Metatranscriptomics reveals unsuspected protistan diversity in leaf litter across temperate beech forests, with Amoebozoa the dominating lineage

Forest litter harbors complex networks of microorganisms whose major components are bacteria, fungi and protists. Protists, being highly selective consumers of bacteria and fungi could influence decomposition processes by shifting competitive
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  FEMS Microbiology Ecology , 95, 2019, fiz142 doi: 10.1093/femsec/fiz142 Advance Access Publication Date: 0 2019Research Article RESEARCH ARTICLE Metatranscriptomics reveals unsuspected protistandiversity in leaf litter across temperate beech forests,with Amoebozoa the dominating lineage Christian Voss 1,2 , Anna Maria Fiore-Donno 1,2, * , † , MarcoAlexandre Guerreiro 3 , Derek Per ˇ soh 3 and Michael Bonkowski 1,2 1 Terrestrial Ecology Group, Institute of Zoology, University of Cologne, Cologne, Germany,  2 Cluster of Excellence on Plant Sciences (CEPLAS), Cologne, Germany and  3 Department of Geobotany, Faculty of Biologyand Biotechnology, Ruhr-University of Bochum, Germany ∗ Corresponding author: Anna Maria Fiore-Donno.  E-mail: One sentence summary:  Metatranscriptomics reveals Amoebozoa as the main protistan lineage in one-year-old beech litter. Editor:  Petr Baldrian † Anna Maria Fiore-Donno, ABSTRACT Forest litter harbors complex networks of microorganisms whose major components are bacteria, fungi and protists.Protists, being highly selective consumers of bacteria and fungi could influence decomposition processes by shifting competitive microbial interactions. We investigated the eukaryotic diversity from 18 samples of one-year beech ( Fagussylvatica ) leaf litter by RNA-based high-throughput sequencing of the small-subunit ribosomal RNA gene. By applying ametatranscriptomics approach, we avoided biases inherent to PCR-based methods, and could therefore focus on elusiveprotistan groups. We obtained 14 589 eukaryotic assembled sequences (contigs) representing 2223 unique taxa. Fungidominated the eukaryotic assemblage, followed by an equal proportion of protists and plants. Among protists, the phylumAmoebozoa clearly dominated, representing more than twice the proportion of Alveolata (mostly ciliates) and Rhizaria(mostly Cercozoa), which are often retrieved as the dominant protistan groups in soils, revealing potential primer biases. Byassigning functional traits to protists, we could assess that the proportion of free-living and heterotrophs was much higherthan that of parasites and autotrophs, opening the way to a better understanding of the role played by the protistancommunities and how biodiversity interacts with decomposition processes. Keywords:  biodiversity; protists; litter; decomposition; metatranscriptomics; functional traits INTRODUCTION Tree litter decomposition has long been recognized as an essen-tial process for recycling nutrients previously immobilized byplants (H ¨ attenschwiler, Tiunov and Scheu 2005). Microorgan-isms are the main drivers of this process, which is characterizedby a succession of bacterial and fungal decomposers with shift-ing enzymatic capabilities, each degrading specific substrates(Baldrian 2017). Bacteria dominate in abundance during thewholeprocess(Purahong  etal. 2016),althoughfungidominateinbiomass with a 60/40% fungal/bacterial ratio of microbial tran-scripts ( ˇ Zif  ˇ c´akov´a  et al.  2017). The energy obtained from litterdecomposition is channeled to higher trophic levels in the soilfood web (Rooney  et al.  2006), with protists seemingly playing an important role (Schr ¨ oter, Wolters and De Ruiter 2003). How-ever, studies based on microscopical observation often focus Received:  27 June 2019;  Accepted:  25 August 2019 C  FEMS 2019. All rights reserved. For permissions, please e-mail: 1  2  FEMS Microbiology Ecology , 2019, Vol. 95, No. 00 on relatively easily identifiable taxa, such as ciliates (Foissner et al.  2005) or testate amoebae (Ehrmann  et al.  