Samples Collection and Documentation

We sampled Archerella flavum and Amphitrema wrightianum (Figure 1) from wet Sphagnum mosses collected from the west shore of Duffey Lake, South Central British Columbia, Canada (50°23′ N 122°27′ W) and Praz-Rodet bog in the Jura Mountains of Switzerland (46°34′ N 6°10′ E). An authorization (No 1449) was delivered by the “Service forêts de la faune et de la nature du canton de Vaud” (state office for nature conservation) for sampling in Praz-Rodet for 2011 and 2012 (January 2011 to end of summer 2012). The Duffey Lake Sphagnum sample was not collected within the Provincial Park and therefore, no permits were required. Cells were extracted from Sphagnum mosses through serial of filtrations, and then were washed 3 to 4 times with distilled water [23], [24]. We prepared seven independent extractions, four from Archerella flavum and three from Amphitrema wrightianum, each of these extractions contained between 50 to 70 cells. Both species were documented using light microscopy (Figure 1).

DNA Extraction, PCR Amplification and Sequencing

DNA was extracted using a guanidine thiocyanate-based protocol [25]. Seven SSU rRNA sequences (four from Archerella flavum and three from Amphitrema wrightianum) were obtained by two amplifications. The first amplification was performed using universal eukaryotic primers, 1EKF (CTGGTTGATCCTGCCAG) and 1498R (CACCTA CGGAAACCTTGTTA) or 1520R (CYGCAGGTTCACCTA), in a total volume of 30 µl with amplification profile consisting of (5 minutes at 95°C followed by 40 cycles of 30 sec at 94°C, 30 sec at 58°C and 1 min 30 sec at 72°C with a final elongation of 10 min at 72°C). The positive products were cloned into pCR2.1 Topo TA cloning vector (Invitrogen) and transformed into E. coli TOP10’ One Shot cells (Invitrogen) according to the manufacturer’s instructions. Clone inserts were amplified with vector T7 and SP6 primers. The expected size clones from PCR amplifications were purified with the NucleoFasts 96 PCR Clean Up kit from Macherey-Nagel (Düren, Germany) and sequenced with an ABI PRISM 3700 DNA Analyzer (PE Biosystems, Genève, Switzerland) using a BigDyeTM Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems). We also designed the following primers for internal sequencing: Archer1F (GTAAATTACCCAATCCYAAMTCG), Archer1R (AAACATTTTGCTTTCGC), and Archer2R (TTTGTCCTGCCCTGCT). The positive products were cloned; and two to five clones from each extraction of Archerella flavum and Amphitrema wrightianum were sequenced. Sequences are deposited in GenBank with the Accession Numbers: Amphitrema wrightianum PR-1 (KC245091); Amphitrema wrightianum PR-2 (KC245092); Amphitrema wrightianum PR-2 (KC245093); Archerella flavum BC-1 (KC245094); Archerella flavum BC-2 (KC245095); Archerella flavum BC-3 (KC245096) and Archerella flavum BC-4 (KC245097).

Alignment and Phylogenetic Analysis

All SSU rRNA gene sequences were submitted to BLAST [26] in order to check their similarity with other available data in Genbank. Related sequences together with our sequences were added to a recent dataset [27] and aligned using the BioEdit software [28]. Introns, insertions and variable regions in the SSU rRNA alignment that could not be aligned unambiguously were removed from the analyses. Phylogenetic trees were reconstructed using both Maximum Likelihood and Bayesian approaches based on 800 bp alignment using some sequences of Rhizaria as outgroup.

The maximum likelihood tree was built using RAxML version 7.2.8 algorithm [29] as proposed on the Black Box portal (http://phylobench.vital-it.ch/raxml-bb/) using the GTR+Г+I model. Model parameters were estimated in RAxML over the duration of the tree search. The obtained tree was compared to the one that built by Bayesian analysis using the software MrBayes v. 3.1.2 [30]. We performed two simultaneous MCMC chains, and 500,000 generations. The generations were added until standard deviation of split frequencies fell below 0.01 according to the manual of MrBayes 3.1. For every 1,000^th generation, the tree with the best likelihood score was saved, resulting in 10,000 trees. The burn-in value was set to 25%. Trees were viewed using FigTree (a program distributed as part of the BEAST package). In addition, we performed approximately unbiased (AU) tests [31] to evaluate the likelihood of different alternative topologies to the obtained tree (see Results section).

1. Phylogenetic Position of Genera Amphitrema and Archerella, and Evolution of the Labyrinthulomycetes

This study demonstrates that the genera Archerella and Amphitrema belong to the Labyrinthulomycetes/Labyrinthulea (Stramenopiles) (Figure 2), rather than to other filose testate amoeba such as Euglyphida, Pseudodifflugia (Cercozoa), or the Foraminifera. Filose pseudopodia therefore appeared several times in eukaryotic evolution, not only within Rhizaria but also in some Opisthokonts (Nuclearia; see [34]) and Heterokonts (Leukarachnion; see [35]) and now, Amphitremida.

