Phylogenomic analysis of Emiliania huxleyi provides evidence for haptophyte–stramenopile association and a chimeric haptophyte nuclear genome

Abstract

Emiliania huxleyi is a haptophyte alga of uncertain phylogenetic affinity containing a secondarily derived, chlorophyll c containing plastid. We sought to characterize its relationships with other taxa by quantifying the bipartitions in which it was included from a group of single protein phylogenetic trees in a way that allowed for variation in taxonomic content and accounted for paralogous sequences. The largest number of sequences supported a phylogenetic relationship of E. huxleyi with the stramenopiles, in particular Aureococcus anophagefferens. Far fewer nuclear sequences gave strong support to the placement of this coccolithophorid with the cryptophyte, Guillardia theta. The majority of the sequences that did support this relationship did not have plastid related functions. These results suggest that the haptophytes may be more closely allied with the heterokonts than with the cryptophytes. Another small set of genes associated E. huxleyi with the Viridiplantae with high support. While these genes could have been acquired with a plastid, the lack of plastid related functions among the proteins for which they code and the lack of other organisms with chlorophyll c containing plastids within these bipartitions suggests other explanations may be possible. This study also identified several genes that may have been transferred from the haptophyte lineage to the dinoflagellates Karenia brevis and Karlodinium veneficum as a result of their haptophyte derived plastid, including some with non-photosynthetic functions.

Introduction

Emiliania huxleyi belongs to the Haptophyceae (= Prymnesiophyceae), a group of chlorophyll a and c containing photosynthetic eukaryotes characterized by a microtubule-based cellular appendage called a haptonema used in feeding and perhaps in other ways. Some prymnesiophytes (including E. huxleyi) lack a haptonema for some or all of their life cycle, but it is a very distinctive organelle and there is little doubt that these organisms constitute a monophyletic group. The evolutionary affinities of haptophytes are poorly understood, but they appear to be at least loosely allied with other chlorophyll c containing organisms. Some haptophytes synthesize CaCO₃ scales called coccoliths with which they coat themselves, earning these single-celled organisms the name coccolithophorids (Andersen, 2004, Graham and Wilcox, 2000). Prymnesiophytes are among the three most abundant groups of eukaryotic phytoplankton in marine ecosystems, the other major lineages being diatoms and dinoflagellates (Falkowski et al., 2004). Haptophytes are important in the carbon cycle because of their photosynthetic activities (Falkowski et al., 2004, Graham and Wilcox, 2000) and their production of coccoliths, which produces CO₂ but also increases the sinking of carbon out of the mixed layer (Vanderwal et al., 1995). Coccoliths make up a large portion of cretaceous sediments and large geologic structures such as the White Cliffs of Dover (Mitchell et al., 1997). Because they are easily cultured, they have been used as models of CaCO₃ biomineralization (Iglesias-Rodriguez et al., 2008, Marsh, 2003, Riebesell et al., 2000). This is of particular concern because of the potential impact of rising CO₂ levels and climate change on CaCO₃ mineralizing organisms including reef-building corals. Haptophytes also are involved in the sulfur cycle as sources of dimethylsulphoniopropionate (DMSP), a major component of atmospheric sulfur (Sunda et al., 2002). DMSP is, in turn, converted to dimethyl sulfide (DMS), which promotes cloud condensation thus leading to an increase in albedo, which has the effect of reducing the warming effects of the sun (Charlson et al., 1987).

