REVIEWS

Evolution and ecology of plant viruses

Pierre Lefeuvre   1, Darren P. Martin2, Santiago F. Elena3,4, Dionne N. Shepherd5,
Philippe Roumagnac   6,7 and Arvind Varsani   8,9*
Abstract | The discovery of the first non-​cellular infectious agent, later determined to be
tobacco mosaic virus, paved the way for the field of virology. In the ensuing decades, research
focused on discovering and eliminating viral threats to plant and animal health. However,
recent conceptual and methodological revolutions have made it clear that viruses are not merely
agents of destruction but essential components of global ecosystems. As plants make up over
80% of the biomass on Earth, plant viruses likely have a larger impact on ecosystem stability
and function than viruses of other kingdoms. Besides preventing overgrowth of genetically
homogeneous plant populations such as crop plants, some plant viruses might also promote
the adaptation of their hosts to changing environments. However, estimates of the extent and
frequencies of such mutualistic interactions remain controversial. In this Review , we focus on the
origins of plant viruses and the evolution of interactions between these viruses and both their
hosts and transmission vectors. We also identify currently unknown aspects of plant virus ecology
and evolution that are of practical importance and that should be resolvable in the near future
through viral metagenomics.

Arthropods Viral metagenomics (or viromics) studies are shifting modes, and examine how these viruses adapt to new host
These invertebrate animals the focus from examining individual virus–host–vector species in changing environments. We suggest ways in
have exoskeletons, segmented systems towards characterizing the diversity and biol- which their capacity to adapt might foster the stability
bodies and paired jointed ogy of viruses in the context of entire environments. The of ecosystems and even provide direct benefits to host
appendages. Arthropods
belong to the phylum
genetic diversity of viruses dwarfs that of the cellular species, and we highlight ongoing challenges of plant
Euarthropoda that includes kingdoms of life1–5, and the present and past ubiquity virus ecology in a bid to fully comprehend the diversity
insects, arachnids, myriapods of viruses suggest that, rather than being entirely patho- and ecological roles of plant viruses.
and crustaceans. logical agents of destruction, viruses have likely been
performing key functions in global ecosystems since What are plant viruses?
the dawn of life3. When attempting to understand the polyphyletic origins
All viral contributions to plant ecosystem function of plant viruses10 it is perhaps helpful to realize first that
must derive from the complex interactions between the very concept of ‘plant viruses’ is problematic. Many
viruses, plants and transmission vectors that developed contemporary plant viruses infect the non-​plant organ-
during the rise of terrestrial plants, fungi and arthropods isms that ferry them between plants and most, if not all,
over the past 450 million years6,7. However, it is still purely plant-​infecting viruses have likely evolved from
uncertain for even the best-​studied interactions — that viruses that once infected non-​plant organisms. The
is, for viral pathogens of crops — how viruses affect the evolutionary history of eukaryotic viruses encompasses
diversity of plant species in natural environments. three main themes11: co-​evolution and co-​divergence
Specifically, their host ranges, their prevalence in of viruses with their hosts (for example, picorna-​like
uncultivated, alternative hosts and the mutualistic– viruses), evolution facilitated by horizontal gene trans-
antagonistic interactions with different hosts are almost fers between viral groups that infect distinct hosts (for
entirely unknown8,9. example, bunyaviruses) and parallel evolution of viral
Just as discoveries made with plant viruses laid the groups with related genetic elements (for example, cir-
foundations of virology and yielded crucial insights into coviruses and geminiviruses). Inter-​family gene trans-
numerous fundamental aspects of cellular and molecu- fers have been common, with many transfers occurring
lar biology (Box 1), understanding the ecological roles between viruses that now exclusively infect hosts in
of plant viruses will likely be pivotal in obtaining an different eukaryotic kingdoms3,12–15.
*e-​mail:
arvind.varsani@asu.edu accurate and robust understanding of global ecosystems. The closest non-​plant-infecting relatives of contem-
https://doi.org/10.1038/ In this Review, we address the evolutionary origins of porary plant viruses (green nodes in Fig. 1) tend to infect
s41579-019-0232-3 plant viruses in relation to their hosts and transmission fungi (brown nodes in Fig. 1) or arthropods (pink nodes

Nature Reviews | Microbiology

Reviews

Box 1 | History of plant virology transcriptome mining23,24 or as integrated sequences

within fungal genomes 25,26 has revealed that these
although plant virology as a discipline is only ~120 years old, historic literature viruses share several ancestors with plant viruses17. For
indicates that humans have been aware of plant virus diseases for over a millennium. example, plant-​infecting ourmiaviruses are apparently
a poem written in 752 ad by the Japanese empress Koken describes a striking evolutionarily related to members of the Narnavirdae
yellow leaf pattern that was subsequently confirmed to be caused by a geminivirus–
family, which replicate in the mitochondria27 of fungi
satellite disease complex149. another, perhaps more famous, early description of plant
virus diseases is tulip breaking disease150. the potyvirus tulip breaking virus causes an and plants28, and whose genomes encode only an RNA-​
elegant swirled colour pattern in tulip flowers, a highly valued commodity during the dependent RNA polymerase. Ourmiaviruses poten-
Dutch ‘tulip mania’ in the seventeenth century, but first described in 1576. aside from tially originated after an ancestral narnavirus acquired
inspiring poetry and instigating a speculative bubble that destroyed the Dutch genes that encode movement proteins and capsid proteins
economy, plant viruses are responsible for major scientific breakthroughs, starting through recombination from a presently unknown
with the science of virology itself. the exact date that the first virus was discovered is plant virus29. These additional genes may have enabled
difficult to pinpoint, but priority of discovery is generally given to russian scientist ourmia­viruses to escape from mitochondria so that they
Dimitri ivanovski, who first documented in 1892 the presence of a filterable, could systemically infect plants.
submicroscopic infectious agent in tobacco plants151. Dutch scientist Martinus The capacity of some fungal viruses to replicate in
Beijerinck was the first to use the term ‘contagium vivum fluidum’ or ‘virus’ for the
plant cells30 and the capacity of some plant viruses to
mosaic disease-​causing infectious agent in 1898 (ref.152). the discovery of this agent,
later called tobacco mosaic virus, then led to a series of scientific firsts, including replicate in fungal cells (in vitro and in vivo)31,32 suggest
confirmation of the nature and structure of viruses via paracrystallization153, electron that a pool of viral groups and/or species exists, possi-
microscopy154 and X-​ray crystallography155, and, finally, confirmation of the universality bly some that are presently defined as plant viruses, that
of the genetic code156. More recently, plant viruses were integral to uncovering the in fact infect plants and fungi. The strong association
mechanism of rNai or gene silencing, a discovery which revolutionized genetics in the between plant and fungal viruses may be in part due to
late 1990s157,158. initially recognized as a plant antiviral defence strategy, gene silencing the complex and evolutionarily ancient symbiotic associ-
or rNai is today known to be an almost universal mechanism of gene regulation and ations, ranging from facilitating nutrient uptake to stress
pathogen defence, functionally analogous across kingdoms159. Now, as we head for the tolerance, of terrestrial plants with fungi33–36.
third decade of the twenty-​first century, plant viruses are yet again at the forefront Further, the hypothesis that many plant and verte­
of a conceptual revolution. Driven by advances in metagenomics, the relatively
brate viruses may have originated from arthropod viruses
unbiased sampling and sequencing of viral genomes within entire environments is
changing the perspective of viruses from agents of disease to vital components of is supported by the astonishing diversity of viruses that
global ecosystems. has been discovered during arthropod transcriptome
studies37,38. Although some viruses that presently infect
plants may have been co-​evolving with their arthropod
RNA-​dependent RNA
in Fig. 1). In fact, even when discounting virus genera vectors since arthropods first evolved, it is also plausible
polymerases that contain both plant viruses and viruses that infect that some viruses that used to infect arthropods more
These enzymes catalyse the fungi or arthropods (for example, members of several recently switched to infecting plants. For example, flock
synthesis of RNA from an RNA genera in the Partitiviridae and Reoviridae families), house virus (in the Nodaviridae family) infects arthro-
template. RNA-​dependent
most viral genera that contain only plant viruses have pods but can also systemically infect plants when it is
RNA polymerases are essential
to the replication of viruses genes and/or genomic structures that are homologous complemented with the movement proteins of either
that have no DNA stage. with arthropod and/or fungal viruses16,17. For example, tobacco mosaic virus or red clover necrotic mosaic virus
the RNA-​dependent RNA polymerases of secoviruses are (both of which are plant viruses)39.
Movement proteins most closely related to those of dicistroviruses, ifla­ More viral sequence data will undoubtedly help
Some plant viruses encode
these proteins to facilitate
viruses, marnaviruses, picornaviruses and unclassified to refine hypotheses about the evolutionary origins
cell-​to-cell movement of viral viruses infecting plant-​parasitic nematodes and other of different plant viruses. For example, recent viral
particles and/or uncoated viral invertebrates18. metagenomic and metatranscriptomic studies37,38 have
nucleic acids. They frequently Viruses of fungi, commonly referred to as myco­ discovered numerous viral sequences (grey dots with
function by increasing the size
viruses, were only discovered in the early 1960s19, and red outlines in Fig. 1) and revealed previously unknown
exclusion limits of
plasmodesmata. have remained far less studied than animal and plant relationships between many viral groups (red edges in
viruses20. However, the discovery of new fungal virus Fig. 1). More data will also indicate how common it is
genomes through viral metagenomics21,22, through that a particular virus species infects hosts in different
kingdoms and whether transmission between hosts
in different kingdoms is epidemiologically and/or
Author addresses ecologically relevant.
1
CiraD, uMr PvBMt, st Pierre, La réunion, France. Further, the refinement of techniques to determine
2
Computational Biology Division, Department of integrative Biomedical sciences, exactly when homologous genome fragments of dis-
institute of infectious Diseases and Molecular Medicine, university of Cape town, tantly related viruses last shared a common ancestor
Cape town, south africa. would help to clarify the deep and frequently com-
3
instituto de Biología integrativa de sistemas (i2sysBio), CsiC-​uv, Paterna, valència, spain. plex evolutionary relationships of different virus fam-
4
the santa Fe institute, santa Fe, NM, usa. ilies. For example, there is strong evidence that many
5
research Office, university of Cape town, Cape town, south africa.
tobamoviruses (other than those that infect brassicas)
6
CiraD, uMr BGPi, Montpellier, France.
7
BGPi, CiraD, iNra, Montpellier supagro, university of Montpellier, Montpellier, France. likely co-​e volved and co-​diverged with their hosts
8
the Biodesign Center for Fundamental and applied Microbiomics, Center for evolution since angiosperms first arose40–42. Assuming that the
and Medicine, school of Life sciences, arizona state university, tempe, aZ, usa. most recent common ancestor of all tobamoviruses
9
structural Biology research unit, Department of integrative Biomedical sciences, existed ~130 million years ago, it has been possible to
university of Cape town, Cape town, south africa. infer the dates for the most recent common ancestors of

www.nature.com/nrmicro
 Reviews

Arthropod 7
Fungi 31
Bird
Fish 29 8
30
Plant
27 6
Mammal
Other (unknown) 28
9
Host group 26
19
Foreground 24 14
colour
23
25 10
Unclassified 13 5
sequences 18