2012). Nonethe-less, even studies based on coarse identification of morpho-types (flagellates, ciliates and amoebae) suggest the existenceof rich and dynamic communities of litter protists (Krivtsov  et al. 2003). Studies based on molecular methods described the com-plex dynamics of bacterial and fungal interactions during lit-ter decomposition (Schneider  et al.  2012; Vo ˇ r´ı ˇ skov´a and Baldrian2013). More recently, the succession of bacteria, fungi and pro-tists during decomposition was investigated, showing the pro-tistan taxa Amoebozoa and Cercozoa to dominate in litter agedfrom one to six months. In addition, their relative abundancesdiffered between substrates, probably in response to C and Navailability and/or to related shifts in the composition of othermicrobial communities (Bonanomi  et al.  2019).In soil, the release of ammonia from protists after consump-tionof bacteria, i.e.themicrobial loop,was longknownasacen-tralmechanismofplantNacquisition(BonkowskiandClarholm2012).However,morerecentstudieshaveshownthatprotistsactat multiple levels in the soil food web, not only consuming bac-teria but also fungi, other protists and even microfauna (Geisen et al.  2016; Bonkowski, Dumack and Fiore-Donno 2019). As selec- tive grazers, they can modify the composition of microbial com-munities (Rosenberg   et al.  2009; Gl ¨ ucksman  et al.  2010; Flues,Bass and Bonkowski 2017) and increase the plant N uptake andbiomass (Bonkowski, Griffiths and Scrimgeour 2000; Bonkowski, Jentschke and Scheu 2001; Koller  et al.  2013). Specific bacteria orfungi, in turn, modify the composition of protistan communi-ties (Xiong   et al.  2018) presumably by developing antipredationstrategies(Jousset2012).Besidesfeeding,protistsinteractwithinthe soil in complex biological networks as antagonists or syn-ergists, parasites or mutualists (Koller  et al.  2013; Kramer  et al. 2016; Hassani, Dur´an and Hacquard 2018), ultimately influenc-ing ecosystem functioning. Because of the increasing evidencethat protists are abundant in different types of litter (Bonanomi et al.  2019), they could play an important role in shaping bacte-rial and fungal communities, in turn affecting the decomposi-tion dynamics.Especially because biodiversity loss has consistent conse-quencesforlitterdecomposition(Gessner etal. 2010;Handa etal. 2014), there is a need to precisely identify the major players indifferent types of litter and how they interact within the litterand the soil microbiome. The potential for specific organismsto contribute to ecosystem processes can be revealed by assign-ing functional traits to them. This strategy has been applied tobacteria in litter (Allison 2012) but, to our knowledge, not yet toprotists in litter. Examples of studies using functional traits of protistsincludetestateamoebae,allowingtosuccessfullyrecon-struct past disturbances in peatlands (Marcisz  et al.  2016) andtheir dependence to plant functional types (Jassey  et al.  2014).Complex predator–prey interaction models could be assessed inciliates (Tirok and Gaedke 2010) .  Different responses to environ-mental gradients, e.g. moisture in grasslands, could be found inCercozoa (Fiore-Donno  et al.  2019).We applied high-throughput sequencing and metatranscrip-tomics on 18 samples of one-year-old beech ( Fagus sylvatica )leaf litter to investigate the eukaryotic diversity, focusing onthe protistan communities. Our study sites are part of a large-scale project, the German Biodiversity Exploratories (Fischer et al.  2010), and were located in different beech forests in north-ern, central and southern Germany. We analyzed metatran-scriptomic data from a previous study of active fungal com-munities in beech leaves and litter (Guerreiro 2019). Our mainaim was to identify the dominant protistan lineages and theirfunctional traits (free-living phagotrophs, animal or plant par-asites, autotrophs), by analyzing the small-subunit ribosomalRNA gene sequences (SSU or 18S). Our work represents aninsight using metatranscriptomics