Journal of Molecular Biology
Volume 382, Issue 3, 10 October 2008, Pages 601-609
Journal home page for Journal of Molecular Biology

Beyond DNA: RNA Editing and Steps Toward Alu Exonization in Primates

https://doi.org/10.1016/j.jmb.2008.07.014Get rights and content

Abstract

The exaptation of transposed elements into protein-coding domains by a process called exonization is one important evolutionary pathway for generating novel variant functions of gene products. Adenosine-to-inosine (A-to-I) modification is a recently discovered, RNA-editing-mediated mechanism that contributes to the exonization of previously unprocessed mRNA introns. In the human nuclear prelamin A recognition factor gene transcript, the alternatively spliced exon 8 results from an A-to-I editing-generated 3′ splice site located within an intronic Alu short interspersed element. Sequence comparisons of representatives of all primate infraorders revealed the critical evolutionary steps leading to this editing-mediated exonization. The source of exon 8 was seeded within the primary transcript about 58–40 million years ago by the head-to-head insertions of two primate-specific Alu short interspersed elements in the common ancestor of anthropoids. The latent protein-coding potential was realized 34–52 million years later in a common ancestor of gorilla, chimpanzee, and human as a result of numerous changes at the RNA and DNA level. Comparisons of 426 processed mRNA clones from various primate species with their genomic sequences identified seven different RNA-editing-mediated alternative splice variants. In total, 30 A-to-I editing sites were identified. The gorilla, chimpanzee, and human nuclear prelamin A recognition factor genes exemplify the versatile interplay of pre- and posttranscriptional modifications leading to novel genetic potential.

Introduction

Evolution promotes new genetic function by at least two different ways: first, by triggering formation of new genes via various duplication events and, second, by including new exons in existing genes. New exons arise, for example, by exonization of intronic sequences, and often are composed, usually in part, of transposed elements. Such elements invade predominantly non-protein-coding regions and are either fortuitously exonized soon after integration or millions of years thereafter by acquisition of critical changes.1, 2 Exonized sequences, usually non-protein coding in their prior existence, contribute to protein-coding domains predominantly by alternative1, 3 and, far less frequently, by constitutive splicing.4, 5 In primates, predominantly Alu short interspersed elements are recruited as novel exons.1, 6 Alu retroposons represent a 7SL RNA-derived primate-specific repeat family.7 They distributed during different temporal waves of primate evolution, leaving behind more than 1.1 million copies in the human genome.5 The major waves of primate Alu insertion activity consist of the old AluJ subfamily [including AluJo and AluJb; about 81 million years ago (MYA)], the middle-aged AluS subfamily (including AluS, AluSx, AluSg, AluSq, AluSp, AluSc, and AluSb; about 48–19 MYA), and the young and still active AluY wave (including AluY, AluYa5, AluYa8, and AluYb8; starting about 6 MYA and continuing to the present).8

Recently, it was demonstrated that certain Alu exonizations might depend particularly on adenosine-to-inosine (A-to-I) RNA editing,6 whereby an adenosine is substituted by an inosine that is then recognized as a guanosine.9 A-to-I RNA editing can generate, delete, or alter the meaning of triplet codons as well as permit new splice sites.10, 11

The editing reaction is catalyzed by members of the enzyme family known as adenosine deaminases that act on RNA (ADARs).12 ADARs are double-stranded RNA (dsRNA)-specific enzymes that generally act on duplexes longer than 30 base pairs.13, 14 They are composed of a domain structure that includes variable numbers of dsRNA binding motifs. Two ADAR enzymes (ADAR1 and 2), varying in their number of dsRNA binding motifs and their editing specificities, have been identified in mammals.15, 16 ADAR2 preferentially edits at a single nucleotide or at few nucleotides, converting the arginine codon (AGA) to glycine (IGA = GGA), whereas ADAR1 edits numerous adenosines in RNA duplexes in a more promiscuous way.17, 18 However, the immediate structural surroundings of the edited site appears to be more important than the sequence motif itself.6, 18