Haptophytes have historically been grouped with chrysophycean heterokonts because of their possession of two flagella and their golden-brown color (Bourrelly, 1965, Pascher, 1914). Hibberd (1976) separated them from the chrysophytes because of the stronger ultrastructural similarities true chrysophytes share with other heterokonts relative to the haptophytes. The main feature that unifies the heterokonts (= stramenopiles) and distinguishes them from haptophytes is the presence in heterokonts of mastigonemes on one flagellum, typically the anterior flagellum, with a naked trailing flagellum. By contrast, haptophytes characteristically have two anterior naked flagella of equal length, as well as a haptonema in some cases (Andersen, 2004, Graham and Wilcox, 2000). Cavalier-Smith (1981) grouped the haptophytes and heterokonts together into the “Chromophyta” because of the similarities between haptophytes and heterokonts including the presence of chlorophylls a and c, plastids surrounded by four membranes with thylakoids in stacks of three, and storage of chrysolaminarin and including the heterotrophic heterokonts because of the flagellar mastigonemes they share with flagellate plastidic heterokonts. Later, he grouped the alveolates and cryptophytes with the chromists unifying all chlorophyll a and c containing organisms into the Chromalveolata (Cavalier-Smith, 1999). The haptophytes were eventually separated from the heterokonts by several studies suggesting that their closest relatives were the cryptophytes and that the rhizaria occupied a position between them and the heterokonts with their sister taxon the alveolates (Burki et al., 2007, Hackett et al., 2007, Patron et al., 2007). The stramenopile, alveolate, and rhizarian group became known as the SAR. The sister relationship between the haptophytes and cryptophytes was later called into question by Burki et al., 2012a, Burki et al., 2012b. While it is generally believed that chlorophyll c containing plastids are descended from engulfed rhodophytes, the inclusion of so many organisms lacking plastids has lead some to point out that multiple acquisitions and/or losses would be required within the chromalveolate/SAR group (Archibald, 2009, Sanchez-Puerta and Delwiche, 2008). Other research has indicated that there may be a significant green algal signal separate from the red algal signal in the genomes of heterokonts and haptophytes implying that they descended from a green alga or they had a green algal plastid before acquiring the red one (Moustafa et al., 2009) although the source and extent of this gene infusion has been questioned by some (Burki et al., 2012a, Woehle et al., 2011). However, several studies have found evidence for at least some EGT (endosymbiotic gene transfer) in several chromalveolate organisms indicating that it has been a factor in the evolution of this diverse group of organisms (Burki et al., 2014, Burki et al., 2012a, Dorrell and Smith, 2011, Moustafa et al., 2009, Woehle et al., 2011).

Despite the proposal of the “chromalveolate hypothesis” more than a decade ago (Cavalier-Smith, 1999), the placement of organisms within the group is still controversial as is the number of independent plastid acquisitions and the extent of tertiary plastid gains (Archibald, 2009). Several studies have placed E. huxleyi and the haptophytes as sister to the cryptophytes (Burki et al., 2007, Hackett et al., 2007, Patron et al., 2007) but a more recent study separated the haptophytes from the cryptophytes, placing the haptophytes closer to the SAR (Burki et al., 2012b). The phylogenetic placement of these organisms is made more difficult by the opportunity for endosymbiotic gene transfer (EGT) within these lineage(s) which implies that discordant gene trees are to be expected and may represent biological reality in some cases. Conventional phylogenomic approaches like supermatrix and supertree methods average the signal from multiple genes thus missing important biological phenomena such as horizontal gene transfer and paralogy (Edwards, 2009). Furthermore, adding more sequence data to concatenated analyses can lead to incorrect conclusions with inappropriately high confidence levels (Delsuc et al., 2005). Thus, phylogenetic methods used to study these organisms should treat gene trees individually to account for the possibility of EGT. This study will leverage an increasing quantity of genomic sequence data from chromalveolate taxa including: Emiliania huxleyi (Read et al., 2013) the cryptophyte Guillardia theta (Curtis et al., 2012) and several heterokonts including but not limited to Aureococcus anophagefferens (Gobler et al., 2011), Phaeodactylum tricornutum (Bowler et al., 2008), and Thalassiosira pseudonana (Armbrust et al., 2004) to do a phylogenomic study centered on E. huxleyi in a way that accounts for the possibility of EGT and paralogy. Our approach, which has similarities to those used by Burki et al. (2014, 2012a), involves quantifying the occurrence of E. huxleyi with other organisms from a group of single protein phylogenetic trees in a way that allows for paralogy as well as variation in taxon content.