22 11

Background 17
colour
Baltimore classification
16 3
dsRNA 4

ssRNA– 15

ssRNA+
21
dsDNA
12 1
ssDNA
20 2
RT-dsDNA

Viral families and genera

1 Amalgaviridae Crustavirus 13 Dicistroviridae 20 Bromoviridae 25 Nodaviridae
2 Quadriviridae Wastrivirus Iflaviridae Closteroviridae 26 Sinaivirus
Botybirnavirus 7 Aspiviridae Marnaviridae Virgaviridae
Secoviridae Blunervirus 27 Alphatetraviridae
3 Partitiviridae 8 Phenuiviridae Bacillarnavirus Cilevirus Carmotetraviridae
4 Endornaviridae 9 Tospoviridae Labyrnavirus Higrevirus Permutotetraviridae
14 Potyviridae Idaeovirus 28 Cyclovirus
5 Reoviridae 10 Fimoviridae
Peribunyaviridae 15 Papanivirus 21 Tymoviridae 29 Alphasatellitidae
6 Bornaviridae
Mymonaviridae 11 Caulimoviridae 22 Luteoviridae Circovirus
16 Benyviridae Nanoviridae
Nyamiviridae Solemoviridae
12 Alphaflexiviridae 17 Aumaivirus
Rhabdoviridae Tombusviridae 30 Geminiviridae
Betaflexiviridae
Anphevirus Gammaflexiviridae 18 Narnaviridae 23 Ourmiavirus Genomoviridae
Arlivirus Deltaflexiviridae 19 Virtovirus 31 Tolecusatellitidae
Chengtivirus 24 Alvernaviridae

Fig. 1 | The many origins of plant viruses. A network of shared genes between plant viruses and other viruses.
In the network , virus genera are represented with nodes of different sizes depending on the number of sequences in
the database. The colours of nodes indicate the most frequent host from which sequences were isolated (see key).
The background colours represent the Baltimore classification160 (that is, the type of genetic material of the virus
genera). Groups of genera with significant similarities are connected with edges of increasing width depending on
the protein similarity. The protein similarity network (minimum E-​value of 1 × 10−10, minimum bit-​score of 100 and
Markov Cluster Algorithm (MCL) inflation parameter of 2 as described in vConTACT161) is based on all reference
sequences described in the virus–host database162 (accessed on October 2018, dataset detailed online: Plant viruses
dataset). Importantly , all shown genera either contain plant viruses or viruses that express proteins with substantial
similarity to those expressed by a plant virus. Whereas some genus clusters contain only plant viruses (for example,
the Potyviridae and the Caulimoviridae family clusters), others contain genera that infect hosts from distinct
kingdoms. As a general trend, plant viruses cluster with either fungal viruses (brown nodes) or arthropod viruses
(purple nodes). Interestingly , the addition to this analysis of the flurry of recently discovered ’omics viruses (grey dots
circled in red37,38) link several clusters that otherwise appear unrelated (see red edges). For clarity , these new sequences
are grouped into clusters that display similar patterns of gene sharing, although many (if not most) of these still
unclassified sequences represent new genus-​level or family-​level taxonomic groupings. ds, double-​stranded;
RT, reverse transcriptase; ss, single-​stranded.

Nature Reviews | Microbiology

Reviews

Brassicas homologous genes that encode RNA-​dependent RNA enable the cell-​to-cell movement of viral particles and/or
Brassica is a genus in the polymerase and capsid proteins in tobamovirus and, for uncapsidated viral genomes50,51. Although this hypoth-
mustard family (Brassicaceae) example, viruses infecting charophyte algae14. esis of multipartitism as a response to plasmodesmata-​
of plants, which includes Despite the complexity of the evolutionary pathways imposed viral particle size constraints is appealing,
cabbage, lettuce and
cauliflower.
that have yielded the present diversity of plant viruses, supporting evidence is lacking52. Therefore, along with
even highly divergent and apparently unrelated plant the many unresolved evolutionary factors impacting
Angiosperms viruses share features that differentiate them from long-​distance intra-​plant virus movements53, the evolu-
Angiosperms are also other viruses. Among these features is the frequent tionary underpinnings of multipartitism in plant viruses
known as flowering plants
occurrence of multipartite genomes, which are packaged remain a mystery.
and are the most diverse
group of land plants. While
as individual genome components into separate viral
both gymnosperms and particles during transmission. Other viruses with mul- Transmission and spread
angiosperms produce seeds, tiple genome components, called segmented viruses, dif- Unlike animal viruses, plant viruses are only rarely trans-
angiosperms are characterized fer from multipartite viruses in that all of their genome mitted by direct contact between infected and unin-
by the presence of flowers, an
endosperm within the seeds
components are packaged into a single viral particle. fected individuals54–56. Although transmission modes
and the inclusion of seeds Whereas almost all known multipartite viruses infect can involve parasitic plants, natural root grafts or con-
within fruits. plants43, segmented viruses infect both plants and ani- taminated soil or water57, the most effective transmission
mals (for example, reoviruses and tospoviruses) or just modes are through vectors, pollen or seeds.
Plasmodesmata
animals (for example, orthomyxoviruses). Hypothetically, for any given plant virus, the relative
These microscopic channels
traverse plant cell walls
Moreover, many nominally monopartite plant viruses effectiveness of these transmission modes should depend
enabling intercellular trafficking associate with self-​replicating satellite nucleic acid mole­ on how strongly it harms the host. At one extreme, the
of macromolecules. cules (for example, begomoviruses and nanoviruses survival of virus species that are exclusively transmitted
associate with members of the Alphasatellitidae family44) through seed or pollen would demand that they mini-
Plasmodiophorids
This class of plant parasites
or satellite viruses, which they trans-​replicate (for exam- mally affect the reproductive success of infected plants
comprises organisms in the ple, tobacco necrosis satellite virus) and with which they or that they provide infected plants with some advantage
orders Plasmodiophorida and are frequently co-​transmitted45. These viruses and their over uninfected plants58,59. At the other extreme, vector-​
Phagomyxida. They have long satellites could represent intermediate stages in the transmitted viruses could be far more antagonistic,
been recognized as a basal
evolution of multipartitism. If the satellites eventually needing to only ensure that infected plants do not either
group to fungi, but recent
molecular phylogenetic
were to stop self-​replication and/or self-​encapsidation die too fast or die in such large numbers that the pool of
analysis suggests that they and become trans-​replicated and/or trans-​encapsidated potential uninfected hosts runs dry.
are more closely related to by the helper virus, they would essentially become The relative frequencies across all plant viruses of verti­
protozoa in the phylum components of multipartite genomes. cal transmission through seeds or pollen versus horizontal
Cercozoa.
Why plant viruses in particular show multipartitism transmission through vectors remain unknown. It is also
Hemiptera remains unknown. Similar to segmented viruses, multi­ unknown how common it is for viruses to be transmis-
This order of insects includes partite viruses can undergo a simple form of genetic sible by both vectors and seeds or pollen. Modern plant
insects such as aphids, cicadas, recombination called genome-​component reassort- virology has mainly focused on antagonistic virus–crop
leafhoppers and planthoppers.
ment. Although reassortment and other forms of genetic interactions, and this has undoubt­edly contributed to a
Most hemipterans feed on
plant sap with their sucking
recombination can potentially uncouple the fate of general underestimation of the role of vertical transmis-
and piercing mouthparts. genome regions carrying deleterious mutations from that sion in plant virus epide­mio­logy and ecology. Nonethe­
of the remainder of the genome46, such processes can also less, vector-​mediated transmission appears to be more
Haemocoel undermine the gene coadaptation within viral genomes47. common than other transmission modes and is therefore
This is the body cavity
in arthropods wherein
Another advantage of multipartite or segmented probably the most epidemiologi­cally and ecologically
haemolymph (plasma with genomes over monopartite or unsegmented genomes important mode of plant virus transmission60.
haemocytes) circulates. is the ability of dynamic changes in the relative copy Whereas plant virus vectors include arachnids, fungi60,
numbers of the genes encoded by different genome nematodes and some protists (plasmodiophorids)60,61,
components. This extra layer of gene regulation could, insects, belonging mostly to the order Hemiptera, transmit
for example, enable multipartite or segmented viruses more than 70% of known plant viruses62. Hemipterans
to respond more rapidly to environmental changes than are particularly well suited to transmitting plant viruses
viruses with monopartite or unsegmented genomes48,49. because of their needle-​like mouthparts that they use for
Whatever the advantages or constraints imposed by sucking sap and/or the contents of plant cells.
having multiple genome components, they should be Insects transmit plant viruses in two distinct ways63.
valid regardless of whether viruses infect plants, ani- The simpler of the two ways is variously called non-​
mals, bacteria, archaea or fungi. The main advantage of persistent, semi-​persistent or non-​circulative. It involves
multipartite genomes over segmented genomes in plants specific and reversible interaction of viral particles with
possibly relates to the size of viral particles. By packaging molecular components of mouthparts, with imbibed
each genome component into a separate particle, multi- sap promoting the binding of virions whereas excreted
partite viruses can increase their genome sizes without saliva promotes their release64. The second transmission
concomitantly increasing their particle sizes, which may mode is called circulative or persistent and the insect
make multipartitism particularly appealing for plant needs to acquire a virus only once for it to be able to
viruses that traffic between cells through plasmodesmata. transmit the virus during the remainder of its lifespan.
Indeed, another feature of currently characterized plant In such cases, ingested viral particles traverse the gut
viruses is the presence of movement proteins that epithelia and enter the haemocoel before moving to, and
increase the size exclusion limits of plasmodesmata to entering, the salivary glands. Sometimes, this circulative

