Evolutionary, structural and functional analysis of the caleosin/peroxygenase gene family in the Fungi

Abstract

Background: Caleosin/peroxygenases, CLO/PXG, (designated PF05042 in Pfam) are a group of genes/proteins with anomalous distributions in eukaryotic taxa. We have previously characterised CLO/PXGs in the Viridiplantae. The aim of this study was to investigate the evolution and functions of the CLO/PXGs in the Fungi and other non-plant clades and to elucidate the overall origin of this gene family.

Results: CLO/PXG-like genes are distributed across the full range of fungal groups from the basal clades, Cryptomycota and Microsporidia, to the largest and most complex Dikarya species. However, the genes were only present in 243 out of 844 analysed fungal genomes. CLO/PXG-like genes have been retained in many pathogenic or parasitic fungi that have undergone considerable genomic and structural simplification, indicating that they have important functions in these species. Structural and functional analyses demonstrate that CLO/PXGs are multifunctional proteins closely related to similar proteins found in all major taxa of the Chlorophyte Division of the Viridiplantae. Transcriptome and physiological data show that fungal CLO/PXG-like genes have complex patterns of developmental and tissue-specific expression and are upregulated in response to a range of biotic and abiotic stresses as well as participating in key metabolic and developmental processes such as lipid metabolism, signalling, reproduction and pathogenesis. Biochemical data also reveal that the Aspergillus flavus CLO/PXG has specific functions in sporulation and aflatoxin production as well as playing roles in lipid droplet function.

Conclusions: In contrast to plants, CLO/PXGs only occur in about 30% of sequenced fungal genomes but are present in all major taxa. Fungal CLO/PXGs have similar but not identical roles to those in plants, including stress-related oxylipin signalling, lipid metabolism, reproduction and pathogenesis. While the presence of CLO/PXG orthologs in all plant genomes sequenced to date would suggest that they have core housekeeping functions in plants, the selective loss of CLO/PXGs in many fungal genomes suggests more restricted functions in fungi as accessory genes useful in particular environments or niches. We suggest an ancient origin of CLO/PXG-like genes in the 'last eukaryotic common ancestor' (LECA) and their subsequent loss in ancestors of the Metazoa, after the latter had diverged from the ancestral fungal lineage.

Keywords: Caleosin; Evolution; Fungi; Lipid droplets; Oxylipins; Peroxygenase; Stress responses; Viridiplantae.

Fig. 1

Motif analysis of 40 selected fungal CLO/PXG sequences from each of the major phyla. a six motif logo generated from the consensus amino acid sequences; b distribution of the motifs in the 40 selected fungal species

Fig. 2

CLO/PXG protein sequence alignments from representative fungal species. a 40 selected fungal species showing the major structural and functional domains; b 23 sequences from non-Dikarya species. The major conserved domains are shown in boxes: N-terminal H-domain, Ca²⁺ binding EF hand, lipid-binding domain, haem binding and kinase domain, C-terminal phosphorylation domain. The two near-invariant haem-coordinating histidine residues are shown with a red star

Fig. 3

Predicted secondary structures and TM of 40 selected fungal CLO/PXG proteins. The red right hand arrow showing the transmembrane domain for each sequence. For secondary structure, the annotation goes as: pink tube shows Alpha helix, the yellow arrow shows beta strand, blue arrow shows turn, grey coil shows the coil of each sequence

Fig. 4

Predicted gene structure of 40 representative fungal CLO/PXGs. The prediction shows the locations of introns (grey) and exons (black). Note that the gene lengths are quite variable so for clarity they have all been scaled to the same lengths here

Fig. 5

Roles of oxylipins in mediating downregulation of AfPXG gene expression and decreased aflatoxicogenicity in A. flavus. a 7-day old fungal growth on PDA-plates in the presence or absence of exogenous oxylipins at concentrations of 50 and 100 μM, referred to as to oxylipin50 and oxylipin100, respectively. b Measurements of conidia number and fungal mycelium dry weight for each treatment compared with controls. c Transcript levels of AfPXG genes, evaluated by RT-qPCR, and peroxygenase activity of AfPXG, measured by the hydroxylation of aniline at 310 nm. d Micrographs of LDs viewed at 40× magnification immediately after preparation. Bar represents 5 μm. e Sections of TLC-plate showing the blue-fluorescent spot under UV corresponding to AFB1 extracted from oxylipin-treated fungi compared to a control. f Quantitative data for AFB1 production estimated by UV-detector HPLC. g Relative quantification RQ ₌ 2^{(−∆∆CT)} of AF-biosynthesis cluster genes in oxylipin-treated fungi compared to controls. The colour scale (white-red-black) indicates relative changes of transcripts of 1, − 20 and − 40 fold, respectively where the expression level for each gene in controls was defined as 1

Fig. 6

Alignment of CLO/PXG sequences from three non-fungal opisthokont species with ten selected fungal and ten plant sequences. The three non-fungal opisthokont are: Planoprotostelium fungivorum (PfuCLO1), Capsaspora owczarzaki (CoCLO1 & 2), and Panagrolaimus spp (PsppCLO1–5). The ten Viridiplantae species are: Chlorella variabilis (CvaCLO1), Volvox carteri (VcCLO1), Klebsormidium nitens(KnCLO1), Marchantia polymorpha(MpCLO1), Cycus revolute (CreCLO1), Amborella trichopoda (AtrCLO1), Elaeis guineensis (EgCLO1), Phoenix dactylifera (PdCLO1), Jatropha curcas (JcCLO1) and Arabidopsis thaliana (AtCLO1). The ten fungal species are: Agaricus bisporus (AbbCLO1), Coprinopsis cinerea (CciCLO1), Aspergillus flavus (AflCLO1), Aspergillus oryzae (AorCLO1), Beauveria bassiana (BbCLO1), Rhizophagus irregularis (RiCLO1), Allomyces macrogynus (AmaCLO1), Spizellomyces punctatus (SpCLO1), Mitosporidium daphnia (MdCLO1) and Rozella allomycis (RaCLO1)

Fig. 7

Representative Maximum Likelihood phylogeny for 199 plant and fungal CLO/PXG proteins. The optimum model of protein substitution was found to be LG + G. Bootstrap resampling (100 iterations) was undertaken and is shown on internal nodes. There are several strongly supported clades but support values inferring sister group relationships between these are extremely low. Species names are coloured relative to their taxonomy. The presence of MEME predicted motifs are shown for individual proteins