into the biodiversity of lit-ter eukaryotes. By adding functional traits to litter inhabiting protistan groups, we made a significant contribution to estimat-ingfunctionalmicrobialdiversityandultimatelytoourabilitytopredict consequences of biodiversity loss in the litter decompo-sition processes. MATERIALS & METHODS Study area, sample collection, RNA extraction andsequencing Litter samples were collected as in Guerreiro (2019) and pro-cessed as previously described (Guerreiro  et al.  2018; Guerreiro2019). The study sites were located in forest plots dominatedby European beech ( F. sylvatica ), selected within the frameworkof the German Biodiversity Exploratory project (Fischer  et al. 2010). In each of the widely separated (300 km) regions of theBiodiversity Exploratories, one-year-old litter samples were col-lected between the 20th and the 25th of October 2014, in 100 m 2 plots, nine in Hainich, six in Schorfheide-Chorin and three inthe Swabian Alps, 18 samples in total. Each plot was sampled infive positions (corners and center, five subsamples of which twoto three were merged and used for nucleic acids coextraction).Samples were stored and transported at 4 ◦ C to the laboratoryand were processed within 12 hours. DNA and RNA were coex-tracted from 0.2 g of litter, by adapting the protocol from Per ˇ soh et al.  (2008) and Guerreiro  et al.  (2018), with special care to avoid-ing coextraction of humic acids.To coextract the nucleic acids, 700 µ l of tris-buffer (0.1 M, pH8) and 35  µ l of aluminum sulfate [Al 2 (SO 4 ) 3 ; 4 M] were added toeach 0.2 g of combined litter sample. Each sample was homoge-nizedinaFastPrep  R  -24Instrument(MPBiomedicals,Eschwege,Germany) for 20 s at 4.0 m s − 1 . The pH was adjusted to 8 byaddingNaOH(4M),andsubsequentlyreadjustedto5.5byadding 3.0 µ l of acetic acid. A mixture of glass beads (0.15 g of Ø 0.1–0.25mm; 0.12 g of Ø 0.25–0.5 mm; 0.12 g of Ø 1.25–1.55 mm) and steelbeads (3 beads of Ø 3 mm), 325 µ l of extraction buffer [0.4 M LiCl,200 mM Tris-HCl (pH 8), 120 mM EDTA (pH 8), 10% SDS] and 500 µ l of phenol were added to the sample. Samples were disruptedusingaFastPrep-24Instrument(MPBiomedicals)for30sat4.0ms − 1 , followed by two steps of 40 s at 6.5 m s − 1 , with an intermedi-ateincubationoniceof5min.Aftercentrifugation(14165gfor1 ′ atroomtemperature),750 µ lphenol:chloroform:isoamylalcohol(25:24:1 v/v) was added to 750 µ l of the supernatant. The samplewas incubated on ice for 5 ′ and mixed every 90 ′′ . After centrifu-gation (17 650 g for 15 ′ at 4 ◦ C), three consecutive centrifugationsteps(17650gfor15 ′ at4 ◦ C),withadditionofanequalvolumeof chloroform:isoamyl alcohol (24:1) and intermediate incubationson ice for 5 ′ , were performed. Nucleic acids were precipitated byadding0.1volumeofNaCl(5M)and0.7volumeofisopropanolto500 µ l of the supernatant. After overnight incubation at  − 20 ◦ C,the sample was centrifuged (17 650 g for 60 ′ at 15 ◦ C), and thepellet was air dried for 3 ′ and resuspended in 20  µ l of RNase-free water.Total RNA was purified by incubating 16  µ l of nucleic acidswith 2 µ l of DNase (1 unit/ µ l, Sigma-Aldrich, Munich, Germany),in 2  µ l of the corresponding buffer, for 15 ′ at room tempera-ture, followed by EDTA adjunction and incubation at 70 ◦ C tostop the reaction. Metatranscriptomic libraries were preparedusing NEBNext  R   Ultra Directional RNA Library Prep Kit for  Voss  et al.  3 Illumina  R   (New England Biolabs GmbH, Frankfurt am Main,Germany), according to manufacturer’s recommendations. Thepoly(A) mRNA workflow was used, which depletes the librariesfrom the dominating rRNAs, but still allows a fraction of themto be sequenced. Although the NEBNext library preparation kitgave consistent results, relatively even transcript coverage andlow rates of PCR duplicates (Vecera  et al.  2019), this procedurecould have an unmeasurable effect on the relative abundancesof the retrieved taxa in our study.The libraries were sequenced with an Illumina MiSeqsequencer (Illumina, Inc., San