Fig 1.
Arrangement of the heme biosynthetic pathway in algae with complex plastids.
Inferred origins of enzymes are represented by colored boxes with flags where applicable. Localizations of B. natans and G. theta enzymes were predicted by SignalP and TargetP (see Material and Methods). Dashed arrows indicate a possible dual localization of UROS in both the cytosol and the plastid of B. natans. Only key metabolites are shown for clarity, for example the starting substrates for mitochondrial-cytosolic C4 (succinyl-CoA and glycine) and plastid C5 pathways (glutamyl-tRNAGlu). Parts of the pathway identical to K. foliaceum are not depicted in the D. baltica and G. foliaceum scheme. Schematic representation of organelles: N, N1 –nucleus of the host; N2 –nucleus of the endosymbiont diatom; n—nucleomorph of the endosymbiont; mt—mitochondrion. Enzymes: ALAS—delta-aminolevulinic acid synthase; GTR—glutamate-tRNA reductase; GSA—glutamate-1-semialdehyde 2,1-aminotransferase; ALAD—aminolevulinic acid dehydratase; PBGD—porphobilinogen deaminase; UROS—uroporphyrinogen-III synthase; UROD—uroporphyrinogen decarboxylase; CPOX—coproporphyrinogen oxidase; PPOX—protoporphyrinogen oxidase; FeCH—ferrochelatase. A typical pathway in a primary heterotroph and a primary autotroph are shown for comparison (Kořený and Oborník 2011).
Fig 2.
Bayesian phylogenetic tree as inferred from ALAD amino acid sequences.
Taxa of interest in this study are highlighted by colored bars: blue for dinotoms, green for Lepidodinium chlorophorum, ochre for Bigelowiella natans and red for Guillardia theta. The tree shows red algal origin for B. natans and G. theta enzymes. For L. chlorophorum, we suggest a gene duplication / loss of paralogs scenario (see text); despite branching as sister to green algae, other dinoflagellates contained in the same clade do not possess a green algal plastid. Numbers near branches indicate Bayesian posterior probabilities followed by the bootstrap of respective clades from the likelihood analysis. Only support values greater than 0.85 (Bayesian) and 50 (likelihood) are shown. dt—different topology in the ML tree, see S2 Fig; a dash indicates an unsupported topology. An asterisk marks inferred bacterial contamination in G. foliaceum data.
Fig 3.
Bayesian phylogenetic tree as inferred from FeCH amino acid sequences.
Taxa of interest in this study are highlighted by colored bars: blue for dinotoms, green for Lepidodinium chlorophorum, ochre for Bigelowiella natans and red for Guillardia theta. We document two orthologs, one of unresolved cyanobacterial origin and the other of eukaryotic origin, for B. natans and G. theta enzymes. The L. chlorophorum sequence branches together with other dinoflagellates, suggesting its origin lies in the peridinin plastid repertoire. Numbers near branches indicate Bayesian posterior probabilities followed by the bootstrap of respective clades from the likelihood analysis. Only support values greater than 0.85 (Bayesian) and 50 (likelihood) are shown. dt—different topology in the ML tree, see S2 Fig; a dash indicates unsupported topology. An asterisk marks inferred bacterial contamination in G. foliaceum and Karenia brevis data.
Fig 4.
Bayesian phylogenetic tree as inferred from GSA-AT amino acid sequences.
Taxa of interest in this study are highlighted by colored bars, blue for dinotoms, green for Lepidodinium chlorophorum, ochre for Bigelowiella natans and red for Guillardia theta. Numbers near branches indicate Bayesian posterior probabilities followed by the bootstrap of respective clades from the maximum likelihood (ML) analysis. Only support values greater than 0.85 (Bayesian) and 50 (ML) are shown. dt—different topology in the ML tree, see S2 Fig; a dash indicates unsupported topology. The tree demonstrates the cyanobacterial origin of canonical GSA-AT, while the non-canonical GSA-AT originates in proteobacteria. Schematics of Karenia brevis transcripts and respective proteins are shown for complete representatives of canonical and non-canonical GSA-AT. The presence of a spliced-leader sequence at the 5’ end suggests nuclear encoding and transcription of these genes. An N-terminal presequence of the resulting protein putatively targets both enzymes into the plastid. The canonical and non-canonical enzymes share motifs for pyridoxal 5’-phosphate binding and a catalytical residue. UTR—untranslated region; ORF—open reading frame; TM—transmembrane domain; SP—signal peptide; SP-TM—signal peptide predicted by the SignalP-TM networks; yellow bar—Panther Class III aminotransferase / glutamate-1-semialdehyde 2,1-aminomutase hit; violet bar—Pfam Class III aminotransferase hit; grey bar—PROSITE Class III aminotransferase hit; numbers represent scale in nt or aa.
Fig 5.
A simplified scheme of the evolution of the tetrapyrrole biosynthesis pathway with an emphasis on the models from this study (black bars).
Primary endosymbiosis (PE) gave rise to the Archaeplastida comprising red algae, green algae and glaucophytes. Following the divergence of main eukaryotic lineages, secondary (SE) or tertiary endosymbiosis (TE) events equipped the ancestors of CASH taxa (cryptophytes, alveolates, stramenopiles and haptophytes) with photosynthetic capabilities. Contradictory evidence has been debated over the last years as for the history of CASH plastid acquisitions (e.g. [5,8,34]). A plastid-early scenario (the chromalveolate hypothesis) posits that all CASH taxa are monophyletic and the CASH plastid was vertically transferred (dashed red line) and lost in extant non-photosynthetic descendants (such as ciliates and most rhizarians). Plastid-late scenarios require multiple lateral acquisitions of the CASH plastid (question marks) but better reflect some current phylogenomic analyses of the plastid recipients (e.g. [18,69]). Loss of photosynthesis/plastids have been documented in many sister lineages, such as oomycetes or apicomplexans, however these are in line with plastid-late scenarios as well. A cryptic SE with a CASH alga or numerous HGT (red arrows) events are inferred before the divergence of extant chlorarachniophytes (this work, [78]), which was masked by the acquisition of the current green algal endosymbiont. A similar situation in L. chlorophorum led to the peridinin plastid replacement with a green plastid, however the majority of red-related heme pathway enzymes remained functionally conserved in the successor plastid. The loss of the heterotrophic pathway possibly occurred several times independently in the stramenopile and dinoflagellate lineages, as Perkinsus marinus, sister to dinoflagellates, still contains a functional mitochondrial-cytosolic pathway [8].