ALOG domains: provenance of plant homeotic and developmental regulators from the DNA-binding domain of a novel class of DIRS1-type retroposons

Abstract

Members of the Arabidopsis LSH1 and Oryza G1 (ALOG) family of proteins have been shown to function as key developmental regulators in land plants. However, their precise mode of action remains unclear. Using sensitive sequence and structure analysis, we show that the ALOG domains are a distinct version of the N-terminal DNA-binding domain shared by the XerC/D-like, protelomerase, topoisomerase-IA, and Flp tyrosine recombinases. ALOG domains are distinguished by the insertion of an additional zinc ribbon into this DNA-binding domain. In particular, we show that the ALOG domain is derived from the XerC/D-like recombinases of a novel class of DIRS-1-like retroposons. Copies of this element, which have been recently inactivated, are present in several marine metazoan lineages, whereas the stramenopile Ectocarpus, retains an active copy of the same. Thus, we predict that ALOG domains help establish organ identity and differentiation by binding specific DNA sequences and acting as transcription factors or recruiters of repressive chromatin. They are also found in certain plant defense proteins, where they are predicted to function as DNA sensors. The evolutionary history of the ALOG domain represents a unique instance of a domain, otherwise exclusively found in retroelements, being recruited as a specific transcription factor in the streptophyte lineage of plants. Hence, they add to the growing evidence for derivation of DNA-binding domains of eukaryotic specific TFs from mobile and selfish elements.

Figure 1

(A) Multiple sequence alignment of the DNA-binding ALOG and catalytic tyrosine recombinase domains. Proteins are labeled by their gene names, species abbreviations and Genbank index numbers separated by underscores. Sequences are colored based on their conservation at 90% consensus. The coloring scheme, consensus abbreviations and secondary structure representation are shown in the key. Absolutely conserved residues are shaded red. For residues encompassing the tyrosine recombinase N-terminal/ALOG domain, the consensus was computed based on the conservation of the alignment positions in ALOG domain-containing proteins. Also highlighted are the DNA-contacting residues derived from crystal structures of tyrosine recombinase DBDs, and the catalytic residues of the tyrosine recombinase catalytic domain. Species abbreviations are as follows. Adig : Acropora digitifera; Alyr : Arabidopsis lyrata; Atha : Arabidopsis thaliana; BPP1 : Enterobacteria phage P1; BPlambda : Enterobacteria phage lambda; Bflo : Branchiostoma floridae; Brap : Brassica rapa; Ccin : Coprinopsis cinerea; Ddis : Dictyostelium discoideum; Ecol : Escherichia coli; Esil : Ectocarpus siliculosus; Lgig : Lottia gigantea; Nvec : Nematostella vectensis; Ogra : Oryza grandiglumis; Osat : Oryza sativa; Ppat : Physcomitrella patens; Sbic : Sorghum bicolor; Skow : Saccoglossus kowalevskii; Smoe : Selaginella moellendorffii; Spra : Spirogyra pratensis; Vcho : Vibrio cholerae; Zmay : Zea mays. (B) Cartoon representation of the N-terminal DBD of the CRE recombinase (PDB: 1CRX) in complex with DNA illustrating the position of the predicted ALOG domain zinc ribbon. Helices in the DBD that are conserved in the ALOG domain are colored red.

Figure 2

(B). Phylogenetic tree of the ALOG domain, domain architectures, and structure of the ALOG containing DIRS-1 transposon. The tree was reconstructed using an approximately maximum-likelihood method implemented in the FastTree 2.1 program (see Material and methods). Clades with boostrap values equal to or above 80% are marked with a red circle. Well-supported clades are collapsed and shown as triangles, which are colored based on their phyletic patterns (shown in the key below). The higher-order relationships should be viewed with caution due to the shortness of the alignment. Phyletic patterns of the collapsed clades are shown next to the clade name in brackets. Species abbreviations are as in Figure 1. For complete details, refer to the supplement. Also shown is the structure of the complete transposon extracted from the Ectocarpus genome. Domains in the architectures are not drawn to scale. X refers to an uncharacterized domain.