Environmental Microbiology (2009) doi:10.1111/j.1462-2920.2009.02023.x Eukaryotic diversity and phylogeny using small- and large-subunit ribosomal RNA genes from environmental samplesemi_2023 1..10 William Marande, Purificación López-García and Introduction David Moreira* The use of molecular methods has played a major role in Unité d’Ecologie, Systématique et Evolution, UMR our understanding of microbial diversity. During the last CNRS 8079, Univ. Paris-Sud 11, bâtiment 360, 91405 two decades, PCR amplification and sequencing of the Orsay Cedex, France. small-subunit ribosomal RNA gene (SSU rDNA) has been extensively applied to survey the huge bacterial and Summary archaeal diversity found in many common and extreme habitats (Barns et al., 1996; Hugenholtz et al., 1998). The recent introduction of molecular techniques in Recently, this technique was also adopted for the analysis eukaryotic microbial diversity studies, in particular of eukaryotic diversity and it revealed a variety of new those based in the amplification and sequencing groups, such as the marine Group I alveolates (López- of small-subunit ribosomal DNA (SSU rDNA), has García et al., 2001; Moon-van der Staay et al., 2001) or revealed the existence of an unexpected variety of the putative new algal clade picobyliphytes (Not et al., new phylotypes. The taxonomic ascription of the 2007), as well as a large diversity within many well-known organisms bearing those sequences is generally groups such as the prasinophyte algae (Guillou et al., deduced from phylogenetic analysis. Unfortunately, 2004; Viprey et al., 2008). In addition, several SSU rDNA the SSU rDNA sequence alone has often not enough sequences that branched deeply in the eukaryotic phy- phylogenetic information to resolve the phylogeny of logeny were impossible to be specifically related to any fast-evolving or very divergent sequences, leading to known group, even at the kingdom level, suggesting the their misclassification. To address this problem, we controversial existence of several novel eukaryotic tried to increase the phylogenetic signal by amplify- kingdom-level groups (Dawson and Pace, 2002; ing the complete eukaryotic rDNA cluster [i.e. the SSU Edgcomb et al., 2002). rDNA, the internal transcribed spacers, the 5.8S rDNA A key component of all the SSU rDNA-based diversity and the large-subunit (LSU) rDNA] from environmen- studies is the taxonomic classification of the organisms tal samples, and sequencing the SSU and LSU rDNA whose sequences are retrieved, which is commonly part of the clones. Using marine planktonic samples, inferred by phylogenetic analysis of the novel environ- we showed that surveys based on either SSU or mental phylotypes. In addition to artefacts derived from SSU + LSU rDNA retrieved comparable diversity pat- undetected chimeric sequences, single-marker phylog- terns. In addition, phylogenetic trees based on the enies have a series of inherent methodological problems, concatenated SSU + LSU rDNA sequences showed such as the lack of enough informative signal for the better resolution, yielding good support for major resolution of deep nodes, which can engender artefac- eukaryotic groups such as the Opisthokonta, Rhizaria tual misplacement in phylogenetic trees (Philippe and and Excavata. Finally, highly divergent SSU rDNA Adoutte, 1998; Stiller and Hall, 1999; Philippe et al., sequences, whose phylogenetic position was impos- 2000). A typical phylogenetic reconstruction problem, the sible to determine with the SSU rDNA data alone, long-branch attraction (LBA) artefact, often causes highly could be placed correctly with the SSU + LSU rDNA divergent SSU rDNA sequences to be placed at the base approach. These results suggest that this method can of phylogenetic trees, leading them to be considered as be useful, in particular for the analysis of eukaryotic potential novel eukaryotic lineages. Recently, it was dem- microbial communities rich in phylotypes of difficult onstrated that from 28 published phylotypes representing phylogenetic ascription. putative novel high-level eukaryotic taxa, only 11 were actual potential new lineages (Berney et al., 2004; Received 22 August, 2009; accepted 26 June, 2009. *For correspon- dence. E-mail [email protected]; Tel. (+33) 169157608; Cavalier-Smith, 2004). For example, the phylotypes Fax (+33) 169154697. DH145-EKD11 and CCW75, putative members of a new © 2009 Society for Applied Microbiology and Blackwell Publishing Ltd 2 W. Marande, P. López-García and D. Moreira eukaryotic clade (López-García et al., 2001; Stoeck and whereas for the DH141 library, which appeared to show Epstein, 2003), were later recognized to be highly diver- the most even taxonomic diversity without any clearly gent phylotypes related to the very fast-evolving ciliates dominant group, we sequenced 96 additional clones (i.e. Myrionecta rubra and Mesodinium pulex (Johnson et al., 144 clones in total) of each type in order to have a library 2004). This is probably the case for many other divergent with a more exhaustive characterization for comparative eukaryotic phylotypes (Berney et al., 