Biogeographic distribution of infected plants and the continental drift theory allow a tentative time calibration of the phylogenetic tree of Pospiviroidae. Hypothetically, viroids evolved in the late Early Cretaceous shortly after the appearance of angiosperms, which constitute their only known hosts. No decline in the estimated divergence rates of Pospiviroidae is observed during the Late Cretaceous but it appears that they abruptly decreased at the Cretaceous-Paleogene boundary. However, an adaptive radiation of Pospiviroidae which occurred in the late Paleocene may reflect a recovery from the Cretaceous-Paleogene (K–Pg) mass extinction. It seems that the evolutionary history of viroids has been in part shaped by radiation and extinction events of angiosperms. Herein, for the first time I show the probable impact of a mass extinction event on the divergence rates of subviral pathogens, which are the simplest known “lifeforms”.

mass extinctions, viroids, paleovirology

Mass extinction events play a fundamental role in shaping the biosphere (Raup and Sepkoski 1982; Novacek 1999; McElwain and Punyasena 2007; Chin et al. 2013; Niedźwiedzki et al. 2016). Wang et al. (2011) suggested that mass extinctions might have also impacted on the evolutionary history of viruses. This is a reasonable concept since the evolution of viruses should have been in part shaped by the history of their hosts (see Thézé et al. 2011). However, data demonstrating the impact of global mass extinctions on the diversity or divergence rates of viral or subviral pathogens have never been discussed.

Viroids are small (246–399 nucleotides), unencapsidated, single-stranded, circular RNAs, which are known to infect solely angiosperm plants (Diener 1971, 2003; Góra-Sochacka 2004; Flores et al. 2005). Viroids differ from viruses by a much smaller genome size and, in contrast to viruses, do not code for any protein. Viroids are classified as subviral particles and, alongside small linear and circular satellite RNAs, constitute the smallest pathogens and “lifeforms” known (Elena et al. 1991; Flores et al. 2014; AbouHaidar et al. 2014). Viroids appear to be evolutionarily related to satellite RNAs, including the human Hepatitis delta virus (Elena et al. 1991, 2001). However, in contrast to satellite RNAs, viroids complete their infectious cycle without resorting to a helper virus (Daròs et al. 2006). Viroids are divided into two families, Pospiviroidae and Avsunviroidae, which replicate either in the nuclei or in the chloroplasts, respectively (Tsagris et al. 2008; Flores et al. 2012; Rao and Kalantidis 2015).

In this paper, I present a time-calibrated phylogenetic tree of members of the family Pospiviroidae. I also estimate the divergence rates of Pospiviroidae throughout the Cretaceous and the Paleogene. The hypothetical impact of the Cretaceous–Paleogene (K–Pg) extinction event on the divergence rates of this viroid family is discussed.

Evolutionary analyses involved 34 nucleotide sequences of viroids belonging to the families Pospiviroidae (30 sequences) and Avsunviroidae (4 sequences) (Table 1), which were downloaded from the National Center for Biotechnology Information, U.S. National Library of Medicine. Sequence alignment was performed by the MUSCLE algorithm. The analyses were conducted in MEGA X (Kumar et al. 2018) under Debian/GNU Linux 9.6. “Stretch”.

Phylogeny

The phylogenetic tree of viroids was constructed by using the Maximum Likelihood method and General Time Reversible model (Nei and Kumar 2000). The tree with the highest log likelihood (-11928.92) is shown in Fig. 1. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree for the heuristic search was obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site (Fig. 1).

Figure 1.

Phylogenetic tree of viroids including Pospiviroidae (in black) and Avsunviroidae (in red). Assignment of the Dahlia latent viroid to the genera Pospiviroid or Hostuviroid is problematic (Verhoeven et al. 2013).

Divergence Rates

The time-calibrated phylogenetic tree (Fig. 2) was used to estimate the divergence rates of Pospiviroidae in different epochs of the Cretaceous and the Paleogene periods. Divergence/speciation rate is understood as the number of divergence events divided by epoch duration. Calculations are provided in the Table 2. Duration of given epochs was estimated based on the International Chronostratigraphic Chart v. 2018/08 (Cohen et al. 2013, updated 2018).

Figure 2.

