Plants, fungi and oomycetes: a 400‐million year affair that shapes the biosphere

Abstract

In a rare gathering, genomics met palaeontology at the 10th New Phytologist Workshop on the ‘Origin and evolution of plants and their interactions with fungi’. An eclectic group of 17 experts met at The Natural History Museum (London, UK) on 9–10 September 2014 to discuss the latest findings on plant interactions with fungi (Eumycota) and oomycetes (Oomycota = Peronosporomycota), with topics ranging from the fossil record and comparative genomics to symbiosis and phytopathology. The discussions were largely disseminated via social media (Box 1). Highly diverse plant–fungal interactions have formed the backbone of land ecosystems and biogeochemical cycles since the Palaeozoic (see Fig. 1 for geological timeframe). As summarized by Christine Strullu-Derrien and Paul Kenrick (The Natural History Museum, London, UK) the first land plants arose c. 470 million years (Myr) ago (Kenrick et al., 2012; Edwards et al., 2014), at which time fungi and oomycetes had already colonized terrestrial ecosystems. Following their terrestrialization, these microbes began to abound within plant fossils (Taylor et al., 2014, and references therein). Ultimately, biological interactions sculpted the genomes of plants, fungi and oomycetes (e.g. Schmidt & Panstruga, 2011; Kohler et al., 2015). Here we illustrate the picture that has emerged from the discussions at the 10th New Phytologist Workshop, and point to some pending questions. As it has now become routine at New Phytologist Workshops and Symposia and other scientific conferences, many delegates live-tweeted scientific highlights and other tidbits. Questions from the community were relayed to the speakers through Twitter. This resulted in broader dissemination of the workshop topic and engagement with many more scientists than those who attended the 10th New Phytologist Workshop. A picture of a 407 Myr old slab of Rhynie Chert, the earliest known ecosystem of plants and fungi, was popular in the Twittersphere and was repeatedly retweeted. The #NPW10 tweets were archived in the Storify platform and received > 10 000 page views a month after the workshop (https://storify.com/KamounLab/npw10-10th-new-phytologist-workshop-origin-and-ev). ‘In a rare gathering, genomics met palaeontology at the 10th New Phytologist Workshop.’ The emergence of land plants is often viewed as a symbiosis between a green alga with photosynthetic abilities that is unable to access water and minerals in the substrate, and a heterotrophic fungus predisposed to exploit soil resources (e.g. Selosse & Le Tacon, 1998; Selosse & Strullu-Derrien, 2015). Lichens have existed since the Devonian (Honegger et al., 2013), but their apparent low frequency in the fossil record has argued against a major role during terrestrialization. As Marc-André Selosse (Muséum national d'Histoire Naturelle, Paris, France) and Dianne Edwards (University of Cardiff, UK) discussed, the affinities of the enigmatic Nematophytales – an extinct group that includes the gigantic Prototaxites, a widespread Devonian columnar organism, which could reach up to 9 m in height and 1 m in diameter – are still debated (Edwards et al., 2013). A fungal affinity is suspected because of their distinctive filamentous organization, but since the contemporary plant biomass was likely insufficient to support large heterotrophs, it has been suggested that Nematophytales had photosynthetic abilities (Edwards et al., 2013), or even a lichen-like nature (Selosse, 2002). Although remains of photobionts are still controversial, recent research bears out the lichenized nature of the much smaller thalloid forms (Edwards et al., 2013; Honegger et al., 2013). Together with evidence from other fossils (e.g. Winfrenatia reticulata; Taylor et al., 1997), it now seems probable that lichens were a component of early terrestrial ecosystems, which bore a resemblance to modern cryptogam covers. The terrestrialization of land plants (Embryophyta) is also a story of fungal symbiosis. Arbuscular mycorrhizas (AM) formed by Glomeromycetes are widespread in living plants, and Silvia Pressel and Jeffrey Duckett (The Natural History Museum, London, UK) summarized recent evidence from most basal lineages (e.g. liverworts, hornworts; Desirò et al., 2013). Similarly, Maarja Öpik presented an overview of species diversity and modern distributions of AM fungi, based on a database hosted by her at the University of Tartu, Estonia (Öpik et al., 2013). As noted by Martin Bidartando (Imperial College, London, UK) and others, the AM symbiosis can be considered ancestral in land plants (Bonfante & Selosse, 2010), and Christine Strullu-Derrien illustrated further support for this view, based on new fossils from the 407 Myr old Rhynie Chert (Strullu-Derrien et al., 2014). Here, the plant hosts were rootless, and arbuscules were formed in photosynthetic aerial axes, a mode of infection unknown in modern vascular plants but closely resembling that in living liverworts and