Comprehensive phylogenetic reconstruction of amoebozoa based on concatenated analyses of SSU-rDNA and actin genes

Abstract

Evolutionary relationships within Amoebozoa have been the subject of controversy for two reasons: 1) paucity of morphological characters in traditional surveys and 2) haphazard taxonomic sampling in modern molecular reconstructions. These along with other factors have prevented the erection of a definitive system that resolves confidently both higher and lower-level relationships. Additionally, the recent recognition that many protosteloid amoebae are in fact scattered throughout the Amoebozoa suggests that phylogenetic reconstructions have been excluding an extensive and integral group of organisms. Here we provide a comprehensive phylogenetic reconstruction based on 139 taxa using molecular information from both SSU-rDNA and actin genes. We provide molecular data for 13 of those taxa, 12 of which had not been previously characterized. We explored the dataset extensively by generating 18 alternative reconstructions that assess the effect of missing data, long-branched taxa, unstable taxa, fast evolving sites and inclusion of environmental sequences. We compared reconstructions with each other as well as against previously published phylogenies. Our analyses show that many of the morphologically established lower-level relationships (defined here as relationships roughly equivalent to Order level or below) are congruent with molecular data. However, the data are insufficient to corroborate or reject the large majority of proposed higher-level relationships (above the Order-level), with the exception of Tubulinea, Archamoebae and Myxogastrea, which are consistently recovered. Moreover, contrary to previous expectations, the inclusion of available environmental sequences does not significantly improve the Amoebozoa reconstruction. This is probably because key amoebozoan taxa are not easily amplified by environmental sequencing methodology due to high rates of molecular evolution and regular occurrence of large indels and introns. Finally, in an effort to facilitate future sampling of key amoebozoan taxa, we provide a novel methodology for genome amplification and cDNA extraction from single or a few cells, a method that is culture-independent and allows both photodocumentation and extraction of multiple genes from natural samples.

Figure 1. Morphology of the amoeboid lineages isolated for this study.

1a–c. Cryptodifflugia operculata: a) Scanning electron micrograph (SEM) of C. operculata in ventral view, showing the distinctive mucous operculum covering the aperture; b) Dorsal view of two C. operculata with a cytoplasmic connection, this state is often seen in culture; c) Differential interference contrast images (DIC) of 3 connected C. operculata individuals. Scale bars are 5 µm. 1d–f. Light microscopy images of the Arcella mitrata individual that was genome amplified to generate the sequences used in this study: d) lateral view showing the typical polygonal profile; e) top view of the same individual, focal plane at the middle of test height; f) top view of the same individual, focal plane at bottom of test height, showing the characteristic rippled apertural margin. Scale bars are 100 µm. 1g–i. Hoffman modulation contrast (HMC) images of cultured individuals of Arcella gibbosa: g) lateral view showing hemispherical profile and pseudopods; h) another individual showing the shell's ridges and depressions; i) lateral view of a third individual. Scale bars are 60 µm. 1j–l. Arcella discoides: j) HMC image of a cultured individual; k) SEM image showing the thin lateral profile; l) close-up on the apertural margin of individual in k, showing pores surrounding the aperture. Scale bars for j, k are 30 µm, for l 3 µm. 1m–n. DIC images of cultured Pyxidicula operculata: m) focal plane at middle of test height showing the nucleus and one contractile vacuole; n) focal plane at the bottom of a different individual, surrounded by bacteria on which it was feeding. Scale bars are 10 µm. 1o–r. DIC images of ‘Govecia fonbrunei’ ATCC® 50196: o) Encysted individual; p) resting individual, note the hyaline covering visible at the top margin; q) individual shape immediately after excystation; r) initial stages of locomotion. Scale bars are 10 µm. 1s–t. HMC images of Hyalosphenia papilio: s) close up on one of the individuals that was genome amplified to obtain sequences in this study, scale bar 30 µm; t) a more general view of the same individual, scale bar 50 µm. 1u–y. Images of ‘Stereomyxa ramosa’ ATCC® 50982: u,v) Phase contrast images of a cultured individual; x) protargol staining, showing the single nucleus; y) DIC image of a ‘S. ramosa’ showing the variety of pseudopods it can produce. Scale bars are 20 µm. 1z–a′. HIC images of Nebela carinata: z) a lateral profile of one of the individuals used to obtain sequences in this study, this image shows the characteristic rim around the margin of the shell; a′) same individual observed in the typical raised shell locomotive position. Scale bars are 20 µm. 1b′–e′. ‘Stygamoeba regulata’ ATCC® 50892: b′) sedentary shape; c′) beginning of movement morphology; d′) start of monopodial movement; e′) polypodial movement. Scale bars are 5 µm. 1f′–h′. Three images of isolate CHINC-5 ATCC® 50979 (misidentified as Sexangularia) showing locomotive form. The absence of a shell, among other significant characters, indicates the identification as Sexangularia is incorrect. Note the finger-like pseudopods, similar to dactylopodids. Scale bars are 10 µm. Images of ATCC® isolates were generated by Jeffrey Cole and kindly provided by Robert Molestina, director of ATCC® collections, except for images on isolate CHINC-5 ATCC® 50979 provided by O. Roger Anderson.

Figure 2. Computational pipeline implemented for phylogenetic analysis.

Grey boxes indicate a dataset, grey arrows indicate phylogenetic analyses performed on that dataset. Black arrows and boxes indicate other types of analyses performed on particular datasets, and the black dotted lines indicate the final analyses performed to obtain scores for each phylogenetic reconstruction.

Figure 3. Phylogenetic reconstruction of the Amoebozoa, based on concatenated analysis of SSU-rDNA and actin genes of 139 lineages.

This reconstruction is the best maximum likelihood tree obtained from the dataset Manual139, which we consider exhibits the optimal combination of tree indices and taxonomic coverage. Both Bayesian posterior probabilities and bootstrap supports are plotted on branches of interest. Branches without any support indication had bootstrap support of less than 70. The three well-supported higher-level groupings are shaded gray. The lower-level, morphologically consistent relationships are indicated. The novel relationships uncovered in the current study are in bold, and the suggested name for the group is shown in single quotes. Terminals in bold indicate lineages for which we are providing novel molecular information. Dashed brackets represent lower-level groups that are morphologically consistent but not recovered in this reconstruction. All branches are drawn to scale, except the branches leading to Myxomycetes, Lindbladia, Vannella CAZ6/I and Clydonnella which were trimmed to half-length for display purposes.

Figure 4. Reconstruction of actin gene family evolution in Amoebozoa, using 140 paralogs.

Triangles indicate multiple paralogs (number indicated in parenthesis), the length of triangle is equal to the length of longest branching paralog within the group.