Emergence and loss of spliceosomal twin introns

Abstract

Background: In the primary transcript of nuclear genes, coding sequences-exons-usually alternate with non-coding sequences-introns. In the evolution of spliceosomal intron-exon structure, extant intron positions can be abandoned and new intron positions can be occupied. Spliceosomal twin introns ("stwintrons") are unconventional intervening sequences where a standard "internal" intron interrupts a canonical splicing motif of a second, "external" intron. The availability of genome sequences of more than a thousand species of fungi provides a unique opportunity to study spliceosomal intron evolution throughout a whole kingdom by means of molecular phylogenetics.

Results: A new stwintron was encountered in Aspergillus nidulans and Aspergillus niger. It is present across three classes of Leotiomyceta in the transcript of a well-conserved gene encoding a putative lipase (lipS). It occupies the same position as a standard intron in the orthologue gene in species of the early divergent classes of the Pezizomycetes and the Orbiliomycetes, suggesting that an internal intron has appeared within a pre-extant intron. On the other hand, the stwintron has been lost from certain taxa in Leotiomycetes and Eurotiomycetes at several occasions, most likely by a mechanism involving reverse transcription and homologous recombination. Another ancient stwintron present across whole Pezizomycotina orders-in the transcript of the bifunctional biotin biosynthesis gene bioDA-occurs at the same position as a standard intron in many species of non-Dikarya. Nevertheless, also the bioDA stwintron has disappeared from certain lineages within the taxa where it occurs, i.e., Sordariomycetes and Botryosphaeriales. Intriguingly, only the internal intron was lost from the Sordariomycetes bioDA stwintron at all but one occasion, leaving a standard intron in the same position, while where the putative lipase stwintron was lost, no intronic sequences remain.

Conclusions: Molecular phylogeny of the peptide product was used to monitor the existence and fate of a stwintron in the transcripts of two neatly defined fungal genes, encoding well conserved proteins. Both defining events-stwintron emergence and loss-can be explained with extant models for intron insertion and loss. We thus demonstrate that stwintrons can serve as model systems to study spliceosomal intron evolution.

Keywords: Aspergillus nidulans; Intron gain; Intron loss; Molecular phylogenetics; Pezizomycotina; Spliceosomal intron evolution; Spliceosomal twin introns.

Fig. 1

The [D1,2] stwintron of the putative lipase encoding gene of A. nidulans. a Depicts the determined A. nidulans lipS intron–exon structure with the exact sizes of the alternating exons and intervening sequences given. The stwintron is the second intervening sequence. b Schematically the structure of the phase-1 stwintron that splits the CCC codon of Pro189, and the two consecutive splicing events necessary to remove it. Exonic sequences are printed in capitals, grouped as the consecutive codons (Val186–Phe187–Ala188–Pro189–Asp190–Tyr191–Arg192). The internal intron is marked by the lighter grey bar; its 5′-donor-, lariat branch point domain- and 3′-acceptor sequences are printed in blue lettering. The external intron is marked by the darker grey bar; its 5′-donor-, lariat branch point domain- and 3′-acceptor sequences are printed in red lettering. c The existence of the splicing intermediate from which the internal intron has been removed by the first excision, has been confirmed experimentally (GenBank MF612152). The newly formed donor splice site sequence of the retained external intron, (5′-gugagu), is underlined

Fig. 2

The [D1,2] stwintron of the putative lipase encoding gene of A. niger ATCC 1015. a Depicts the determined A. niger lipS intron–exon structure with the exact sizes of the exons and intervening sequences given: The stwintron is the second intervening sequence. b Schematically the structure of the stwintron that splits the CCU codon of Pro184, and the two consecutive splicing events necessary to remove it. Exonic sequences (in capitals) are grouped as the consecutive codons (Val181–Phe182–Ala183–Pro184–Lys185–Tyr186–Arg187). The internal intron is marked by the lighter grey bar; its 5′-donor-, lariat branch point domain- and 3′-acceptor sequences are printed in blue lettering. The external intron is marked by the darker grey bar; its 5′-donor-, lariat branch point domain- and 3′-acceptor sequences are printed in red lettering. c The existence of the splicing intermediate from which the internal intron has been removed by the first excision, has been confirmed experimentally (GenBank MF612153). The newly formed donor splice site sequence of the retained external intron, (5′-gugagu), is underlined