hornworts. There is also a growing body of molecular developmental evidence: Medicago mutants deficient for the pathway required to perceive Glomeromycete signals that are critical for fungal colonization can be rescued by ortholog genes from liverworts or hornworts, supporting the ancestral origin of the plant–Glomeromycetes symbiosis (Wang et al., 2010). Research on the origin of the genes acting in this pathway now focuses on algal lineages related to land plants, such as charophytes. Pierre-Marc Delaux (University of Wisconsin, Madison, WI, USA) suggested a stepwise evolution of the plant symbiotic ‘toolkit’ in algal ancestors, with several components predating the first land plants (Delaux et al., 2012). Elements of this ‘toolkit’ may, therefore, first have facilitated the interactions between aquatic charophytes and diverse symbiotic microorganisms, later being recruited and further developed for AM evolution on land. A broader survey of the distribution and function of these genes within green algae is now desirable, and the investigation of living and fossil Charophyta–fungus interactions may offer further insights (Taylor et al., 1992). Similarly, the strigolactones, another key component of the symbiotic plants–Glomeromycetes dialogue (Delaux et al., 2012), might have been involved initially in non-symbiotic signalling pathways. Results are beginning to show how the origins of the finely interwoven relationship between Glomeromycetes and plants can be unpicked though a combination of palaeontological and molecular developmental approaches. Recently, however, a new chapter in the story of land colonization through symbiosis was opened by two nearly concomitant discoveries: first, as Martin Bidartondo showed, extant basal land plants, such as liverworts, hornworts and lycopods, associate with Mucoromycetes (Desirò et al., 2013; Rimington et al., 2015) in a symbiosis whose mutualistic nature is suspected in some cases at least (Field et al., 2014); and second, Christine Strullu-Derrien showed a similar symbiosis in Devonian fossil land plants (Strullu-Derrien et al., 2014). The reasons for such an ancient co-existence of two symbiotic lineages, the Glomeromycetes and the Mucoromycetes (two sister lineages; Tisserant et al., 2012; Lin et al., 2014), remain unclear: are their effects on the host plant redundant, which would make their co-existence more difficult to explain, or more likely, do they provide different benefits to the host? To answer these questions, more knowledge on extant Mucoromycetes associations would be desirable. In addition to deriving mutual benefits through symbiosis, the rise of the cooperation between fungi and land plants is now thought to have had major impacts on biogeochemical cycles. Christopher Berry (University of Cardiff, UK) showed that complex plant communities with high biomass, such as forest-type ecosystems, arose as early as the Mid-Devonian (Stein et al., 2012), probably in part due to efficient nutrient extraction by fungi supported by plant photosynthates. Two important consequences of the evolution of forests were increased weathering of calcium–magnesium silicates and the release of growth-limiting nutrients such as phosphorous from surface rocks, with diverse consequences that were emphasized by Timothy Lenton (University of Exeter, UK) and Jonathan Leake (University of Sheffield, UK). First, the higher land-to-ocean calcium–magnesium flux increased ocean carbonate precipitation, drawing down atmospheric CO2. When incorporated into models of the geochemical carbon cycle, this effect equates to as much as an 80% drop in atmospheric CO2 following the rise of land plants, contributing to a global cooling and possibly to the Late Ordovician glaciations (Bergman et al., 2004; Lenton et al., 2012). Second, Timothy Lenton stressed that biomass production, and the burial of a portion in sediments, caused a rise of atmospheric O2, with many biotic and abiotic consequences, such as the radiation of large predatory fish with high oxygen requirements (Dahl et al., 2010) and the development of wildfire in terrestrial ecosystems, first seen in Late Silurian charcoals (Lenton, 2001; Scott et al., 2014). Following terrestrialization, changes in atmospheric CO2 concentrations through the Phanerozoic are thought to have exerted critical controls on rates of plant-driven fungal weathering via photosynthate fluxes, regulating the land-to-ocean calcium, magnesium and phosphorous fluxes and stabilizing atmospheric chemistry. As noted by Jonathan Leake, the evolution of ectomycorrhizas (EM), where Ascomycetes or Basidiomycetes interact with vascular plants (gymnosperm and angiosperm tree taxa), appears to have selected for intensification of weathering, since EM display higher weathering abilities than AM roots (Quirk et al., 2012, 2014; Thorley et al., 2014). Direct fossil evidence of EM is sparse, with the oldest records dating to the Eocene (LePage et al., 1997). However, fossils of EM hosting plant families appear in the fossil record or diversified