www.nature.com/nrmicro
 Reviews

process can also involve interactions of virus particles

Begomovirus (n = 536)
with insect endosymbionts, or the uncoating and repli-
Crinivirus (n = 23) Bemisia tabaci Viruses vectored
Ipomovirus (n = 8) hosts by Bemisia tabaci cation of viruses in insect cells, in which case the viruses
might accurately be defined as being both insect and
Euphorbiaceae
Linaceae plant viruses65,66.
Rosaceae Many insect-​transmitted plant viruses have likely
Ulmaceae
Moraceae also evolved mechanisms to manipulate vector behav-
Urticaceae iour. They either make infected plants more attractive
Rhamnaceae
Cucurbitaceae to sap-​feeding insects than uninfected plants and/or
Fagaceae
Fabaceae ensure that infected plants produce chemicals that,
Polygalaceae upon ingestion by insects, promote insect behaviours
Anacardiaceae
Rutaceae that are most conducive to transmission67. In fact, seve­
Simaroubaceae
Moringaceae ral lines of evidence indicate that viruses that are non-​
Salvadoraceae persistently transmitted might manipulate host plants in
Brassicaceae
Capparaceae ways that encourage vectors to feed for shorter periods,
Caricaceae
Malvaceae whereas viruses that are persistently transmitted might
Melastomataceae manipulate plants to encourage vectors to feed for longer
Myrtaceae
Onagraceae periods63. Further, as is exemplified by pepper cryptic
Lythraceae virus 1 (a deltapartitivirus), some plant viruses that are
Vitaceae
Haloragaceae not transmitted by insects may manipulate their hosts
Scrophulariaceae
Lamiaceae to discourage potentially destructive insects such as
Verbenaceae aphids from feeding on them, presumably to promote
Acanthaceae
Pedaliaceae host survival68.
Oleaceae
Convolvulaceae The complexity of such interactions suggests long-​
Solanaceae term co-​evolution of plant viruses with specific com-
Rubiaceae
Apocynaceae binations of vectors and host species69. In contrast to
Gentianaceae
Asteraceae viruses that are non-​persistently transmitted by arthro-
Campanulaceae pods and which commonly have relaxed vector associ-
Araliaceae
Aquifoliaceae ations (for example, cucumber mosaic virus70), most of
Primulaceae the known plant viruses that are persistently transmitted
Ericaceae
Polemoniaceae by arthropods are transmitted effectively by only one or
Balsaminaceae
Caryophyllaceae a small number of closely related insect species, with the
Plumbaginaceae host ranges of these viruses generally corresponding to
Ranunculaceae
Berberidaceae the host ranges of their vector species71.
Commelinaceae
Zingiberaceae Plant viruses commonly have a host range that
Arecaceae includes several species from one or more different plant
Poaceae
Iridaceae families; however, they rarely are transmitted by more
Liliaceae than a few very closely related insect species71 (an exam-
Dioscoreaceae
Araceae ple of begomoviruses, criniviruses and ipomoviruses,
Alismataceae
Magnoliaceae all transmitted by the hemipteran Bemisia tabaci, is
Lauraceae provided in Fig. 2).
Piperaceae
Given the strong vector restriction of many persis-
tently transmitted viruses, it is likely that adaptation
Fig. 2 | overlap between the plant hosts of a whitefly vector species and to being transmitted by a new vector species may be
transmitted viruses. The taxonomic tree of the plant families that the whitefly a more difficult evolutionary challenge than adapta-
Bemisia tabaci163 feeds on and which of these families contains species that are tion to infecting a new plant host. Consistent with this
infected with viruses in the genera Begomovirus, Crinivirus and Ipomovirus that
hypothesis is the finding that the capsid protein phy-
this insect transmits (as defined in the virus–host database162). The number of
unique plant species–virus species pairs is indicated in brackets for each virus logenies of some plant virus families, such as the gemi-
genus. Notably , this figure represents only the known (and therefore the minimal) niviruses, mirror those of their vector species far more
host ranges of B. tabaci and the viruses in the three transmitted genera. Despite closely than those of their host species (Fig. 3). In fact,
this bias, it is evident that the known natural virus host ranges (that is, plants from strict vector specificity and the congruence of vector and
which viruses have been sampled in ‘the wild’) span much of the B. tabaci range. plant virus host ranges imply that the primary evolu-
Also striking is that a single vector transmits such a large number of distinct virus tionary imperative of most plant viruses that are persis-
species164. However, it is unlikely (although certainly not impossible) that an tently transmitted by one or a small number of vectors
individual begomovirus, crinivirus or ipomovirus has a natural host range that is not maintenance of optimal virus–host interactions
includes all of the plant species in all of these families. New plant virus species are but maintenance of optimal virus–vector interactions.
now being discovered at a rate that surpasses our capacity to biologically
The roles of viral capsids as structural proteins and as
characterize them. As a consequence, a decreasing proportion of described
plant viruses have either identified vector species or known host ranges. mediators of specific interactions between plant viruses
Metagenomic plant virus studies that focus on unbiased sampling of insects and their vectors may impose strong constraints on
and plants within ecosystems should, in the near future, be powerful enough the evolution of these proteins. These constraints may
to reveal the host ranges and vector species of many poorly characterized severely reduce the probability of viruses evolving to be
virus species. persistently transmitted by new vectors.

Nature Reviews | Microbiology

Reviews

Geminivirus genera
Becurtovirus
Begomovirus
Capulavirus Geminivirus capsid protein phylogeny
Curtovirus
Eragrovirus
Grablovirus
Mastrevirus PgaV MG001959
Topocuvirus CaCDaV MG452759
Turncurtovirus CCDaV KF561253
Unclassified PCMoV MG696802
MCaV KR131749
Capsid protein
similarity networks
Insect vector known

GraGV KX618694
JmaV MG001958

Insect vector unknown

Australian streak mastreviruses

ACSV KJ437671
RLV2 KY962381

MSMV MF167297 African streak

Geminivirus vector mastreviruses
SWSV KJ187745
Bemisia sp.
SStV-A KR150789
Nesoclutha sp.
Aphis sp.
SStV-D KX352044
Nesoclutha sp.
DfasMV JX458741
Cicadulina sp. SSMV2 JQ948052

Orosius sp. SSMV1 JQ948051

SMaV KF806701
Psammotettix sp.

Circulifer sp. Dicot-infecting

mastreviruses
SpSMV1 KY565231

Micrutalis sp.
Eurasian dwarf mastreviruses
Spissistilus sp.
EMSV JF508490
0.2 nucleotide MiSV E02258
substitutions/site
RLV1 KY962380
LaaV MG001961

Vector taxonomy
Family Subfamily Genus TaGV MF072689
Aleyrodidae Aleyrodinae Bemisia ToALCV MG491195
Aphididae Aphidinae Aphis
Cicadellidae Deltocephalinae Nesoclutha
Cicadellidae Deltocephalinae Cicadulina
Cicadellidae Deltocephalinae Orosius
Cicadellidae Deltocephalinae Psammotettix
Cicadellidae Deltocephalinae Circulifer
Membracidae Smiliinae Micrutalis 0.5 amino acid
Membracidae Smiliinae Spissistilus substitutions/site

www.nature.com/nrmicro
 Reviews

▶ Fig. 3 | co-​divergence of a viral capsid protein and transmission vector sequences. viruses can adopt a continuum of evolutionary strategies
The evolutionary relationships between capsid proteins of viruses in the Geminiviridae ranging from generalization to specialization. On the one
family are broadly congruent with the evolutionary relationships of the different insect hand, a virus species may undergo adaptive radiation,
species that transmit geminiviruses (in this case determined using cytochrome oxidase I which enables species to adapt to heterogeneous habi-
gene nucleotide sequences). This pattern strongly supports the hypothesis that the capsid
tats. Adaptive radiations occur when either different lin-
protein sequences of geminiviruses are co-​evolving with their vectors. Capsid protein
sequences were extracted from all available geminivirus full genome sequences (n = 6,760) eages within a species or different species within a given
before clustering them using CD-​HIT165. Only one exemplar sequence was retained from biome become adaptively specialized to more effectively
each cluster of sequences sharing >90% identity to leave only a subset of sequences use particular resources within that biome, while con-
representative of the entire known breadth of geminivirus capsid protein sequence diversity comitantly becoming maladapted to use alternative
(n = 276). These sequences were aligned using MAFFT166 with the satellite tobacco necrosis resources76. This type of specialization can minimize
virus capsid protein sequence as an outgroup and used to construct a maximum likelihood competition between the different lineages or species
phylogenetic tree with PhyML 3.0 (ref.167) with the LG+I+G amino acid substitution model. and result in highly polymorphic viral populations.
The cytochrome oxidase I gene nucleotide sequences of known geminivirus vector species On the other hand, generalist viruses that infect multiple
were obtained from the GenBank168 and BOLD169 databases and aligned with Nemouridae host species (such as cucumber mosaic virus77,78) have
sequences (in the order Plecoptera) as outgroups using MAFFT166. This alignment was used
access to a larger array of resources but compete with
to infer a maximum-​likelihood tree using PhyML 3.0 (ref.167). Branches with less than 0.8
approximate likelihood ratio test support were collapsed in both the geminivirus capsid
other viruses for these resources. Such a situation should
protein and the cytochrome oxidase I gene phylogenetic trees. The geminivirus capsid yield a low-​diversity viral population dominated by one
sequence similarity networks for different groups of geminiviruses were generated from or a few of the best-​adapted viral genotypes76.
the representative 276 sequence dataset using the Enzyme Function Initiative-​Enzyme Obviously, whether a virus evolves as a specialist or
Similarity Tool170 with an alignment score threshold of 60 and minimum E-​value of 1 × 10−5. as a generalist will strongly depend on the feeding pref-
Note that one of the geminivirus genera, Mastrevirus, has viruses with capsid proteins so erences of its vectors79,80. Viruses that have generalist
diverse that they fall into seven distinct clusters and three singletons, whereas capsid vectors or several vectors should have more opportu­
protein sequences from the other genera all fall within individual clusters. ACSV, Axonopus nities to infect different plant species. In this situation,
compressus streak virus; CaCDaV, Camellia chlorotic dwarf-​associated virus; CCDaV, the virus may evolve as a generalist or as a specialist.
Camellia citrus chlorotic dwarf-​associated virus; DfasMV, dragonfly-​associated mastrevirus;
By contrast, viruses that have specialist vectors will
EMSV, Eragrostis minor streak virus; GraGV, grapevine geminivirus; JmaV, Juncus
maritimus-​associated virus; LaaV, Limeum africanum-​associated virus; MCaV, mulberry encounter a limited number of potential hosts and might
crinkle-​associated virus; MiSV, Miscanthus streak virus; MSMV, maize streak Reunion virus; therefore be expected to evolve as specialists. For exam-
PCMoV, passion fruit chlorotic mottle virus; PgaV, Polygala garcinii-​associated virus; ple, although aphids are quite specific with regards to
RLV1, rice latent virus 1; RLV2, rice latent virus 2; SMaV, switchgrass mosaic-​associated virus; which plants they feed on, they probe every plant that
SpSMV1, sweetpotato symptomless mastrevirus 1; SSMV1, Sporobolus striate mosaic they encounter, thus facilitating the transmission of
virus 1; SSMV2, Sporobolus striate mosaic virus 2; SStV-​A , sugarcane striate virus A ; non-​persistent viruses such as the generalist cucumber
SStV-​D, sugarcane striate virus D; SWSV, sugarcane white streak virus; TaGV, tomato-​ mosaic virus81. Conversely, the acquisition by aphid vec-
associated geminivirus; ToALCV, tomato apical leaf curl virus. tors of phloem-​restricted viruses requires that the insects
feed on virus-​infected tissues. Possibly as a consequence
of this requirement, viruses such as barley yellow dwarf
Although the inherent genetic plasticity of viruses virus have specialized for transmission by a particular
may predispose them to infect new host species and to aphid species and, by extension, for infecting the plant
eventually emerge as pathogens, it is likely that many species that its aphid vector feeds on82.
emergence events have been facilitated by the existence In keeping with the old adage that a ‘jack of all trades
of arthropod vectors with very broad host ranges (that is, is a master of none’, it has been suggested that evolution
polyphagous vectors)72–74. Accordingly, several emerging should favour specialists because of expected trade-​offs
crop diseases are attributed to plant viruses that are vec- that limit the fitness of generalists in their alternative
tored by invasive polyphagous arthropods (for example, hosts83,84. One of the reasons for such trade-​offs, called
begomoviruses; Fig. 3). During their invasive spread, antagonistic pleiotropy, might arise when mutations
these vectors provide the plant viruses that they trans- that are beneficial in one host are deleterious in another
mit with an expanded menu of potential hosts, some of host85. A fitness trade-​off might also arise if neutral
which may be important crop species. Such new host mutations accumulate by random genetic drift in genes
encounters are a vital first step in expansion of the virus that are dispensable in the current host but are useful in
host range and the emergence process. other hosts86. Given the compactness of viral genomes, in
particular those of small RNA and single-​stranded DNA
Host-​range evolution viruses, which have many multifunctional proteins, this
Aphids Many viruses rapidly adapt to changes in host genotypes75 second mechanism seems less plausible than the first.
These small sap-​sucking after new host encounters, which, in some cases, results in However, our current knowledge of viruses in their
insects are members of the epidemics. Within a host, viruses evolve towards optimal natural ecosystems87–89 is still too limited to satisfac­
superfamily Aphidoidea in the
virus–host interactions that enable efficient viral repli- torily establish how each of these mechanisms explains
Hemiptera order.
cation and movement inside the host. To maximize the the observed host ranges of plant viruses. Although
Phloem overall fitness of a virus, evolution must constantly tweak metagenomic studies are providing valuable information
The phloem is the vascular the efficiencies of viral replication, movement inside the on the prevalence of different viruses in different host
system in plants within which host and transmission inside the heterogeneous host species, this information is not yet linked to quantitative
soluble organic compounds
that are produced during
populations in which the virus circulates. measures of viral fitness in each host.
photosynthesis are When confronted with complex environments con- To compensate for this lack of observational data,
transported. taining multiple host plant species or genotypes, plant experimental evolution in controlled greenhouse