Time-calibrated phylogenetic tree of Pospiviroidae. A–E Calibration points (see Material and Methods). Viroids typically infecting plants native to South America are marked in red.

Table 2.

Download as

CSV

XLSX

Divergence rates of Pospiviroidae throughout the Cretaceous and the Paleogene.

Period	Cretaceous		Paleogene
Epoch	Early Cretaceous	Late Cretaceous	Paleocene	Eocene	Oligocene
Divergence events	3	12	1	8	5
Epoch duration (million years)	16.1*	34.5	10	22.1	10.87
Divergence rates (divergence events / epoch duration)	0.186	0.347	0.100	0.361	0.459

*since the divergence of viroids 116.6 million years ago

Because there is no fossil record of viruses, paleovirology relies on analyses of modern genetic information (Patel et al. 2011; Aswad and Katzourakis 2012; Feschotte and Gilbert 2012; Taylor et al. 2014). The origin of certain virus groups has been dated back to the Mesozoic (Suh et al. 2013, 2014) and even to the Paleozoic (Thézé et al. 2011). It appears that a possible impact of mass extinctions on the evolution of viruses and subviral pathogens might be relevant at least in case of the end-Permian (251.9 mya), end-Triassic (201.3 mya), and the end-Cretaceous (66 mya) events.

Viroids and viroid-like satellite RNAs have been suggested to possibly represent relics of the primordial RNA world (Diener 1989; Chela-Flores 1994; Flores et al. 2014), which hypothetically existed before the beginning of cellular life about 4 billion years ago (Gilbert 1986; Cech 2012). This concept is based on the small and circular genome and no protein-coding capacity in all viroids, and also on the presence of hammerhead ribozymes in members of the family Avsunviroidae, which act as catalitic RNAs, as expected for primitive self-replicating lifeforms of the primordial RNA world (Diener 1989; Chela-Flores 1994; Flores et al. 2014). Since members of the family Avsunviroidae replicate in chloroplasts, their ancestor could hypothetically infect endosymbiotic cyanobacteria (Chela-Flores 1994).

However, members of the family Avsunviroidae are not capable of replication without resorting to a host despite their ribozymatic activity (Rao and Kalantidis 2015). They are solely known to infect angiosperms (an evolutionarily young group) and unknown to replicate in chloroplasts of other eukaryotes. On the other hand, viroids are usually described from cultivated plants, such as fruit trees, vegetables, and ornamental plants, whereas pathogens of most plant groups seem poorly studied. However, the Early Cretaceous origin of viroids inferred from molecular dating in this study and their close association with angiosperm plants do not support the hypothesis of viroids as relics from the primordial RNA world. Angiosperms, which constitute the only known hosts of viroids, evolved over 124.6 mya (Sun et al. 2002), which was soon followed by the appearance of viroids ≈116.6 mya, as inferred from molecular dating (Fig. 3). Possible pre-Cretaceous origin of angiosperm plants is still a subject of debate (Herendeen et al. 2017; Wang 2017; Fu et al. 2018).

The last common ancestor of the closely related Coconut cadang-cadang and Coconut tinangaja viroids infecting coconut palms (Gitau et al. 2009) is dated to ≈47.27 mya based on molecular data analyzed in this study (Fig. 2). Interestingly, this would imply their split shortly after the appearance of coconut palms, which evolved in the early Eocene (Shukla et al. 2012). However, certain viroids are capable of infecting a variety of host taxa (Verhoeven et al. 2017). Rather than to simply mirror the taxonomy of infected angiosperms the phylogenetic tree of viroids has been also shaped by the paleobiogeographic distribution of hosts and the continental drift (Fig. 2).

The estimated divergence rates of Pospiviroidae were particularly low in the Paleocene Epoch (Table 2; Fig. 3), which may be interpreted as a possible effect of the end-Cretaceous extinction. The K–Pg mass extinction event caused a disappearance of ≈57% of all plant species in North America (Wilf and Johnson 2004). Decrease in host population size and host isolation would have probably led to a mass extinction of viruses (Wang et al. 2011). Viroids can be transmitted by a direct contact between plants, seeds, pollen, and occasionally by insects when encapsidated by a virus (Syller et al. 1997; Card et al. 2007; Van Bogaert et al. 2014). Studies on insect-feeding damages of angiosperm leaves suggest a disappearance of most specialized plant–insect associations at the K–Pg boundary (Labandeira et al. 2002) and severely unbalanced food webs in North America (Wilf et al. 2006). However, such alterations in the plant–insect interactions have not been observed in the Paleocene ecosystems of Europe (Wappler et al. 2009). It should be also noted that insect vectors appear to have a smaller relevance for viroid epidemiology than in the case of viruses (Van Bogaert et al. 2014).