much earlier during the Cretaceous (see Fig.1 for geological time frame; Taylor et al., 2011). Modelling approaches suggest that the rise of EM and the associated increase in weathering would have contributed to the CO2 drawdown observed during the Cretaceous and over the past 120 Myr (Taylor et al., 2011). This may be one of the causes for the lasting trend of reduction in atmospheric CO2 and consequent climate cooling since the Paleocene–Eocene thermal maximum. In temperate soils, climatic conditions limit mineral weathering and microbial mineralization, so that EM plants, whose fungi have increased access to mineral resources as well as to organic nitrogen and phosphorus (see later), are particularly well adapted in these environments where they now dominate (Read, 1991; Selosse & Le Tacon, 1998). Understanding the relationships and feedbacks among palaeo-environmental conditions, biogeochemical cycles, and the evolution of the land flora is an exciting and developing area (Beerling & Berner, 2005). As Timothy Lenton pointed out, modelling is a powerful tool for testing how the successive stages in the evolution of plant–fungal interactions contributed quantitatively to changes in Earth climate and geochemistry (Bergman et al., 2004; Taylor et al., 2011). The evolution of land plants and their ancestors selected new enzymatic abilities in fungi feeding on plant tissues. For example, Mary Berbee (University of British Columbia, Vancouver, Canada) showed that pectinase genes are homologous in Ascomycetes, Basidiomycetes, and early diverging fungi, including extant aquatic species living on algal remains: molecular clock dating indicates that 750 Myr, the estimated maximum age of origin of the pectin-containing ancestors of land plants, may represent the maximum age for the fungal ancestor that evolved such pectinases. Understanding this evolution again calls for more research on the interactions of extant aquatic fungi with algae closely related to land plants, and their use of algal pectin equivalents. Later, land plants evolved new polymers such as lignin that in turn selected for new saprotrophic abilities in fungi. David Hibbett (Clark University, Worcester, MA, USA) reported that, based on comparative analyses of Agaricomycetes genomes (Basidiomycetes; Floudas et al., 2012; Kolher et al., 2015), peroxidases allowing lignin degradation by white-rot fungi might have originated during the Carboniferous period. The diversification of ligninolysis coincided with the decrease in the rate of organic carbon burial around the end of the Permo-Carboniferous, and the possibility of a causal role was much debated during the workshop. One caveat is our ignorance of when lignin degradation emerged in Ascomycetes, which can explain the report of white-rot as early as in the Upper Devonian (see Stubblefield et al., 1985; although some doubt persists over the interpretation of the fossil fungi in this study). Moreover, the classical view of coal formation in wetlands under high rainfall and associated with tectonically induced subsidence (e.g. forearc basins or rift basins) holds for coals that accumulated later in geological history (e.g. Cretaceous coals; McCabe & Parrish, 1992). Even nowadays, lignolysis requires high levels of oxygen and does not occur under the waterlogged conditions dominating the mire sediments where Carboniferous coals were formed. Thus, whereas the geological context of the formation of Euro-American Carboniferous coals remains unchallenged, the quality or the abundance of white-rot fungi could have placed further limits on organic matter recycling. Further research on fossil fungi in Late Palaeozoic ecosystems and on the evolution of white-rot in Ascomycetes will clarify the relative impacts of fungal evolution on carbon sequestration, especially during the Carboniferous. Not all fungi live on decaying land plants as saprotrophs: some evolved direct interactions with living tissues (biotrophy), and considerable genomics advances on biotrophic fungi were reported in London. Francis Martin (INRA, Nancy, France) explained how the evolution of EM, which occurred more than 80 times within Ascomycetes and Basidiomycetes (Tedersoo & Smith, 2013), entailed genome erosion with extensive loss of lignocellulose decay genes. Phylogenomic reconstructions suggest that extant EM clades derived with repeated gene loss from ancestral lineages of white-rot, brown-rot and litter decayers (Plett & Martin, 2011; Floudas et al., 2012; van der Heijden et al., 2015; Kohler et al., 2015). This nicely explains why this transition occurred so frequently, whereas the reversion to saprotrophy, which would require the gain of many genes, is hitherto unknown. Nevertheless, some EM lineages have retained some plant cell-wall degrading capabilities, which they do not utilize to assimilate carbon substrates but to access organic nitrogen and phosphorous (Rineau et al., 2013). As Francis Martin pointed out, the ability to degrade plant cell wall materials found in some EM fungi means that they