Nature Reviews | Microbiology

Reviews

conditions is being increasingly used to study aspects infections of resistant hosts are rare; or viruses infect-
of host-​range expansion83,84,90,91. Of note, in these exper- ing resistant hosts tend to not be transmitted further.
imental studies, evolution in a single new host always Alternatively, one might also argue that in natural settings
increases fitness in that host but, in most cases, con- the conditions that favour the emergence of specialists,
comitantly decreases fitness in the original host. At the such as serial passages through genetically homogeneous
molecular level, multiple cases of convergent evolution host plants without any competition from other viruses,
have been described, often involving the same mutations would occur infrequently. Unfortunately, we presently
in several independent experiments83,91–93. This suggests have no data on the relative frequencies of specialist and
that trade-​offs may occur as a consequence of a limited no-​cost generalist viruses in any ecosystems.
number of alternative evolutionary pathways towards Nevertheless, even if the appropriate conditions are
adaptation to a novel host, such that adaptive mutations only infrequently met outside the laboratory, it is prob-
in a novel host are maladaptive in the original host. able, given the size and diversity of environments on
By contrast, if viruses are alternatively passaged Earth, that both types of viruses would naturally occur.
through several host species, they will frequently evolve Specifically, the diversity and distribution in the time and
as ‘no-​cost generalists’85. These generalist viruses are as space of hosts and viruses in a given environment will
fit as specialist viruses in all hosts, without paying the determine how viruses use the resources in that environ-
fitness cost predicted by the ‘jack of all trades but mas- ment91. Whether specialists, no-​cost generalists or inter-
ter of none’ hypothesis. In other words, under fluctua­ mediates between these two extremes are most common
ting host conditions, ‘master of all’ viruses can evolve. in nature will hopefully be revealed in the near future
Although the evolution of no-​cost generalists during by ecosystem-​scale viromics studies that are designed
such experiments has been frequently observed in ani- to quantitatively evaluate both the distributions of indi-
mal viruses94–96, only a single instance has been reported vidual viral lineages in different host species (preferably
for plant viruses, in which lineages of tobacco etch virus with enough resolution to differentiate between geno-
were evolved in Solanaceae hosts97. Whereas single-​host types) and the relative titres of different viral variants at
virus lineages evolved as specialists and displayed the host sites from which the viruses are acquired by vectors.
expected fitness trade-​offs, lineages alternatively pas-
saged between hosts evolved to be as fit as the single-​ Ecology of plant viruses
host lineages in their corresponding hosts, thus behaving The pervasive perception that plant viruses are pri-
as true no-​cost generalists. Interestingly, all generalist marily pathogens has meant that most plant virology
lineages fixed convergent mutations. The fact that some research has focused on the causes and consequences of
of these mutations were also detected in single-​host lin- virus pathology. Within the pathology field, emerging
eages suggests that non-​epistatic host species-​specific diseases have garnered most attention because they
mutations were responsible for the observed fitness cause the most damage to economically important food
increases in the alternative hosts. and ornamental plant species. Key examples of viruses
In other experimental evolution studies, viruses that are responsible for well-​studied emerging diseases
were passaged through differentially virus-​resistant are cassava-​infecting begomoviruses103, closteroviruses
genotypes of a single host species that differed in their causing grapevine leafroll disease104, luteoviruses such
level of virus resistance84. Some reports have evaluated as barley yellow dwarf virus105 and sobemoviruses such as
natural variations in the resistance of host genotypes rice yellow mottle virus106.
to virus infection98,99, whereas others have focused After decades of inventorying, tracking and analysing
on artificially engineered virus resistance genes100,101. plant viruses107, it is now apparent that the emergence of
In both cases, fitness trade-​offs apparently depend on new diseases following changes in viral host ranges is
the particular host genotypes in which they evolve, with driven by adaptive viral evolution in response to novel
more-​resistant host genotypes selecting highly virulent ecological conditions108. These novel conditions include
generalists and more-​permissive host genotypes select- the introduction of viruses108 and vectors to new areas, the
ing low-​virulence specialists. These differences could be intensification of agriculture and urbanization, and eco-
partially explained by the differential effects that each logical changes in response to changing climatic con-
type of viruses has on the transcriptional regulation of ditions109,110. In agriculture, which accounts for almost
host genotypes102. Whereas generalists influence the all studies of plant virus emergence, emergence is likely
same set of transcripts in all host genotypes, resulting facilitated by persistent, mostly dead-​end ‘spillover’
in a relatively homogeneous response, specialists influ- transmissions of viruses from uncultivated plant spe-
ence different sets of transcripts in each host genotype, cies to genetically homogeneous crop populations at so-​
resulting in heterogeneous responses. called ‘agro-​ecological interfaces’ between cultivated and
If no-​cost generalists can experimentally evolve in natural ecosystems109,110. At such interfaces, the detected
resistant plants or following passages through different proportion of infected individuals (a proxy for virus
hosts, the question arises of why specialists exist in ecosys- prevalence) has been found to be higher for cultivated
tems that commonly contain both several potential host species than for uncultivated species8,111,112.
species and individual plants with natural antiviral resist- Although the frequency of detected emergence
ance. A possible answer is that access to resistant hosts events has increased concomitantly with the pace with
that select for generalist viruses may be restricted under which natural biomes have been replaced by managed
natural conditions for any combination of the following croplands109,113,114, the ‘emergence potential’ of viruses
reasons: resistant hosts occur infrequently; successful is a natural property of virus populations, which likely is