Fig. 3

Maximum likelihood phylogeny of the LipS orthologue in the Ascomycota: Emergence of the [D1,2] stwintron. Taphrinomycotina proteins constitute the outgroup. Branch statistics are given as Approximate Likelihood Ratio Test values (0–1) at each node. The branches are color coded to distinguish the classes of Pezizomycotina; olive green: Pezizomycetes, auburn: Orbiliomycetes, blue: Sordariomycetes, green: Eurotiomycetes (NB. Including A. nidulans), red: Dothideomycetes, orange: Leotiomycetes, violet: Lecanoromycetes. The latter five classes belong to the Leotiomyceta super class. For the sake of simplicity, the clades for the four well-represented classes—Sordariomycetes, Eurotiomycetes, Dothideomycetes and Leotiomycetes—are collapsed. Classes in which the stwintron occurs are underlined. Species of Pezizomycetes and Orbiliomycetes, in which the stwintron position (second intron position in Aspergillus) is occupied by a standard intron in the orthologue gene, have their names printed in light blue. The two defining events in the formation of the stwintron are indicated by the yellow arrows at the left. The data do not allow to determine whether the stwintron emerged in Leotiomyceta before or after the divergence of the Sordariomycetes

Fig. 4

Maximum likelihood phylogeny of LipS: Instances of stwintron loss in Eurotiomycetes. A subtree of the phylogenetic analysis depicted in Fig. 3, is shown in detail to highlight the loss of the stwintron from taxa of Eurotiomycetes. Class-specific color coding is the same as in Fig. 3. Species that have lost the stwintron from their lipS gene have their name printed in red lettering. For the sake of simplicity, we have collapsed groups of related fungi that behave identically with respect to stwintron presence/absence. All the species in the collapsed clades for the Chaetothyriales order and the Penicillium genus have no intronic sequences at the stwintron position (second intron position in Aspergillus) and the taxon names are therefore marked with red boxes to highlight stwintron loss. Independent events of stwintron loss are also indicated by red triangles on the directly preceding branches (NB. The position of the triangles does not correspond with the exact time point at which stwintron loss has taken place)

Fig. 5

Maximum likelihood phylogeny of LipS: Instances of stwintron loss in Leotiomycetes. A subtree of the phylogenetic analysis depicted in Fig. 3, is shown in detail to highlight the loss of the stwintron from taxa of Leotiomycetes. Class-specific color coding is the same as in Fig. 3. Species that have lost the stwintron from their lipS gene have their name printed in red lettering. Independent events of stwintron loss are also indicated by red triangles on the directly preceding branches (NB. The position of the triangles does not correspond with the exact time point at which stwintron loss has taken place)

Fig. 6

Maximum likelihood phylogeny of the BioDA protein in the Ascomycota: Emergence of the [D2,3] stwintron. Taphrinomycotina proteins constitute the outgroup. Branch statistics are given as Approximate Likelihood Ratio Test values (0–1) at each node. The branches are color coded to distinguish classes of Pezizomycotina, as described in the legend to Fig. 3. Some fungal taxa are collapsed, to the level of whole classes for most of the Pezizomycotina. Taxa in which the stwintron occurs—Sordariomycetes and Botryosphaeriales—are underlined. The two groups of Sordariomycetes that stand out for the absence of intronic sequences at the bioDA stwintron position—members of two families of Diaporthales and five species of the order of the Ophiostomatales, respectively—are cartooned (blue) rather than collapsed. The branch where the stwintron has emerged at the position of an ancient intron is indicated by the upper yellow arrow. The latter “host” intron occurs in many species of non-Dikarya (see “Results” section): To indicate its kingdom-wide existence, the lower yellow arrow points at the origin of the tree. Note that a Maximum Likelihood phylogeny based on an alignment calculated using more stringent parameters (similarity matrix BLOSUM62) suggested that the Botryosphaeriales clade is paraphyletic to the (main) Sordariomycetes clade

Fig. 7

Maximum likelihood phylogeny of the BioDA protein: Instances of stwintron loss. A subtree of the phylogenetic analysis depicted in Fig. 6, is shown in detail to highlight the loss of the stwintron from taxa of Sordariomycetes and Botryosphaeriales. Class-specific color coding is the same as in Fig. 3. Species that have lost the internal intron from the stwintron in their bioDA gene (retaining a standard intron at the stwintron position) have their name printed in green lettering. For the sake of simplicity, we have collapsed groups of related fungi that behave the same with respect to the stwintron. Independent events of internal intron loss are also indicated by the green triangles on the directly preceding branches. (NB. The position of the triangles does not correspond with the exact time point at which intron loss has taken place). The complete absence of intronic sequences at the stwintron position from the genes encoding the BioDA proteins in the separate Diaporthales clade and in two of the Botryosphaeriales (7 proteins in red lettering) may have occurred either in one or in two consecutive events of intron loss. Complete stwintron loss was indicated by a red triangle (as in the legend to Fig. 4). We annotated the tree for the two-step process in the case of P. capitalensis, first losing the internal intron in an ancestor shared with P.citricarpa (green triangle). The subsequent loss of the standard intron that remained after the first event (red triangle) resulted in the complete absence of intronic sequences from P. capitalensis bioDA