Figure 3.

Relative divergence rates of Pospiviroidae symbolized by the column width; drawn based on the Table 2.

Molecuar dating in this study relies on the geographic provenance of viroids and their hosts. This is particularly important for the branch A that includes viroids infecting predominantly plants native to South America (Fig. 2). For example, the Potato spindle tuber viroid infects mainly Solanum tuberosum (potato), Solanum lycopersicum (tomato), and Capsicum annuum (pepper), which are all native to South America, whereas infections in other hosts are symptomless (Owens and Verhoeven 2009). However, the origin of viroids is often difficult to trace, which is caused by the human activity. For example, the Citrus exocortis viroid is well-known to infect citrus trees in Australia, which are native to Asia. However, it also infects tomato plants in Asia (Mishra et al. 1991), which are native to South America. The Citrus exocortis viroid is in fact closely related to several viroids typically infecting plants native to South America (Fig. 2).

Reliability of the method used in this study is partly dependent on the sample size. Since viroids constitute a small (or a poorly studied) group of pathogenic agents, nucleotide sequences of only 30 members of Pospiviroidae could be used to build the phylogenetic tree, whereas 4 members of Avsunviroidae were used for tree rooting (Table 1). The International Committee on Virus Taxonomy currently recognizes only 32 viroid species (Di Serio et al. 2017), which is even fewer than the 34 sequences studied herein. Notably, there are no divergence events recorded for the Santonian and Bartonian ages of the mid-Late Cretaceous and mid-Eocene epochs, respectively. This is possibly an effect of the poor sampling because the Santonian and the Bartonian are among the shortest ages of the interval studied (see Cohen et al. 2013, updated 2018). For this reason, more reliable results are obtained if the divergence rates are calculated for epochs (longer intervals) rather than for ages (Table 2; Fig. 3). On the other hand, the particularly long-lasting (8-million-years-long) lack of divergence events, which followed the K–Pg mass extinction event, is here tentatively interpreted as a biotic event.

It should be noted that speciation rates inferred from extant species data are underestimated because the phylogenetic trees are obtained by suppressing all extinct lineages (Stadler 2011). For the same reason, the methodology applied herein cannot estimate the extinction rates. Nevertheless, this study shows that relative divergence rates of Pospiviroidae varied between geologic epochs. These differences may reflect paleoecologic changes and particularly events concerning the biodiversity of infected angiosperm plants or possibly interactions of plants with vectors (such as insects) transmitting plant viruses and viroids.

The lack of divergence events during the Neogene and the Quaternary on the constructed phylogenetic tree can be explained as an artifact caused by evolution on a subspecies level. Only nucleotide sequences arbitrarily classified as separate viroid “species” in literature and the NCBI database were included. The analyses have not involved different isolates of the same viroid.

The evolution of Pospiviroidae might have been partly shaped by the evolutionary history of their hosts. It appears that a collapse of food chains (the lack of appropriate vectors transmitting pathogens) and a mass extinction of species (the paucity and isolation of infected hosts) may severely impact viral and subviral pathogens resulting in a decrease of their divergence rates. This hypothetically occurred during the K–Pg mass extinction but results obtained in this study are preliminary and require a thorough testing in future research. As viroids constitute a small group, the study of divergence rates of viruses would provide an important control. Intriguingly, given their abundance in most environments, viruses and subviral pathogens might potentially be useful in the study of the state of the biosphere.

Piotr Bajdek conceived and designed the experiments, contributed analysis tools, performed the experiments, analyzed the data, prepared figures and tables, and wrote the manuscript.

I would like to thank two anonymous journal reviewers and the editor Prof. Frans Jorissen for thoughtful comments, which allowed to improve the quality of the work.