www.nature.com/nrmicro
 Reviews

crucial for maintaining the stability of unmanaged eco- most plant virus infections are essentially harmless to
systems. The ecological impacts of plant–virus interac- host plants, however, has remained largely untested and
tions may be similar to those of interactions between is still controversial131. Viruses are, after all, parasitic in
lytic viruses and their bacterial hosts, in which viruses their reliance on host-​provided enzymes and cofactors
limit the outgrowth of genetically homogeneous bac- for their replication, and the existence of evolutionarily
terial populations during ecological competitions115. ancient antiviral defence strategies throughout the plant
In stable unmanaged ecosystems, plant virus emergence kingdom132 implies that plants would rather remain
events may preserve the genetic richness of ecosystems uninfected. Accordingly, although some viruses cause
by preventing them from being overgrown by genetically fewer and/or less severe symptoms in uncultivated hosts
homogeneous plants and, thereby, may foster the capacity than in cultivated hosts, viruses still generally inflict
of ecosystems to endure environmental changes116,117. measurable harm on their uncultivated hosts133–136.
In the context of invading exotic plants, the role of Demonstrating that a virus benefits a host is diffi-
viruses can be mixed. On the one hand, they can facil- cult as the virus and the host do not exist in isolation
itate invasion if an invasive plant species has fewer nat- in an ecosystem. Instead, they are part of a holobiont,
ural pathogens than the indigenous plant species (the encompassing the host plant and all bacteria, archaea,
enemy release hypothesis118) or if the invasive species eukaryotes and viruses that it interacts with137,138. In this
induces an increase of virus loads in the indigenous plant regard, the benefit or harm of a virus for a host prob-
species with which it competes119,120. On the other hand, ably depends on how direct and indirect interactions
viruses may help prevent invasions if indigenous plant between the virus and other members of the holobiont
species carry viruses that infect the invader and are vir- community affect the composition of the holobiont and,
ulent enough to prevent the invader from overrunning, ultimately, the fitness of the host.
or even surviving long term in, the ecosystem121. Interactions between hosts and parasites or sym-
When a virus emerges in a new host, the amount bionts can be divided into two broad categories: pre-​
of harm that the virus inflicts on its new host will, infection interactions, which affect the susceptibility
initially at least, not be at an optimal level for onward and/or exposure of a host to a second parasite or sym-
transmission. Often, an emergent virus causes severe biont; and post-​infection interactions, which affect the
disease symptoms owing to maladaptation to its new quality of within-​host nutrient resources for, or host
host122. Various trade-​off hypotheses123,124 predict that immune responses to, a second parasite or symbiont139.
the basic reproductive number of a virus (that is, how For example, a virus infection might reduce the sus-
many new hosts a virus will, on average, be transmit- ceptibility of plants to mammal herbivory140 or fungal
ted to) will be maximized if the virus can balance its infections139. Although several greenhouse studies sug-
transmission rate against the duration of infection, the gest that virus infection might slow plant desiccation
virus-​induced mortality and the reproductive capaci- during droughts141–144, others have shown that heat,
ties of hosts. For a virus that mainly infects artificially drought or salt stressors increase plant susceptibility to
managed crops, the lifespans and/or reproductive capac- viral pathogens145 or that virus infection has no effect
ities of infected plants might not affect the size of future on drought tolerance146. Also, even when elements of
host populations, simply because farmers will ensure a plant–virus interactions are genuinely symbiotic, the
constant stream of susceptible plants. In such cases, the degree of mutualism versus pathogenicity can depend
evolutionarily optimal amount of harm a virus inflicts both on the host and virus genotypes and on the inten-
might be much higher than if the virus mainly infected sity of the environmental factor that the virus provides
uncultivated plants in natural ecosystems. some resistance against91,147.
As intuitively appealing as such trade-​off hypotheses
are, few reports that test these for plant viruses exist125 Conclusions
and the outcomes for viruses with different host-​range Our rapidly expanding view of viral diversity is illumi-
sizes and/or transmission modes are unclear 126,127. nating the deep evolutionary relationships of viruses3,148
For example, viruses with broad host ranges that include and indicates that an appreciable fraction of many plant,
both cultivated and uncultivated hosts might optimally fungal and arthropod viruses share common origins16.
have high virulence in their cultivated hosts and low The future large-​scale sampling and analysis of viral
virulence in their uncultivated hosts, but be constrained genomic sequence data from diverse environments will
to have intermediate virulence across all their hosts. undoubtedly reveal additional evolutionary relation-
Alternatively, high virulence might be more beneficial ships, which may reveal the distant origins of all of the
for viruses that are primarily transmitted through soil major virus groups (Fig. 1).
or water following the death and decay of plants than With proper design, these large-​scale virus-​sequencing
it is for viruses that are primarily vertically transmitted projects could also yield insights into the roles of viruses
through seeds or pollen. Therefore, in practice, the opti- in different ecological contexts. Of particular interest
mal degree of virulence that a virus should have in a are the collateral impacts of human activities on these
given host will depend on the relative epidemiological ecological roles. Although there are very few terres-
importance of that host as a reservoir for onward trial environments that have remained uninfluenced by
transmission and the different transmission modes. human activities, the degree to which ecosystems have
Empirically, in natural ecosystems at least, the opti- been disturbed by humans ranges from minor for mostly
mal degree of harm that plant viruses inflict on their pristine preserved environments through to extreme for
hosts is probably very low128–130. The hypothesis that industrial wastelands. Crucially, relatively undisturbed

Nature Reviews | Microbiology

Reviews

Towards understanding the ecology and evolution of plant viruses at different tiers

Ecosystem ecology
Impact of
O=C=O • abiotic stress on virus–vector–host dynamics
• changes in land use on viral dynamics
• climate change
• expansion of geographical range of arthropod vector
• legal and illegal trafficking of infected plants
• spillover of agriculturally relevant viruses to endemic plants and vice versa
Viral community ecology
Impact of
• co-infections (rapid progression to disease or protection)
• recombination and reassortment
• emergence, evolution and selection of viral subgenomic and satellite molecules
• dynamics of cross-kingdom infecting viruses (plants, arthropods and fungi, in
particular)
• shifts in viral ecological networks

Virus–host–vector interactions
Understand
• viral evolution and adaptation to host and/or vector
• fitness of virus in host
• fitness of virus in vector and trade-offs with fitness in host
• transmission of viruses (arthropod, seed or pollen)
• symbiosis type with the host and vector (antagonism to mutualism)

Fig. 4 | Addressing important questions in the ecology and evolution of plant viruses. High-​throughput sequencing
approaches, including single-​molecule sequencing, coupled with traditional viral pathology methods can be used to study
interactions up to the ecosystem scale. In general, these methods are becoming common for the rapid identification of
plant viruses and, in certain cases, involve sequencing ‘in the field’ with portable sequencing machines such as the MinION
(Oxford Nanopore Technologies). In the next decade, these methods will enable rapid mapping of viral sequence spaces
and illuminate spatial and temporal virus–host–vector dynamics. Among other things, this will reveal spillover of viruses
from agro-​ecosystems to endemic wild flora (and vice versa), tentative expansions of virus host and/or vector ranges over
time and exposure of plants in a specific region to imported viruses. On a broader scale, next-​generation metagenomic
studies should enable ecosystem-​scale analyses of interaction networks between viruses, hosts, vectors and other
microscopic and macroscopic flora and fauna in the environment. One of the big questions that such analyses could
answer is whether it is possible to construct and apply synthetic microbial communities of beneficial viruses, bacteria and
fungi to protect natural ecosystems from being overrun by invasive viruses, plants and insects, and to reduce the use of
chemicals in modern agriculture.

ecosystems frequently occur side by side with moderately relative epidemiological and evolutionary importance of
disturbed agro-​ecosystems. Viromics studies at these different host and vector species for individual virus spe-
agro-​ecosystem interfaces that include the collection of cies, such studies could also expose epidemiological links
plant tissues, trapped or captured insects and soil over between viruses associated with different plants, fungi and
ecologically relevant areas and over multiple time points arthropods. We anticipate that recent and future method-
will likely prove crucial for understanding the impacts of ological and analytical innovations will enable the design
ecological disturbances on the population and evolution- and implementation of plant viral metagenomics studies
ary dynamics of plant viruses (Fig. 4). Such studies will be that reveal the spatial and temporal dynamics of virus
particularly powerful and quantitative (relative frequen- infections and the effects on plant communities. These
cies of genomic sequence reads correspond to the relative hypothesis-​driven next-​generation metagenomics studies
frequencies of sequences within the virome), especially will definitively determine if, and how, plant viruses either
when virus-​sequence reads are linked to specific geo- foster or harm the stability and productivity of ecosystems.
graphical locations and the host or vector. In addition to
enabling the large-​scale quantitative determination of the Published online xx xx xxxx

1. Edwards, R. A. & Rohwer, F. Viral metagenomics. 5. Gregory, A. C. et al. Marine DNA viral macro- and 8. Bernardo, P. et al. Geometagenomics illuminates
Nat. Rev. Microbiol. 3, 504–510 (2005). microdiversity from pole to pole. Cell 177, the impact of agriculture on the distribution and
2. Greninger, A. L. A decade of RNA virus metagenomics 1109–1123 (2019). prevalence of plant viruses at the ecosystem scale.
is (not) enough. Virus Res. 244, 218–229 (2018). This study provides the most comprehensive ISME J. 12, 173–184 (2018).
3. Koonin, E. V. & Dolja, V. V. Metaviromics: a tectonic inventory of marine viral diversity to date. This article is a viromics study of agro-​ecological
shift in understanding virus evolution. Virus Res. 246, 6. Morris, J. L. et al. The timescale of early land plant interfaces that demonstrates the impacts of
A1–A3 (2018). evolution. Proc. Natl Acad. Sci. USA 115, agriculture on the diversity and prevalence of
This review highlights how our understanding of E2274–E2283 (2018). plant-​associated viruses.
virology is changing as a consequence of advances 7. Mushegian, A., Shipunov, A. & Elena, S. F. 9. Muthukumar, V. et al. Non-​cultivated plants of the
in metagenomics. Changes in the composition of the RNA virome mark Tallgrass Prairie Preserve of northeastern Oklahoma
4. Suttle, C. A. Viruses: unlocking the greatest evolutionary transitions in green plants. BMC Biol. frequently contain virus-​like sequences in particulate
biodiversity on Earth. Genome 56, 542–544 (2013). 14, 68 (2016). fractions. Virus Res. 141, 169–173 (2009).