The labyrinthulida and thraustochytrida are characterized by the presence of a bothrosome or a ‘sagenetosome’ (sometimes also called ‘sagenogen’), an organelle that produces cytoplasmic networks (extensions of the plasma membrane) to absorb the nutrients from the surrounding environments, similar in that to fungi [36], [37]. Other genera such as Labyrinthula and Aplanochytrium also use these cytoplasmic networks for gliding [38]. The Labyrinthulomycetes are mainly osmotrophic protists. They are extremely common in marine environments, are often associated with decaying plants such as mangrove leaves [39], and less frequently parasitic [40]. The labyrinthulids and thraustochytrids exhibit a typical dimorphic life cycle with a vegetative absorptive stage and a flagellated zoosporic stage. Although genera Archerella, Amphitrema, Diplophrys and Amphifila move also by filose ectoplasmic extensions they do not possess a true bothrosome, and biflagellated stage have not yet been observed [14] (Table 1).

Our phylogenetic analysis suggests that the typical organisation with two symmetrical pseudopodial tufts found in Amphitrema, Archerella, Diplophrys and Amphifila marina might be ancestral to both Amphitremida and Amphifilidae. The AU test does not reject the existence of one clade grouping both Amphitremida and Amphifilidae and a second clade uniting the thraustochytrida together with labyrinthulida, an evolutionary pathway that appears the most parsimonious because it implies a single appearance of the bothrosome, a unique feature of thraustochytrida and labyrinthulida (Figure 2), and a simultaneous loss of pseudopodia and phagotrophy [38]. Under this evolutionary hypothesis, the bilateral symmetry of the cells would be a synapomorphy of Amphitremida+Amphifilidae, a character shared by all known members. Because osmotrophic state is not likely to be reversed back into a phagotrophic state (this would require regaining structures necessary for phagocytosis), we can hypothesize that the ancestral Labyrithulomycetes were phagotrophic and amoeboid organisms, possibly with a bilateral symmetry. In support to this interpretation, the basal-branching Schizochytrium mangrovei and Thraustochytrium striatum can shift from osmotrophic vegetative stage to phagotrophic amoeboid stage ingest through pseudopodia, if kept in culture together with bacteria, illustrating the dual nature of these organisms [41]. Diplophrys sp and Amphifila marina are both phagotrophic, and recent studies based on stable isotope ratios suggest that Archerella flavum is bacterivorous (Vincent Jassey, unpublished data), in contrast to earlier suppositions [20].

Another possible candidate for the assignment to both Amphitremida and Amphifilidae is Sorodiplophrys stercorea, an organism isolated from cow and horse dung with bilateral symmetry and filose pseudopodia, and devoid of bothrosome; interestingly, it relies entirely on osmotrophy [42].

2. Environmental Diversity of the Amphitremida

Our tree analysis including environmental clones revealed an unexpected diversity of organisms branching within Amphitremida that derived from a very wide range of environments (Figure 2). These include a freshwater biofilm clone from New Zealand, 528-O7 [43] that showed more than 98% similarity to our Archerella flavum sequences, and also, surprisingly, fourteen sequences obtained from anoxic and micro-oxic water column from the Cariaco Basin in the Caribbean Sea [44]. Their pervasive presence suggests that they are either genuine members of planktonic communities, or that they are associated to sinking debris in the water column. In addition, one pyro-tag from activated sludge F5K2Q4C04J2WC4 (GU923337) and one environmental sequence B216R from the surface of marine macro-algae (GU949602), both unpublished data from GenBank have been found to have high similarity with Archerella and Amphitrema respectively, but were not included in our analysis due to their short length. Nevertheless, these sequences further illustrate the diversity of habitats colonized by the hitherto unrecognised clade of Amphitremida.

The environmental sequence PR3_4E_52 (GQ330589) obtained by Lara et al., [27] from a eukaryotic diversity survey of the central pool in a pristine peat bog in the Swiss Jura Mountains branched together with our Amphitrema wrightianum sequences, from which it differs by three nucleotides at most. The different extractions of A. wrightianum also showed small differences in their sequences (up to two nucleotides). This diversity suggests the presence of several genotypes within the morpho-species A. wrightianum and/or the existence of several closely related taxa (cryptic species). Indeed, confusion with the very similar-looking peat bog species A. stenostoma [45] cannot be excluded. The two other described Amphitrema species (A. lemanense [46] and A. congolense [47]) do not possess endosymbiotic algae, and their phylogenetic position within the genus still needs to be determined. In contrast, our analyses did not reveal any intra-species genetic variability within Archerella flavum.

The Amphitremida represents a novel major clade within the Labyrinthulomycetes. This clade has colonised environments as divergent as peat bogs, freshwater and the oceans, ranging from nitrogen-depleted environments to eutrophic (sludge); their metabolisms vary from aerobic to anaerobic/micro-aerophilic, they can be phagotrophic or mixotrophic and have a planktonic or a benthic lifestyle. This illustrates the immense versatility of this group that certainly encompasses an even larger environmental genetic diversity than currently known. The true magnitude of this diversity will most probably be revealed by future massive environmental sequencing studies.