contribute to decomposition, but not that they are saprotrophs sensu stricto, since the mutualistic association remains their main or sole source of carbon. Polyphyletic evolution of the EM lifestyle is also marked by rapid genetic turnover in symbiosis-induced genes, some of which may reflect lineage-specific functional innovations, such as mycorrhiza-induced small secreted effector proteins (MiSSPs; Kohler et al., 2015). Some of these effector proteins in EM fungi, such as MiSSP7, locally control plant immunity (Plett et al., 2014); interestingly, this represents a convergence with effectors reported from Glomeromycetes, such as SP7 that has a role similar to MiSSP7 (Kloppholz et al., 2011). Such effectors, in the similar context of a loss of the enzyme required for free life, are also abundantly found in the genomes of parasitic fungi (Schmidt & Panstruga, 2011), and represent a general adaptation to biotrophic life, as detailed in a recent Virtual Special Issue of New Phytologist (Kuhn & Panstruga, 2014). Plant parasitism evolved independently in the oomycetes, a group of filamentous heterotrophic stramenopiles (Heterokonts) phylogenetically unrelated to Eumycota. Oomycetes, such as the Irish potato famine agent Phytophthora infestans, are common in modern ecosystems and often cause destructive epidemics in agriculture. The 10th New Phytologist Workshop highlighted their evolution to parasitism. Although distinctive features of oomycetes are somewhat difficult to detect in fossils, Christine Strullu-Derrien and Ben Slater (University of Birmingham, UK) reported recent discoveries of well-preserved materials that place the origin of oomycete–plant associations in the Devonian, if not earlier (Taylor et al., 2006). Early plant-associated oomycetes colonized tissues of lycophytes, ferns, seed ferns and gymnosperms in Late-Palaeozoic–Early Mesozoic ecosystems and by that stage exhibited extensive diversification (Strullu-Derrien et al., 2011; Slater et al., 2013). The earliest evidence of parasitism dates back to the Carboniferous (c. 300 Myr; Strullu-Derrien et al., 2011). However, how and when plant parasitism evolved in oomycetes remains unclear. On the one hand, it may have directly evolved from an ancestral algal parasitism, since many basal oomycete lineages are aquatic algal parasites (Beakes et al., 2012). On the other hand, a secondary evolution from land saprotrophic oomycetes cannot be excluded, and would account for the absence of direct evidence for parasitism in the most ancient fossils. The evolution of plant parasitism in oomycetes may have been partly fuelled by gene transfer from pre-existing parasitic fungi (Richards et al., 2011), and obligately parasitic oomycetes lost genes and pathways related to saprophytic life (Baxter et al., 2010; Kemen et al., 2011). Sophien Kamoun UK) reported that some such as infestans, evolved genomes with a distinctive in which adaptation and with the host plants over time et al., 2010; & 2012). are in genes that are involved in host interactions and in such as effectors, in with the and host to unrelated species & 2012; et al., 2014). oomycetes between and symbiosis. (University of UK) reported that Medicago mutants deficient in AM colonization also display by the oomycete Phytophthora et al., 2015), supporting the that have the pathways by which plants mutualistic One of these is in the of that colonization by AM fungi and (Wang et al., 2012). The is that has evolved to AM symbiosis, and was by oomycetes to host roots (Wang et al., 2012). Thus, oomycetes have evolved in many to fungi in their ability to interact with plants, especially that plants have evolved general to In the further on evolution between oomycetes and fungi will from palaeontological and but also from the use of common hosts as for their et al., 2015). The combination of palaeontological and approaches in the workshop illustrated how the of plants, fungi and oomycetes, over > Myr, has their evolution and the geochemistry of ocean and key would from further First, the in which and their is of the of the fossil record for in for Second, since all have a a knowledge of their interactions in extant aquatic environments is to the origins of their terrestrial interactions – and now be and are beginning to in of cryptogam – including their fungal component – in early terrestrial ecosystems, the to which these modern and the place of lichen-like associations remain questions. of the of the to geochemical cycles through the Phanerozoic is and we now to further of geochemical weathering under different and we a of the of lignocellulose during the to which may have to the modern carbon are of key to geochemical Earth that the in palaeontology as or and the of genomics (e.g. with the van der Heijden et al., 2015) will the in London, further and our view of the between plants, fungi and The the of the New Phytologist and all for their to the of this exciting workshop. The also the for their of this is part of the The of is supported by the received support from the under the