www.nature.com/nrmicro
 Reviews

10. Koonin, E. V., Dolja, V. V. & Krupovic, M. Origins symbioses in vascular plants (a review). Proc. Natl 58. Hamelin, F. M., Allen, L. J., Prendeville, H. R.,
and evolution of viruses of eukaryotes: the ultimate Acad. Sci. USA 77, 2113–2118 (1980). Hajimorad, M. R. & Jeger, M. J. The evolution of plant
modularity. Virology 479–480, 2–25 (2015). 34. Bonfante, P. & Genre, A. Mechanisms underlying virus transmission pathways. J. Theor. Biol. 396,
11. Nasir, A. & Caetano-​Anolles, G. A phylogenomic beneficial plant–fungus interactions in mycorrhizal 75–89 (2016).
data-​driven exploration of viral origins and evolution. symbiosis. Nat. Commun. 1, 48 (2010). 59. Hamelin, F. M. et al. The evolution of parasitic
Sci. Adv. 1, e1500527 (2015). 35. Redman, R. S., Sheehan, K. B., Stout, R. G., and mutualistic plant–virus symbioses through
12. Dolja, V. V. & Koonin, E. V. Common origins and host-​ Rodriguez, R. J. & Henson, J. M. Thermotolerance transmission–virulence trade-​offs. Virus Res. 241,
dependent diversity of plant and animal viromes. generated by plant/fungal symbiosis. Science 298, 77–87 (2017).
Curr. Opin. Virol. 1, 322–331 (2011). 1581 (2002). 60. Nault, L. R. Arthropod transmission of plant viruses:
13. Wolf, Y. I. et al. Origins and evolution of the global 36. Rodriguez, R. & Redman, R. More than 400 million a new synthesis. Ann. Entomol. Soc. Am. 90, 521–541
RNA virome. mBio 9, e02329–18 (2018). years of evolution and some plants still can’t make it (1997).
This study yields new insights relating to the on their own: plant stress tolerance via fungal 61. Tamada, T. & Kondo, H. Biological and genetic diversity
evolution of RNA viruses in light of novel RNA symbiosis. J. Exp. Bot. 59, 1109–1114 (2008). of plasmodiophorid-​transmitted viruses and their
viruses that have recently been discovered using 37. Li, C. X. et al. Unprecedented genomic diversity vectors. J. Gen. Plant Pathol. 79, 307–320 (2013).
metagenomics techniques. of RNA viruses in arthropods reveals the ancestry of 62. Hogenhout, S. A., Ammar el, D., Whitfield, A. E. &
14. Volk, M., Gibbs, A. J. & Suttle, C. A. Metagenomes of negative-​sense RNA viruses. eLife 4, e05378 (2015). Redinbaugh, M. G. Insect vector interactions with
a freshwater charavirus from British Columbia provide 38. Shi, M. et al. Redefining the invertebrate RNA persistently transmitted viruses. Annu. Rev.
a window into ancient lineages of viruses. Viruses 11, virosphere. Nature 540, 539–543 (2016). Phytopathol. 46, 327–359 (2008).
299 (2019). This study identified ~1,400 RNA viruses using 63. Dader, B. et al. Insect transmission of plant viruses:
15. Krupovic, M. & Koonin, E. V. Multiple origins of viral metatranscriptomics, which tremendously expands multilayered interactions optimize viral propagation.
capsid proteins from cellular ancestors. Proc. Natl our current knowledge of RNA virus diversity. Insect Sci. 24, 929–946 (2017).
Acad. Sci. USA 114, E2401–E2410 (2017). 39. Dasgupta, R., Garcia, B. H. 2nd & Goodman, R. M. 64. Uzest, M. et al. A protein key to plant virus transmission
16. Dolja, V. V. & Koonin, E. V. Metagenomics reshapes Systemic spread of an RNA insect virus in plants at the tip of the insect vector stylet. Proc. Natl Acad.
the concepts of RNA virus evolution by revealing expressing plant viral movement protein genes. Sci. USA 104, 17959–17964 (2007).
extensive horizontal virus transfer. Virus Res. 244, Proc. Natl Acad. Sci. USA 98, 4910–4915 (2001). 65. Ammar, E.-D., Tsai, C. W., Whitfield, A. E.,
36–52 (2018). 40. Gibbs, A. J., Wood, J., Garcia-​Arenal, F., Ohshima, K. Redinbaugh, M. G. & Hogenhout, S. A. Cellular and
17. Roossinck, M. J. Evolutionary and ecological links & Armstrong, J. S. Tobamoviruses have probably co-​ molecular aspects of rhabdovirus interactions with
between plant and fungal viruses. New Phytol. 221, diverged with their eudicotyledonous hosts for at least insect and plant hosts. Annu. Rev. Entomol. 54,
86–92 (2019). 110 million years. Virus Evol. 1, vev019 (2015). 447–468 (2009).
This review describes our current knowledge of 41. Stobbe, A. H., Melcher, U., Palmer, M. W., 66. Brault, V., Herrbach, E. & Reinbold, C. Electron
mycoviruses and their evolutionary relationships Roossinck, M. J. & Shen, G. Co-​divergence and microscopy studies on luteovirid transmission by
with plant viruses. host-​switching in the evolution of tobamoviruses. aphids. Micron 38, 302–312 (2007).
18. Vieira, P. & Nemchinov, L. G. A novel species of RNA J. Gen. Virol. 93, 408–418 (2012). 67. Blanc, S. & Michalakis, Y. Manipulation of hosts and
virus associated with root lesion nematode 42. Gibbs, A. How ancient are the tobamoviruses? vectors by plant viruses and impact of the environment.
Pratylenchus penetrans. J. Gen. Virol. 100, 704–708 Intervirology 14, 101–108 (1980). Curr. Opin. Insect Sci. 16, 36–43 (2016).
(2019). 43. Varsani, A., Lefeuvre, P., Roumagnac, P. & Martin, D. 68. Safari, M., Ferrari, M. J. & Roossinck, M. J.
19. Hollings, M. Viruses associated with a die-​back Notes on recombination and reassortment in Manipulation of aphid behavior by a persistent plant
disease of cultivated mushroom. Nature 196, multipartite/segmented viruses. Curr. Opin. Virol. 33, virus. J. Virol. 93, e01781–18 (2019).
962–965 (1962). 156–166 (2018). 69. Mauck, K., Bosque-​Perez, N. A., Eigenbrode, S. D.,
20. Pearson, M. N., Beever, R. E., Boine, B. & Arthur, K. 44. Briddon, R. W. et al. Alphasatellitidae: a new family De Moraes, C. M. & Mescher, M. C. Transmission
Mycoviruses of filamentous fungi and their relevance with two subfamilies for the classification of mechanisms shape pathogen effects on host–vector
to plant pathology. Mol. Plant Pathol. 10, 115–128 geminivirus- and nanovirus-​associated alphasatellites. interactions: evidence from plant viruses. Funct. Ecol.
(2009). Arch. Virol. 163, 2587–2600 (2018). 26, 1162–1175 (2012).
21. Mu, F. et al. Virome characterization of a collection of 45. Gnanasekaran, P. & Chakraborty, S. Biology of viral 70. Gallitelli, D. The ecology of Cucumber mosaic virus and
S. sclerotiorum from Australia. Front. Microbiol. 8, satellites and their role in pathogenesis. Curr. Opin. sustainable agriculture. Virus Res. 71, 9–21 (2000).
2540 (2017). Virol. 33, 96–105 (2018). 71. Power, A. G. & Flecker, A. S. in Infectious Disease
22. Marzano, S. L. & Domier, L. L. Novel mycoviruses 46. Lucia-​Sanz, A. & Manrubia, S. Multipartite viruses: Ecology: The Effects of Ecosystems on Disease and
discovered from metatranscriptomics survey of adaptive trick or evolutionary treat? NPJ Syst. Biol. of Disease on Ecosystems Ch. 2 (eds Ostfeld, R. S.,
soybean phyllosphere phytobiomes. Virus Res. 213, Appl. 3, 34 (2017). Keesing, F. & Eviner, V. T.) (Princeton Univ. Press,
332–342 (2016). 47. Escriu, F., Fraile, A. & Garcia-​Arenal, F. Constraints 2010).
23. Gilbert, K., Holcomb, E. E., Allscheid, R. L. & to genetic exchange support gene coadaptation in 72. Anderson, P. K. et al. Emerging infectious diseases
Carrington, J. Discovery of new mycoviral genomes a tripartite RNA virus. PLOS Pathog. 3, e8 (2007). of plants: pathogen pollution, climate change and
within publicly available fungal transcriptomic 48. Sicard, A. et al. Gene copy number is differentially agrotechnology drivers. Trends Ecol. Evol. 19,
datasets. Preprint at bioRxiv https://www.biorxiv.org/ regulated in a multipartite virus. Nat. Commun. 4, 535–544 (2004).
content/10.1101/510404v1 (2019). 2248 (2013). 73. Gilbertson, R. L., Batuman, O., Webster, C. G. &
24. Marzano, S. L. et al. Identification of diverse mycoviruses 49. Wu, B., Zwart, M. P., Sanchez-​Navarro, J. A. Adkins, S. Role of the insect supervectors Bemisia
through metatranscriptomics characterization of the & Elena, S. F. Within-​host evolution of segments tabaci and Frankliniella occidentalis in the emergence
viromes of five major fungal plant pathogens. J. Virol. ratio for the tripartite genome of alfalfa mosaic virus. and global spread of plant viruses. Annu. Rev. Virol. 2,
90, 6846–6863 (2016). Sci. Rep. 7, 5004 (2017). 67–93 (2015).
25. Frank, A. C. & Wolfe, K. H. Evolutionary capture of 50. Benitez-​Alfonso, Y., Faulkner, C., Ritzenthaler, C. 74. Fereres, A. Insect vectors as drivers of plant virus
viral and plasmid DNA by yeast nuclear chromosomes. & Maule, A. J. Plasmodesmata: gateways to local and emergence. Curr. Opin. Virol. 10, 42–46 (2015).
Eukaryot. Cell 8, 1521–1531 (2009). systemic virus infection. Mol. Plant Microbe Interact. 75. Simmonds, P., Aiewsakun, P. & Katzourakis, A.
26. Taylor, D. J. & Bruenn, J. The evolution of novel fungal 23, 1403–1412 (2010). Prisoners of war — host adaptation and its constraints
genes from non-​retroviral RNA viruses. BMC Biol. 7, 51. Lucas, W. J. Plant viral movement proteins: agents for on virus evolution. Nat. Rev. Microbiol. 17, 321–328
88 (2009). cell-​to-cell trafficking of viral genomes. Virology 344, (2019).
27. Hillman, B. I. & Cai, G. The family Narnaviridae: 169–184 (2006). This study provides both a summary of viral
simplest of RNA viruses. Adv. Virus Res. 86, 149–176 52. Sicard, A., Michalakis, Y., Gutierrez, S. & Blanc, S. evolutionary rates and a tentative framework to
(2013). The strange lifestyle of multipartite viruses. accommodate potentially conflicting long-​term and
28. Nerva, L. et al. Biological and molecular PLOS Pathog. 12, e1005819 (2016). short-​term evolutionary rate inferences.
characterization of Chenopodium quinoa mitovirus 1 This article reviews the ‘lifestyles’ of multipartite 76. Stroud, J. T. & Losos, J. B. Ecological opportunity
reveals a distinct small RNA response compared to viruses, their peculiarities and the gaps in our and adaptive radiation. Annu. Rev. Ecol. Evol. Syst.
those of cytoplasmic RNA viruses. J. Virol. 93, understanding of their biology. 47, 507–532 (2016).
e01998–18 (2019). 53. Gilmer, D., Ratti, C. & Michel, F. Long-​distance 77. Scholthof, K. B. et al. Top 10 plant viruses in molecular
29. Rastgou, M. et al. Molecular characterization of the movement of helical multipartite phytoviruses: keep plant pathology. Mol. Plant Pathol. 12, 938–954
plant virus genus Ourmiavirus and evidence of inter-​ connected or die? Curr. Opin. Virol. 33, 120–128 (2011).
kingdom reassortment of viral genome segments as its (2018). 78. Jacquemond, M. Cucumber mosaic virus. Adv. Virus
possible route of origin. J. Gen. Virol. 90, 2525–2535 54. Fauquet, C. M., Mayo, M. A., Maniloff, J., Res. 84, 439–504 (2012).
(2009). Desselberger, U. & Ball, L. A. Virus Taxonomy: VIIIth 79. Dietzgen, R. G., Mann, K. S. & Johnson, K. N.
30. Nerva, L., Varese, G. C., Falk, B. W. & Turina, M. Report of the International Committee on Taxonomy Plant virus–insect vector interactions: current and
Mycoviruses of an endophytic fungus can replicate in of Viruses (Academic Press, 2005). potential future research directions. Viruses 8, E303
plant cells: evolutionary implications. Sci. Rep. 7, 55. Liu, S. et al. Fungal DNA virus infects a mycophagous (2016).
1908 (2017). insect and utilizes it as a transmission vector. 80. Bragard, C. et al. Status and prospects of plant virus
31. Andika, I. B. et al. Phytopathogenic fungus hosts a Proc. Natl Acad. Sci. USA 113, 12803–12808 control through interference with vector transmission.
plant virus: a naturally occurring cross-​kingdom viral (2016). Annu. Rev. Phytopathol. 51, 177–201 (2013).
infection. Proc. Natl Acad. Sci. USA 114, 56. Sacristan, S., Diaz, M., Fraile, A. & Garcia-​Arenal, F. 81. Eastop, V. F. in Aphids As Virus Vectors (eds Harris, K. F.
12267–12272 (2017). Contact transmission of tobacco mosaic virus: a & Maramorosch, K.) 3–62 (Academic Press, 1977).
32. Mascia, T. et al. Infection of Colletotrichum acutatum quantitative analysis of parameters relevant for virus 82. Li, C., Cox-​Foster, D., Gray, S. M. & Gildow, F.
and Phytophthora infestans by taxonomically different evolution. J. Virol. 85, 4974–4981 (2011). Vector specificity of barley yellow dwarf virus (BYDV)
plant viruses. Eur. J. Plant Pathol. 153, 1001–1017 57. Jones, R. A. C. Plant and insect viruses in managed transmission: identification of potential cellular
(2018). and natural environments: novel and neglected receptors binding BYDV-​MAV in the aphid,
33. Malloch, D. W., Pirozynski, K. A. & Raven, P. H. transmission pathways. Adv. Virus Res. 101, 149–187 Sitobion avenae. Virology 286, 125–133
Ecological and evolutionary significance of mycorrhizal (2018). (2001).