Diego, CA, USA), with 2  ×  300 bppaired end sequencing (MiSeq Reagent Kit v3 Chemistry, Illu-mina, Inc., San Diego, CA, USA). The raw sequence data weredeposited in EBI Metagenomics database (,accessionnumbersPRJEB7861andPRJEB8419);the assembled contigs were deposited in ENA under the acces-sion number PRJEB33161. Bioinformatics pipeline, taxonomicand functional assignment.The obtained paired end reads were assembled  de novo  byCLC Genomics Workbench (CLCbio, Qiagen, Valencia, CA, USA),using default settings for trimming, automatic word and bub-ble size and a minimum contig length of 110 bp. To determinethe coverage of each contig, the reads were mapped against theassembled contigs. SSU contigs were taxonomically identifiedin the PR 2 database, version gb203 (Guillou  et al.  2013) using BLAST +  (Camacho  et al.  2008) with an E-value of 1E + 50 andkeepingonlythebesthit.Wethenassignedfunctionalclassestoeachtaxonomicentry,distinguishingbetweenlivingandfeeding modes. As living modes, we selected free-living, plant parasiticand animal parasitic or endosymbiotic, while feeding modesweredividedbetweenheterotrophic,autotrophicormixotrophic(Table S2, Supporting Information). Functional assignment wasmainly based on Adl  et al.  (2019), except for Cercozoa, for whichwe followed Fiore-Donno  et al.  (2019). Table S2 (Supporting Infor-mation) was also deposited in the Biodiversity Exploratoriesdatabase (dataset 25 407,, last accessed June 2019) and it will be publiclyaccessible after an embargo of three years. Please note that thetaxonomic relative abundance may be biased by differences inrRNAcontentpercellbetweentaxa(WeberandPawlowski2013). Statistical analyses We investigated if the communities of protists were influencedbytheregionofcollection,thetypeofsoilaccordingtotheWorldReference Base (WRB) and forest management (Table S1, Sup-porting Information—parameters measured in the frameworkof the Biology Exploratories, extracted from dataset 20826 avail-able at, last accessed July 2019). Statistical analyses were conductedusing R v. 3.4.0 (R studio v. 1.0.153) and the following functionsin the ‘vegan’ package (Oksanen  et al.  2013), based on the Bray–Curtis dissimilarity distance matrix of the 18 samples (999 per-mutations,  α  = 0.05). To evaluate whether more sampling effortwould have revealed more richness, we carried out a rarefactionanalysis function  specaccum . Diversity indices and nonmetricmultidimensionalscaling(NMDS)plotswerecomputedwiththefunction  metaMDS . Permutational Multivariate Analysis of Vari-ance (PERMANOVA),  adonis  function, was generated to test forsignificant differences in the litter community structure acrossregions, management types and soil types. RESULTS AND DISCUSSION Summary of taxonomic assignments The metatranscriptomics approach generated 1 287 115 contigswith a minimum length of 200 bp and a maximum of 16 443bp (mean: 365.6, SD:  +  / −  187.1). We were provided with 14 589unique eukaryotic SSU contigs, representing c. 38 million readsand 2223 unique BLAST hits. The 14 589 unique eukaryotic con-tigs assigned as SSU comprised a minimum length of 200 bp anda maximum of 4213 bp (mean: 680.4, SD:  +  / −  384.8). Species–rarefaction curves showed that our sampling was sufficient andthat the extant diversity would already have been revealed with10 samples (Figure S1, Supporting Information). Sequence sim-ilarities with the best BLAST hit ranged from 100 to 74%; onlya minority were 100% similar to an existing sequence (133 con-tigs) (Figure S2, Supporting Information). A large majority of thecontigs was only  < 97% similar to known sequences (the usualthreshold to identify operational taxonomic units) (Grossmann et al.  2016; Muller Barboza  et al.  2018) and therefore likely to rep-resent the unknown (or not yet sequenced) eukaryotic diversityinlitter.Weobtainedonaverage11522contigspersample(max-imum = 13 083, minimum = 9617, SE = 279.63) (Figure S3, Sup-porting Information). Eukaryotic diversity in beech leaves litter Our