Human Alu sequences appear to be preferred targets for A-to-I editing, a phenomenon that may be explained by their high incidence (e.g., 90% of all known human genes contain Alu insertions that, in case they are in tandem inverse orientation, can form intermolecular dsRNA structures that are subject to ADAR editing).13 Recently, Lev-Maor et al. described the A-to-I RNA editing and resultant alternative splicing of one such Alu element that led to the birth of exon 8 of the human nuclear prelamin A recognition factor (NARF; Fig. 1).6 This posttranscriptional editing generates a 3′ splice site functionally equivalent to AG that facilitates exonization and eliminates a premature stop codon important for maintaining an open reading frame (ORF). The newly described splice variant is expressed in diverse human cell lines and tissues.6

Although the above study demonstrated a distinctly novel pathway to generate novel gene domains, the evolutionary steps leading to such an event and its primate wide distribution have yet to be delineated. In the present study, we successfully PCR-amplified the NARF gene between exons 7 and 9, including their embedded intron(s), in representatives of all primate infraorders. Analyses of these products enabled us to outline specific details of the evolutionary history of the described splice variant of NARF, including the insertion of the targeted Alu element in the lineage leading to anthropoids. The critical RNA-editing steps leading to exonization of exon 8 were analyzed by reverse transcription (RT)-PCR and cDNA sequencing. We identified 10 alternative splice forms, 7 of which were mediated by A-to-I editing, and dozens of further modification sites. Our analyses enabled a retrospective delineation of the period and mode of exonization.

Section snippets

Results and Discussion

To compare the prerequisites for alternative splicing and editing at the DNA level, we cloned and sequenced amplicons of approximately 2.6 kb spanning the sequences between exons 7 and 9 of representatives from all primate infraorders. Then, we derived the mRNA sequences coding for the spliced forms by RT-PCR, cloning, and cDNA sequencing of 900 clones. Because the levels of ADAR-editing activities vary from tissue to tissue,6 it was necessary to analyze the same tissue for all species

Conclusions

Four major events distributed over more than 50 million years of primate evolution gave rise to the birth of the novel exon 8 in the NARF gene (Fig. 4). This chain of events started with the insertion of two head-to-head-oriented AluSx elements, already harboring one of the potential essential A-to-I editing sites, in the intron between exons 7 and 9 in the common ancestor of anthropoid primates. During the same period, a genomic C-to-GT change near the 3′ end of the sense-oriented AluSx

DNA extraction

Following standard protocols,22 genomic DNA was isolated from tissue samples of the Hominoidea: G. gorilla, P. pygmaeus, and H. lar; the Cercopithecoidea (Old World monkeys): C. guereza and T. auratus; the Platyrrhini (New World monkeys): C. jacchus, L. lagotricha, and S. sciureus; the Tarsiiformes: T. syrichta; and the Strepsirrhini: E. coronatus and M. murinus. Genomic DNA from Hominoidea blood of H. sapiens (four different ethnic groups: Japanese, Chinese, Koreans, and African Americans) and

Acknowledgements

We thank Michael Haberl, Bernd Karger, Andreas Ochs, Sandra Silinski, Svante Pääbo, and Peter Wohlsein for providing us with tissue samples and are grateful to Marsha Bundman for editorial assistance. We thank Silke and Andy Scheffel for providing the primate silhouettes of Fig. 4. We appreciate the valuable comments of the two anonymous referees. This work was supported by the Nationales Genomforschungsnetz (0313358A to J.B. and J.S.) and the Deutsche Forschungsgemeinschaft (SCHM1469 to J.S.

References (22)

  • C. Basilio et al.

    Synthetic polynucleotides and the amino acid code

    Proc. Natl Acad. Sci. USA

    (1962)
  • Cited by (0)

    1

    Present address: M. Möller-Krull, Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.

    View full text