Nature Reviews | Microbiology

Reviews

83. Bedhomme, S., Hillung, J. & Elena, S. F. 107. Fargette, D. et al. Molecular ecology and emergence 128. Roossinck, M. J. A new look at plant viruses and their
Emerging viruses: why they are not jacks of all trades? of tropical plant viruses. Annu. Rev. Phytopathol. 44, potential beneficial roles in crops. Mol. Plant Pathol.
Curr. Opin. Virol. 10, 1–6 (2015). 235–260 (2006). 16, 331–333 (2015).
84. Elena, S. F. Local adaptation of plant viruses: 108. Jones, R. A. Plant virus emergence and evolution: 129. Roossinck, M. J. Plants, viruses and the environment:
lessons from experimental evolution. Mol. Ecol. 26, origins, new encounter scenarios, factors driving ecology and mutualism. Virology 479–480, 271–277
1711–1719 (2017). emergence, effects of changing world conditions, and (2015).
85. Remold, S. Understanding specialism when the Jack of prospects for control. Virus Res. 141, 113–130 (2009). This review addresses the paradigm shift away
all trades can be the master of all. Proc. Biol. Sci. 279, This article reviews various anthropogenic factors from viewing viruses as antagonistic pathogens
4861–4869 (2012). associated with the emergence of plant viruses and towards viewing them as possible mutualists.
86. Kawecki, T. J. Accumulation of deleterious mutations the ongoing challenges in mitigating emerging 130. Shates, T. M., Sun, P., Malmstrom, C. M., Dominguez, C.
and the evolutionary cost of being a generalist. disease threats. & Mauck, K. E. Addressing research needs in the field
Am. Nat. 144, 833–838 (1994). 109. Pagan, I. et al. Effect of biodiversity changes in disease of plant virus ecology by defining knowledge gaps and
87. Cooper, I. & Jones, R. A. Wild plants and viruses: risk: exploring disease emergence in a plant–virus developing wild dicot study systems. Front. Microbiol. 9,
under-​investigated ecosystems. Adv. Virus Res. 67, system. PLOS Pathog. 8, e1002796 (2012). 3305 (2018).
1–47 (2006). 110. Roossinck, M. J. & Garcia-​Arenal, F. Ecosystem 131. Fraile, A. & Garcia-​Arenal, F. The coevolution of plants
88. Rodriguez-​Nevado, C., Montes, N. & Pagan, I. simplification, biodiversity loss and plant virus and viruses: resistance and pathogenicity. Adv. Virus
Ecological factors affecting infection risk and population emergence. Curr. Opin. Virol. 10, 56–62 (2015). Res. 76, 1–32 (2010).
genetic diversity of a novel potyvirus in its native wild 111. Rodelo-​Urrego, M. et al. Landscape heterogeneity 132. Calil, I. P. & Fontes, E. P. B. Plant immunity against
ecosystem. Front. Plant Sci. 8, 1958 (2017). shapes host–parasite interactions and results in viruses: antiviral immune receptors in focus. Ann. Bot.
89. Susi, H., Filloux, D., Frilander, M. J., Roumagnac, P. apparent plant–virus codivergence. Mol. Ecol. 22, 119, 711–723 (2017).
& Laine, A. L. Diverse and variable virus communities 2325–2340 (2013). 133. Malmstrom, C. M. & Alexander, H. M. Effects of crop
in wild plant populations revealed by metagenomic 112. Rocha, C. S. et al. Brazilian begomovirus populations viruses on wild plants. Curr. Opin. Virol. 19, 30–36
tools. PeerJ 7, e6140 (2019). are highly recombinant, rapidly evolving, and (2016).
90. Elena, S. F., Fraile, A. & García-​Arenal, F. Evolution segregated based on geographical location. 134. Prendeville, H. R., Ye, X., Morris, T. J. & Pilson, D.
and emergence of plant viruses. Adv. Virus Res. 88, J. Virol. 87, 5784–5799 (2013). Virus infections in wild plant populations are both
161–191 (2014). 113. Keesing, F. et al. Impacts of biodiversity on the frequent and often unapparent. Am. J. Bot. 99,
91. McLeish, M. J., Fraile, A. & Garcia-​Arenal, F. emergence and transmission of infectious diseases. 1033–1042 (2012).
Ecological complexity in plant virus host range Nature 468, 647–652 (2010). 135. Remold, S. K. Unapparent virus infection and host
evolution. Adv. Virus Res. 101, 293–339 (2018). This article addresses the impacts of reduced fitness in three weedy grass species. J. Ecol. 90,
This article presents a comprehensive review of biodiversity on disease transmission. 967–977 (2002).
plant virus ecology. 114. Lima, A. T. et al. Synonymous site variation due to 136. Fraile, A. et al. Environmental heterogeneity and the
92. Cuevas, J. M., Willemsen, A., Hillung, J., Zwart, M. P. recombination explains higher genetic variability in evolution of plant-​virus interactions: viruses in wild
& Elena, S. F. Temporal dynamics of intrahost begomovirus populations infecting non-​cultivated pepper populations. Virus Res. 241, 68–76 (2017).
molecular evolution for a plant RNA virus. Mol. Biol. hosts. J. Gen. Virol. 94, 418–431 (2013). 137. Faure, D., Simon, J. C. & Heulin, T. Holobiont: a
Evol. 32, 1132–1147 (2015). 115. Thingstad, T. F. & Lignell, R. Theoretical models for the conceptual framework to explore the eco-​evolutionary
93. Minicka, J., Rymelska, N., Elena, S. F., Czerwoniec, A. control of bacterial growth rate, abundance, diversity and functional implications of host–microbiota
& Hasiów-​Jaroszewska, B. Molecular evolution of and carbon demand. Aquat. Microb. Ecol. 13, 19–27 interactions in all ecosystems. New Phytol. 218,
Pepino mosaic virus during long-​term passaging in (1997). 1321–1324 (2018).
different hosts and its impact on virus virulence. 116. Coutinho, F. H. et al. Marine viruses discovered 138. Grasis, J. A. The intra-​dependence of viruses and the
Ann. Appl. Biol. 166, 389–401 (2015). via metagenomics shed light on viral strategies holobiont. Front. Immunol. 8, 1501 (2017).
94. Ciota, A. T. et al. Experimental passage of St. Louis throughout the oceans. Nat. Commun. 8, 139. Harth, J. E., Ferrari, M. J., Tooker, J. F.
encephalitis virus in vivo in mosquitoes and chickens 15955 (2017). & Stephenson, A. G. Zucchini yellow mosaic
reveals evolutionarily significant virus characteristics. 117. Coutinho, F. H., Gregoracci, G. B., Walter, J. M., virus infection limits establishment and severity
PLOS ONE 4, e7876 (2009). Thompson, C. C. & Thompson, F. L. Metagenomics of powdery mildew in wild populations of Cucurbita
95. Greene, I. P. et al. Effect of alternating passage sheds light on the ecology of marine microbes and pepo. Front. Plant Sci. 9, 792 (2018).
on adaptation of sindbis virus to vertebrate and their viruses. Trends Microbiol. 26, 955–965 (2018). 140. Gibbs, A. A plant virus that partially protects its wild
invertebrate cells. J. Virol. 79, 14253–14260 118. Mitchell, C. E. & Power, A. G. Release of invasive legume host against herbivores. Intervirology 13,
(2005). plants from fungal and viral pathogens. Nature 421, 42–47 (1980).
96. Turner, P. E. & Elena, S. F. Cost of host radiation in an 625–627 (2003). 141. Davis, T. S., Bosque-​Perez, N. A., Foote, N. E.,
RNA virus. Genetics 156, 1465–1470 (2000). 119. Borer, E. T., Hosseini, P. R., Seabloom, E. W. & Magney, T. & Eigenbrode, S. D. Environmentally
97. Bedhomme, S., Lafforgue, G. & Elena, S. F. Dobson, A. P. Pathogen-​induced reversal of native dependent host–pathogen and vector–pathogen
Multihost experimental evolution of a plant RNA virus dominance in a grassland community. Proc. Natl Acad. interactions in the Barley yellow dwarf virus
reveals local adaptation and host-​specific mutations. Sci. USA 104, 5473–5478 (2007). pathosystem. J. Appl. Ecol. 52, 1392–1401 (2015).
Mol. Biol. Evol. 29, 1481–1492 (2012). 120. Malmstrom, C. M., McCullough, A. J., Johnson, H. A., 142. Hily, J. M., Poulicard, N., Mora, M. A., Pagan, I.
This study provides experimental evidence for the Newton, L. A. & Borer, E. T. Invasive annual grasses & Garcia-​Arenal, F. Environment and host genotype
possible existence of no-​cost generalists in plant indirectly increase virus incidence in California native determine the outcome of a plant–virus interaction:
viruses. perennial bunchgrasses. Oecologia 145, 153–164 from antagonism to mutualism. New Phytol. 209,
98. Hillung, J., Cuevas, J. M., Valverde, S. & Elena, S. F. (2005). 812–822 (2016).
Experimental evolution of an emerging plant virus in This article addresses the impacts of an invasive This study provides compelling evidence that
host genotypes that differ in their susceptibility to plant species on viral dynamics in native plants. conditional interactions of plant viruses, plant
infection. Evolution 68, 2467–2480 (2014). 121. Faillace, C. A., Lorusso, N. S. & Duffy, S. Overlooking genotypes and the environment modulate the
99. Lalic, J., Agudelo-​Romero, P., Carrasco, P. the smallest matter: viruses impact biological outcome of symbiosis.
& Elena, S. F. Adaptation of tobacco etch potyvirus invasions. Ecol. Lett. 20, 524–538 (2017). 143. Westwood, J. H. et al. A viral RNA silencing suppressor
to a susceptible ecotype of Arabidopsis thaliana This article reviews the impacts of rapidly evolving interferes with abscisic acid-​mediated signalling and
capacitates it for systemic infection of resistant plant viruses on plant community structure within a induces drought tolerance in Arabidopsis thaliana.
ecotypes. Phil. Trans. R. Soc. B 365, 1997–2007 biological invasion framework. Mol. Plant Pathol. 14, 158–170 (2013).
(2010). 122. Cervera, H., Ambros, S., Bernet, G. P., Rodrigo, G. & 144. Xu, P. et al. Virus infection improves drought tolerance.
100. Charron, C. et al. Natural variation and functional Elena, S. F. Viral fitness correlates with the magnitude New Phytol. 180, 911–921 (2008).
analyses provide evidence for co-​evolution between and direction of the perturbation induced in the host’s 145. Prasch, C. M. & Sonnewald, U. Simultaneous
plant eIF4E and potyviral VPg. Plant J. 54, 56–68 transcriptome: the Tobacco Etch Potyvirus-​Tobacco application of heat, drought, and virus to Arabidopsis
(2008). Case Study. Mol. Biol. Evol. 35, 1599–1615 (2018). plants reveals significant shifts in signaling networks.
101. Jenner, C. E., Wang, X., Ponz, F. & Walsh, J. A. 123. Alizon, S., Hurford, A., Mideo, N. & Van Baalen, M. Plant Physiol. 162, 1849–1866 (2013).
A fitness cost for Turnip mosaic virus to overcome Virulence evolution and the trade-​off hypothesis: 146. Berges, S. E. et al. Interactions between drought
host resistance. Virus Res. 86, 1–6 (2002). history, current state of affairs and the future. and plant genotype change epidemiological traits
102. Hillung, J., Garcia-​Garcia, F., Dopazo, J., Cuevas, J. M. J. Evol. Biol. 22, 245–259 (2009). of cauliflower mosaic virus. Front. Plant Sci. 9, 703
& Elena, S. F. The transcriptomics of an experimentally 124. Anderson, R. M. & May, R. M. Coevolution of hosts (2018).
evolved plant–virus interaction. Sci. Rep. 6, 24901 and parasites. Parasitology 85, 411–426 (1982). 147. Bera, S., Fraile, A. & Garcia-​Arenal, F. Analysis of
(2016). 125. Doumayrou, J., Avellan, A., Froissart, R. & Michalakis, Y. fitness trade-​offs in the host range expansion of an
103. McCallum, E. J., Anjanappa, R. B. & Gruissem, W. An experimental test of the transmission–virulence RNA virus, tobacco mild green mosaic virus. J. Virol.
Tackling agriculturally relevant diseases in the staple trade-​off hypothesis in a plant virus. Evolution 67, 92, e01268–18 (2018).
crop cassava (Manihot esculenta). Curr. Opin. Plant 477–486 (2013). 148. Zhang, Y. Z., Shi, M. & Holmes, E. C. Using
Biol. 38, 50–58 (2017). 126. Froissart, R., Doumayrou, J., Vuillaume, F., Alizon, S. metagenomics to characterize an expanding
104. Almeida, R. P. et al. Ecology and management of & Michalakis, Y. The virulence–transmission trade-​off virosphere. Cell 172, 1168–1172 (2018).
grapevine leafroll disease. Front. Microbiol. 4, 94 in vector-​borne plant viruses: a review of (non-)existing 149. Saunders, K., Bedford, I. D., Yahara, T. & Stanley, J.
(2013). studies. Phil. Trans. R. Soc. B 365, 1907–1918 Aetiology: the earliest recorded plant virus disease.
105. Walls, J., Rajotte, E. & Rosa, C. The past, present, and (2010). Nature 422, 831 (2003).
future of barley yellow dwarf management. Agriculture This article reviews our current understanding of 150. Lesnaw, J. A. & Ghabrial, S. A. Tulip breaking:
9, 23 (2019). the transmission–virulence trade-​off hypothesis in past, present, and future. Plant Dis. 84, 1052–1060
106. Pinel-​Galzi, A., Traore, O., Sere, Y., Hebrard, E. & the context of plant viruses. (2000).
Fargette, D. The biogeography of viral emergence: 127. Leggett, H. C., Buckling, A., Long, G. H. & Boots, M. 151. Ivanowski, D. Ueber die Mosaikkrankheit der
rice yellow mottle virus as a case study. Curr. Opin. Virol. Generalism and the evolution of parasite virulence. Tabakspflanze. Bull. Acad. Imp. Sci. 35, 67–70
10, 7–13 (2015). Trends Ecol. Evol. 28, 592–596 (2013). (1892).

