The evolution of the Puf superfamily of proteins across the tree of eukaryotes

Abstract

Background: Eukaryotic gene expression is controlled by a number of RNA-binding proteins (RBP), such as the proteins from the Puf (Pumilio and FBF) superfamily (PufSF). These proteins bind to RNA via multiple Puf repeat domains, each of which specifically recognizes a single RNA base. Recently, three diversified PufSF proteins have been described in model organisms, each of which is responsible for the maturation of ribosomal RNA or the translational regulation of mRNAs; however, less is known about the role of these proteins across eukaryotic diversity.

Results: Here, we investigated the distribution and function of PufSF RBPs in the tree of eukaryotes. We determined that the following PufSF proteins are universally conserved across eukaryotes and can be broadly classified into three groups: (i) Nop9 orthologues, which participate in the nucleolar processing of immature 18S rRNA; (ii) 'classical' Pufs, which control the translation of mRNA; and (iii) PUM3 orthologues, which are involved in the maturation of 7S rRNA. In nearly all eukaryotes, the rRNA maturation proteins, Nop9 and PUM3, are retained as a single copy, while mRNA effectors ('classical' Pufs) underwent multiple lineage-specific expansions. We propose that the variation in number of 'classical' Pufs relates to the size of the transcriptome and thus the potential mRNA targets. We further distinguished full set of PufSF proteins in divergent metamonad Giardia intestinalis and initiated their cellular and biochemical characterization.

Conclusions: Our data suggest that the last eukaryotic common ancestor (LECA) already contained all three types of PufSF proteins and that 'classical' Pufs then underwent lineage-specific expansions.

Keywords: Giardia intestinalis; LECA; Puf superfamily proteins; RNA processing; RNA-binding protein.

Fig. 1

Clustering analysis of Puf family proteins. PufSF proteins were analysed by CLANS. A total of 4469 proteins formed three major clusters of proteins corresponding to Nop9 homologues (Nop9 cluster), ‘classical’ Puf proteins (Puf cluster), and PUM3 orthologues (PUM3 cluster). The protein sequences were colour-coded according to the taxonomic affiliation to the eukaryotic supergroup. The details of each large cluster show mostly lineage-specific subclusters, except two large subclusters of Puf proteins. The diagrams above the cluster details depict the general domain arrangement of PufSF proteins; PF00806 corresponds to Puf RNA-binding repeat. The x and y axes represent relative positions of the protein sequence in 2D CLANS

Fig. 2

PufSF proteins in major groups of eukaryotes. Puf family proteins identified in all UniProt reference proteomes were grouped according to taxonomic affiliation as specified in Additional file 2: Table S2 and showed in circle packing plot. Each dot corresponds to a genome of particular eukaryotic species, and the size of the dot represents a number of Puf family proteins in the predicted proteome and the colour-codes to the taxonomic affiliation to the eukaryotic supergroup. Grey circles depict taxonomic groups

Fig. 3

Three types of PufSF proteins in major groups of eukaryotes. a The number of Nop9, Puf, and PUM3 orthologues was identified for each proteome in the dataset, and the values were averaged for the taxonomic group of eukaryotes. The error bars correspond to the standard deviation of variance of values within the particular taxonomic group. Error bars for Streptophyta and Ciliophora were cut for better visualization. b The number of Pufs normalized with respect to the total number of the protein coding; colour-codes correspond to the taxonomic affiliation to the eukaryotic supergroups

Fig. 4

Phylogenetic analysis of PufSF proteins. Maximum likelihood phylogenetic inference of PufSF proteins with sequences from H. sapiens, C. elegans, S. cerevisiae, and A. thaliana is shown as triangles with the indicated colours. For visualization purposes, support values were removed. Full phylogenies of the PUM3, Nop9, and Puf proteins can be found in Additional file 3: Fig. S1

Fig. 5

Domain structure of G. intestinalis PufSF proteins. a The Pumilio homology domain containing Puf repeats was predicted using HHPred against Pfam database and denoted as oval for each of the Giardia intestinalis proteins (shades of brown). The numbers in the ovals represent the position of the domain within the protein (black lines). The expectation value (E value) of the domain detection is shown in brackets. The fruit fly, human, and fungal orthologues are shown for comparison. Gi, Giardia intestinalis; Dm, Drosophila melanogaster; Hs, Homo sapiens; Sc, Saccharomyces cerevisiae. b Protein sequence alignment of G. intestinalis Pufs with selected proteins from S. cerevisiae, D. melanogaster, and H. sapiens. Open red rectangles highlight TRMs. Light and dark grey rectangles depict Puf repeats. Fraction of identical amino acids at the particular position is coloured: dark blue > 80%, blue > 60%, light blue > 40%, white < 40%

Fig. 6

Cellular localization of G. intestinalis Puf and Nop9 proteins. a The scores obtained by DeepLoc prediction indicate cytosolic localization of all but GiPUM3 protein, which is predicted as nuclear protein. Colouring gradient represents the values from 0 (red) to 1 (green). Cyt, cytosol; Nuc, nucleus; CM, cytoplasmic membrane; ER, endoplasmic reticulum; Mit, mitochondrion; Extra, extracellular. b Western blot analysis of G. intestinalis expressing BAP-tagged PufSF proteins (one of at least three independent cell experiments is show). SDS-PAGE and immunoblots show total lysate (Lys), cytosolic (Cyt), and high-speed pellet (HSP) fraction. c Immunofluorescence analysis of the same cell lines shows cellular localization of the proteins. PufSF proteins in green, Sec20-endomembrane system marker in red. DIC, Differential Interference Contrast. d Detailed imaging of GiPUM3 by confocal and 2D STED (GiPUM3 in green, DNA in blue). 3D STED of GiPUM3 with the orthogonal projections (GiPUM3 in red)

Fig. 7

Predicted binding motifs of G.intestinalis PufSF proteins. a The tripartite recognition motifs (TRMs) of each Puf repeats of G.intestinalis Pufs were predicted, and the resulting sequence was used to search the conceptual transcriptome. The number of putative mRNA targets, which contain the motif in the 3′-UTR, is shown in bold. Asterisk denotes the same putative mRNA recognized by two different Pufs. b Sequence alignment of 18S rRNA shows the conservation of the sequence recognized by Nop9 as it was experimentally identified for S. cerevisae Nop9. Fraction of identical nucleotides at the particular position is coloured: dark blue > 80%, blue > 60%, light blue > 40%, white < 40%