2004; Cavalier- purposes. For each SSU or SSU + LSU rDNA clone, we Smith, 2004). sequenced ~800 bp of the 5′ region of the SSU rRNA A natural way to alleviate the misplacement of environ- gene using the primer Euk-42F. A total of 508 partial mental sequences due to the intrinsic limitation of single- sequences were thus determined (see Table 1). The marker data would be to increase the number of overall taxonomic diversity retrieved was large, revealing phylogenetic informative sites in phylogenetic analyses by the presence of typical marine planktonic eukaryotic concatenating large-subunit (LSU) and SSU rDNA groups in the samples, including alveolates, cryptophytes, sequences. This strategy was applied recently, producing streptophytes, chlorophytes, haptophytes and heter- eukaryotic phylogenetic trees with much better resolution okonts, with variable frequencies depending on the than the SSU rDNA alone and better statistical support for sample analysed (Fig. 1). For each sample, we compared the major eukaryotic groups (Moreira et al., 2007). The the proportions of the different taxonomic groups for the LSU and SSU rRNA genes are generally adjacent in the two types of libraries, SSU and SSU + LSU rDNA. We genome, so that both markers could be easily retrieved by detected the groups that normally dominate surface PCR amplification also from environmental DNA samples. marine environments (Díez et al., 2001) and observed The feasibility of this approach has already been shown quite similar taxonomic compositions for the most abun- for marine planktonic bacteria (Suzuki et al., 2001) but dant groups in both types of libraries (Fig. 1). Differences never tried for eukaryotes. In this work, we analysed the between SSU and SSU + LSU rDNA libraries from a same results of parallel SSU rDNA and SSU + LSU rDNA sample mostly concerned the less abundant groups surveys to check whether the use of this larger marker (Table 2). This likely reflected, at least in part, that the retrieves comparable eukaryotic diversity profiles. In addi- library surveys were not exhaustive, namely that we had tion, we tested if the use of the SSU + LSU rDNA concat- not reached a plateau in the saturation curves for each enations allows the correct placement of fast-evolving library (data not shown). However, this was not the only phylotypes in eukaryotic phylogenetic trees. reason. For example, we did not retrieve any haptophyte sequence in the SSU + LSU rDNA libraries, despite the fact that members of this widespread photosynthetic Results and discussion group were identified, although in relatively small pro- portions, in all the corresponding SSU rDNA libraries. Diversity analysis and comparison between SSU rDNA An inspection of the available haptophyte LSU rDNA and complete rDNA cluster libraries sequences revealed that the eukaryotic ‘universal’ primer We successfully amplified the eukaryotic rDNA cluster 28S-4R, used for the construction of the SSU + LSU rDNA from DNA extracted from four different surface (25 m) libraries, showed two mismatches with those haptophyte marine planktonic samples (Table 1), obtaining PCR sequences, which probably explains the bias against the products of 5–6 kbp. To compare the genetic diversity haptophytes observed in the SSU + LSU rDNA libraries. retrieved from the complete rDNA cluster amplification Nevertheless, this was the only clear bias that we could with the classical SSU rDNA analysis, we generated in identify in the taxonomic composition, considered at large parallel SSU rDNA libraries from the same four samples. scale, in our libraries. Preferential PCR amplification and For the DH18, DH114 and Ma131 libraries, we selected cloning of rDNA clusters of small size is another possible 48 positive clones of each type (SSU and SSU + LSU), bias that may have escaped from our attention. It is known Table 1. Sampling sites and number of clones sequenced. Volume Clones sequenced Sample Station Coordinates Depth (m) filtered (l) (SSU/SSU + LSU) DH18 DHARMA 5 62°22′11′′ 53°35′56′′ 25 19 46/36 South Atlantic DH114 DHARMA 32 54°59′44′′ 58°22′17′′ 25 14 47/48 South Atlantic DH141 DHARMA 18b 59°22′45′′ 55°46′27′′ 25 14 138/114 South Atlantic Ma131 Marmara Sea 40°50.295 N 28°1.397E 25 4 42/37 © 2009 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology Eukaryotic diversity and phylogeny using SSU + LSU rDNA 3 90 DH141 90 DH114 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 Clone frequency (%) frequency Clone Clone frequency (%) frequency Clone 0 0 s tes tes mids ophytes Animal Alveolates Cercozoa Alveolates HeterokontsHapt Telone Heterokonts CryptophytesChlorophytes CryptophyChlorophytes Glaucophy Haptophytes Choanoflagellates 90 90 DH18 Ma131 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 Clone frequency (%) frequency Clone 0 (%) frequency Clone 0 es nts es tes es tes onts mals olaria Fungi Ani Alveolates Alveolat Radi Excava Heteroko Heterok CryptophytChlorophytes StreptophytHaptophytes CryptophyChlorophytes Haptophytes SSU rDNA clones SSU + LSU rDNA clones Fig. 1. Eukaryotic diversity identified using SSU and SSU + LSU rDNA libraries.