www.nature.com/nrmicro
 Reviews

152. Beijerinck, W. M. Ueber ein contagium vivum fluidum 163. Evans, G. A. Host Plant List of The Whiteflies Nation University, Australia) for helpful comments and sugges-
als Ursache der Fleckenkrankheit der Tabaksblatter. (Aleyrodidae) of The World (USDA Animal Plant tions. P.L. was supported by the European Union: European
Verh. Kon. Akad. Wetensch. 5, 3–21 (1898). Health Inspection Service, 2007). Regional Development Fund (ERDF), by the Conseil Régional
153. Stanley, W. M. Isolation of a crystalline protein 164. De Barro, P. J., Liu, S. S., Boykin, L. M. & Dinsdale, A. B. de La Réunion and by the Centre de Coopération internation-
possessing the properties of tobacco-​mosaic virus. Bemisia tabaci: a statement of species status. ale en Recherche agronomique pour le Développement
Science 81, 644–645 (1935). Annu. Rev. Entomol. 56, 1–19 (2011). (CIRAD). S.F.E. was supported by a grant (BFU2015-65037-P)
154. Kausche, G. A., Pfankuch, E. & Ruska, H. 165. Huang, Y., Niu, B., Gao, Y., Fu, L. & Li, W. CD-​HIT Suite: f ro m S p a i n M i n i s t r y o f S c i e n c e, I n n ova t i o n a n d
Die Sichtbarmachung von pflanzlichem Virus a web server for clustering and comparing biological Universities–ERDF.
im Übermikroskop. Naturwissenschaften 27, sequences. Bioinformatics 26, 680–682 (2010).
292–299 (1939). 166. Katoh, K. & Standley, D. M. MAFFT multiple sequence
Author contributions
155. Bernal, J. D. & Fankuchen, I. Structure types of alignment software version 7: improvements in
P.L., D.P.M., S.F.E., D.N.S., P.R. and A.V. wrote and edited the
protein crystals from virus-​infected plants. Nature performance and usability. Mol. Biol. Evol. 30,
manuscript. P.L. and A.V. undertook the analyses for the data
139, 923–924 (1937). 772–780 (2013).
presented in figures 1–3.
156. Fraenkel-​Conrat, H. The genetic code of a virus. 167. Guindon, S. et al. New algorithms and methods to
Sci. Am. 211, 47–54 (1964). estimate maximum-​likelihood phylogenies: assessing
157. Anandalakshmi, R. et al. A viral suppressor of gene the performance of PhyML 3.0. Syst. Biol. 59, Competing interests
silencing in plants. Proc. Natl Acad. Sci. USA 95, 307–321 (2010). The authors declare no competing interests.
13079–13084 (1998). 168. Clark, K., Karsch-​Mizrachi, I., Lipman, D. J., Ostell, J.
158. Hamilton, A. J. & Baulcombe, D. C. A species of small & Sayers, E. W. GenBank. Nucleic Acids Res. 44,
Peer review information
antisense RNA in posttranscriptional gene silencing in D67–D72 (2016).
Nature Reviews Microbiology thanks M. Roossinck,
plants. Science 286, 950–952 (1999). 169. Ratnasingham, S. & Hebert, P. D. bold: the Barcode
A. Whitfield and the other, anonymous, reviewer(s) for their
159. Guo, Z. et al. Identification of a new host factor of Life Data System (http://www.barcodinglife.org).
contribution to the peer review of this work.
required for antiviral RNAi and amplification of viral Mol. Ecol. Notes 7, 355–364 (2007).
siRNAs. Plant Physiol. 176, 1587–1597 (2018). 170. Gerlt, J. A. et al. Enzyme Function Initiative-​Enzyme
160. Baltimore, D. Expression of animal virus genomes. Similarity Tool (EFI-​EST): a web tool for generating Publisher’s note
Bacteriol. Rev. 35, 235–241 (1971). protein sequence similarity networks. Biochim. Springer Nature remains neutral with regard to jurisdictional
161. Bolduc, B. et al. vConTACT: an iVirus tool to classify Biophys. Acta 1854, 1019–1037 (2015). claims in published maps and institutional affiliations.
double-​stranded DNA viruses that infect Archaea
and Bacteria. PeerJ 5, e3243 (2017). Acknowledgements Related links
162. Mihara, T. et al. Linking virus genomes with host The authors are grateful to Y. Michalakis (Centre national de la Plant viruses dataset: https://lefeup.github.io/plantviruses/
taxonomy. Viruses 8, 66 (2016). recherche scientifique, France) and A. Gibbs (Australian

Nature Reviews | Microbiology