrate animals. Genome Biol. Evol. 6, 1375-1386 (2014).
100. Blanchard, J. L. & Lynch, M. Organellar genes. Trends Genet. 16, 315-320 (2000).
101. Lynch, M. Mutation accumulation in transfer RNAs: molecular evidence for Muller's ratchet in mitochondrial genomes. Mol. Biol. Evol. 1 3, 209-220 (1996).
102. Palmer, J. D. et al. Dynamic evolution of plant mitochondrial genomes: mobile genes and introns and highly variable mutation rates. Proc. Natl Acad. Sei. USA 97, 6960-6966 (2000).
103. Rice, D. W. et al. Horizontal transfer of entire genomes via mitochondrial fusion in the angiosperm Amborella. Science 342, 1468-1473(2013).
This paper reports the acquisition of several mitochondrial genomes by the mitochondria in Amborella trichopoda a basal flowering plant.
104. Edwards, A. W. F. Cogwheels of the Mind: The Story of Venn Diagrams (JHU Press, 2004).
105. Morris, J. J., Lenski, R. E. & Zinser, E. R. The Black Queen Hypothesis: evolution of dependencies through adaptive gene loss. mBio 3, e00036-1 2 (2012).
This paper explains how the interplay between cheating and selection for streamlined genomes can give rise to shared genomic resources.
106. Lobkovsky, A. E., Wolf, Y. l.&Koonin, E. V. Estimation of prokaryotic supergenome size and composition from gene frequency distributions. BMC Genomics 15, SI4 (2014).
107. Lapierre, P. & Gogarten, J. P. Estimating the size of the bacterial pan-genome. Trends Genet. 25, 107-110 (2009).
108. Puigbo, P., Lobkovsky, A. E., Kristensen, D. M., Wolf, Y. I. & Koonin, E. V. Genomes in turmoil: quantification of genome dynamics in prokaryote supergenomes. BMC Biol. 12, 66 (2014).
109. Baumdicker, F, Hess, W. R. & Pfaffelhuber, P.
The infinitely many genes model for the distributed genome of bacteria. Genome Biol. Evol. 4, 443-456 (2012).
110. Hooper, L. V. & Gordon, J. I. Commensal host-bacterial relationships in the gut. Science 292, 1115-1118(2001).
111. Lederberg, J. & McCray, A. 'Ome sweet 'omics — a genealogical treasury of words. Scientist 1 5, 8 (2001).
11 2. Becker, E, A. et al. Phylogenetically driven sequencing of extremely halophilic archaea reveals strategies for static and dynamic osmo-response.. PLoS Genet. 10, e1004784 (2014).
113. Groussin, M et al. Origins of major archaeal clades do not correspond to gene acquisitions from bacteria. BioRxiv http://dx.doi: 10.1101/019851 (201 5).
Acknowledgements
Work in the authors' laboratories was supported through grants from the National Science Foundation Grant (DEB 0830024), NASA Exobiology (NNX1 3AI03G), Binational Science Foundation (BSF 2013061), NSFC Oversea, Hong Kong, Macao collaborative grant (31 328003), and the CAS/ SAFEA International Partnership Program for Creative Research Teams. The authors would also like to thank K. Swithers for providing insightful comments and discussion pertaining to the body of this text.
Competing interests statement
The authors declare no competing interests.
SUPPLEMENTARY INFORMATION
See online article: SI (table)
ALL LINKS ARE ACTIVE IN THE ONLINE PDF
© 2015 Macmillan Publishers Limited. All rights reserved
www.nature.com/reviews/genetics