study, based on metatranscriptomics, revealed that Fungidominated the eukaryotic assemblage (42% of the contigs, 58%of the total reads), with a dominance of Ascomycota (23% of thecontigs) followed by Basidiomycota (11% of the contigs) in con-cordance with what is expected in one-year-old litter (Vo ˇ r´ı ˇ skov´aand Baldrian 2013; Purahong   et al.  2016) (Fig. 1A). We also found 15 contigs attributed to the basal lineage Nuclearida (Holomy-cota), single-celled amoebae at the base of the fungi. Nucle-arida have been described from aquatic environments, but wererecently also reported from litter and mosses (Heger  et al.  2018).Plantae (24% of the contigs/23% of the reads) were repre-sented by a majority of Streptophyta (22% of the contigs, mostlyEmbryophyceae)andaminorityofChlorophyta(2%),mostlyTre-bouxiophyceae. While the latter, mostly unicellular algae, maybe active in the litter, we assume that the sequences of the vas-cular plants came from pollen.Metazoa (9% of the contigs/12% of the reads) were mostlyrepresented by Mollusca (3%) and Nematoda (2%). The pres-ence of Bryozoa (1%) in a nonaquatic habitat is puzzling. Ourcontigs match a single reference sequence, EU650324  Plumatella sp., a genus of freshwater bryozoans. However,  Plumatella  pro-duces massive amounts of dormant stages in floating form thatcould spread to soils. Arthropoda accounted for 1%, with 46%Arachnida (Fig. 1A), of which the majority were soil mites (TableS2, Supporting Information). The choanoflagellates (belonging to Metazoa but ‘protists’ since unicellular) were represented by75 contigs. Formerly considered to be exclusively aquatic, theirpresence in terrestrial environments and especially in forestsis increasingly evidenced by environmental PCR studies (Geisen et al.  2015b; Ferreira de Araujo  et al.  2018; Heger  et al.  2018). Inconformity with Heger  et al.  (2018) and Ferreira de Araujo  et al. (2018),wefoundsequencesrelatedtotheaquaticholozoanpara-sites Ichthyosporea (23 contigs, 14 taxa) and Filasterea, with fivecontigs assigned to  Capsaspora owczarzaki , a parasite of fresh-water slugs. In Ichthyosporea, we found sequences of   Anurofecarichardsi , which was isolated from tadpole faeces (six contigs),  Amoebidium parasiticum  (four contigs) a parasite of freshwater  4   FEMS Microbiology Ecology , 2019, Vol. 95, No. 00 Figure 1.  Sankey diagrams of the diversity of the eukaryotic groups found inthe 18 samples of one-year-old beech leaf litter.  (A)  Relative abundance of theeukaryotic main lineages.  (B)  Living modes.  (C)  Feeding modes. Only groups rep-resented by  > 1% of all contigs are shown. arthropods, and the fish parasite  Sphaerothecum destruens  (threecontigs). Together these results suggest that at least resting stages of Ichthyosporea and Filasterea (perhaps present in fae-cesofbirdsoranimalpredatorsoftheirhostspecies)arepresentin soil. Another hypothesis would be that the host range of Ichthyosporea and Filasterea is much wider than assumed andincludes terrestrial animals. Living and feeding modes Trait-based approaches allow a much deeper understanding of the ecological processes than information on taxonomy alone(Krause  et al.  2014), since functional traits not only reflect adap-tations of individual organisms to physical, chemical and bio-logical pressures of their environment, but also determine howtheorganismsaffectecosystemfunctioning.Definingfunctionaldiversity of microorganisms remains one of the biggest chal-lenges for trait-based approaches to microbial diversity, andallows answering questions such as the relative proportion of heterotrophs/autotrophs or free-living/parasites in litter.We found that free-living eukaryotes were much moreabundant (63%) than animal parasites or endosymbionts (22%)(Fig. 1B). The proportion of free-living protists was even higher:93.7%, while that of parasites and endosymbionts accountedonly for 3.8% (Table S2, Supporting Information). This is in sharpcontrast with the high abundance of animal parasites (Apicom-plexa)foundintropicalforestsoils(Mah´e etal. 2017)butinaccor-dance with Heger  et al.  (2018) who found a great majority of free-living protists in litter. Therefore it seems unlikely that animalparasitesdominateineveryterrestrialhabitat.Theirdominancein tropical forests may be linked to the high diversity of arthro-pods (Basset  et al.  2012) compared to temperate forests.The protistan plant parasites were not abundant (5%). Wehypothesized that they were either related to fine roots pene-trating the litter or more likely present as dormant spores orcysts. Plant parasites were found to decrease from grasslandsto forests (Ferreira de Araujo  et al.  2018).Among the eukaryotic feeding types, heterotrophs were themost abundant (75%) (Fig. 1C). This is accordance with littersupporting mainly saprotrophic taxa able to decompose recal-citrant plant-derived biopolymers (Baldrian 2017); and also tothe results of Heger  et al.  (2018), who found a great majorityof heterotrophs in litter and mosses, followed by autotrophs. Ingrasslands, similar proportions of heterotrophs and autotrophsamongeukaryoteswereretrieved(Venter etal. 2017).Inourdata,the autotrophs were also second (25%) and mostly composedof Plantae (Table S2, Supporting Information). The mixotrophscounted for  < 1% (thus not shown in Fig. 1C). Diversity of protists Protists were indisputably major players in the one-year-oldbeech leaf litter, representing nearly a quarter of the eukaryoticdiversity (24% of the contigs and 7% of the reads) (Fig. 1A). Othermetatranscriptomic studies have reported similar proportionsof protists in soils, from 16% (Urich  et al.  2008) to  > 20%, domi-nating the eukaryotic assemblage in arctic peat soils (Tveit  et al. 2013). Also some amplicon-based studies indicated the impor-tance of protists in surveys of eukaryotes in soil, e.g. in semi-arid plains (protists 31%, Fungi 29%) (Baldwin  et al.  2013), whilein most soils fungi appear to dominate, like in the tundra (48%fungi, 11% protists) and in the Brazilian Cerrado (41% fungi, 30%protists) (Ferreira de Araujo  et al.  2018).In litter protists, Amoebozoa was the most abundant group(37% of the protistan contigs/34% of the reads) dominating in allsamples, followed by Rhizaria (Fig. 2). To our knowledge, there istodateasinglestudyusingmetatranscriptomicstoassessbeechlitter diversity, and in this study Rhizaria was dominating (49%)over Amoebozoa (21%) (Geisen  et al.  2015b). A recent amplicon-based sampling of litter found Amoebozoa to dominate in mostlitter types, although not in N-rich litter, where Cercozoa weremore abundant (Bonanomi  et al.  2019). Together, these resultsshowthattheabundanceofthemajorprotistangroupsmayvaryeven within the same habitat, probably depending on complexinteractions between soil and litter biotic and abiotic factors,including climate. Most environmental studies using generaleukaryotic primers reported a dominance of Alveolata (ciliatesand apicomplexan) in soils (Baldwin  et al.  2013; Bates  et al.  2013;Shen  et al.  2014; Shi  et al.  2015; Mah´e  et al.  2017). Most likely, thisis due to an amplification artifact since most general eukary-otic primers have strong biases against Amoebozoa, and tendto favor Alveolata (Geisen  et al.  2015a; Fiore-Donno  et al.  2016;Bonkowski,Dumack andFiore-Donno2019). Alveolata, however,may dominate in peatlands (Tveit  et al.  2013; Geisen  et al.  2015b)and in tropical forest soils (Mah´e  et al.  2017). Amoebozoa  comprise a wide variety of amoebae and flag-ellates, many of which are key inhabitants of soil. They mea-sure from few microns to several meters across, are free-living or parasites. Amoebozoa may count  > 2400 species (Pawlowski  Voss  et al.  5 Figure 2.  Sankey diagrams of the diversity at the level of the classes and orders of the four main protistan groups.  (A)  Amoebozoa.  (B) . Rhizaria.  (C)  Excavata.  (D) Alveolata. Only groups represented by  > 1% of all contigs are shown.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!