Tissue-specific modification of gld-2 mRNA in C. elegans: likely C-to-U editing

Abstract

Seventeen years after the discovery of tissue-specific apoB mRNA editing, only three nucleus-encoded mRNAs have been shown to undergo C-to-U editing. All three mRNAs occur in mammals. apoB mRNA editing is tissue-specific and occurs normally, whereas NF1 and NAT1 mRNA editing is found largely in tumors. Here we report the first example of C-to-U RNA editing in Caenorhabditis elegans. The gld-2 gene encodes an atypical poly(A) polymerase that governs the mitosis/meiosis decision in the germ line as well as progression through meiosis and early embryogenesis. At least two of its alternatively spliced transcripts are germline-specific. We find that most and perhaps all germline-specific transcripts generated by the gld-2 gene undergo C-to-U editing, but that somatic transcripts show no detectable editing. The gld-2 C-to-U editing event changes the codon from CCG to CUG, which is predicted to cause a proline to leucine substitution in the protein sequence. Our findings suggest the presence of a sequence- and tissue-specific cytidine deaminase acting on RNA, or CDAR. This CDAR modifies a specific base in gld-2 mRNA, and acts only in the germline.

FIGURE 1.

Germline-specific gld-2 transcripts undergo C-to-U editing. (A) Comparison of sequences derived from genomic DNA and cDNA. Genomic DNA: The sequences of genomic DNA were derived from three sources. The sequence of cosmid ZC308 was determined by the C. elegans Sequencing Consortium. ZC308 carries a 19,559-bp insert of wild-type C. elegans genomic DNA, including the region that encodes all the gld-2 transcripts. It possesses a C at position 10791. The other two sequences were derived from PCR products (designed to include the region that contains nucleotide 10791) from genomic DNA of two gld-2 mutants, gld-2(q497) and gld-2(h292). These two mutants have nucleotide-substitution mutations elsewhere in the gld-2 gene, and were confirmed in the sequence analysis (see Wang et al. 2002). However, both DNAs possessed a C at position 10791, corroborating the C at position 10791 in the cosmid sequence. mRNA: Sequences of germline and somatic mRNAs were derived in three ways. The first was derived from the sequence of a gld-2 cDNA clone isolated from a phage cDNA library. The library was prepared using total mRNA extracted from mixed-stage worms (including both germline and somatic tissues). A T was detected at position 10791, indicating editing from C to U in the mRNA. The second analysis (“germline-specific mRNAs [PCR1]”) detected germline-specific transcripts. It was performed using the PCR1 reaction (B), using primers in exons 1 and 11. Two germline-specific transcripts exist, designated 4.7L and 4.7S, differing in the use of an exon 3 splice site and inclusion or exclusion of exon 4. The sequence of these PCR products includes a T at position 10791. The third analysis (“somatic mRNAs [PCR2]”) specifically identifies the 4.6-kb somatic mRNA, which carries a C at position 10791, matching genomic DNA. PCR was performed using a primer in the first exon of the somatic mRNA (exon 1 [4.6]), which is missing in germline-specific mRNAs. Underline, AlwNI site. AlwNI cleaves the CAGNNNCTG sequence, which is found in the region of cDNAs but not genomic DNA. Arrowhead, position of cleavage by AlwNI. (B) gld-2 exon/intron structure: two germline-specific transcripts and one soma-enriched transcript. The two germline-specific transcripts, 4.7L and 4.7S, are detected in the PCR1 reaction. This PCR reaction used primers in exons 1 and 11 (open arrows). Exon 1 is present only in germline-specific mRNAs. Exon 1′ (4.6) is present only in the 4.6-kb RNA that is highly enriched in somatic cells. For PCR1 and PCR2 reactions, the primer indicated by a black arrow (in exon 19) was used to prime the reverse transcriptase on total mRNA from mixed-stage worms and q224 worms (germline-less worms), respectively (see Materials and Methods for details). Open boxes, exons; thin lines, introns; arrows, PCR primers; black arrowhead, position of nucleotide 10791 in exon 10. All the gld-2 mRNAs are trans-spliced to SL1. The lengths of the predicted PCR products derived from mRNAs are indicated below the lines. For additional information on mRNA analysis, see Wang et al. (2002). (C) Schematic diagram of the GLD-2 protein sequence. GL-specific, germline-specific region, encoded by exons specific to the germline transcripts corresponding to the 4.7S and 4.7L RNAs (Wang et al. 2002). Black and dark gray, central and catalytic domains of the GLD-2 nucleotidyltransferase (Aravind and Koonin 1999; Wang et al. 2002); light gray, region of GLD-2 protein present only in the germline.

FIGURE 2.

Editing of gld-2 mRNA is efficient and germline-specific. Reverse transcriptase reactions were performed with the same primer (indicated as a black arrow) on poly(A+) RNA from wild-type mixed-stage and q224 worms. q224 mutants possess essentially no germline (Kodoyianni et al. 1992). PCR products of 1464 and 1338 bp were obtained from the germline-specific mRNAs, corresponding to the mRNAs designated 4.7L and 4.7S in Figure 1B. A PCR product of 1269 bp was obtained from the soma-enriched transcripts, corresponding to the mRNA designated 4.6 in Figure 1B. The PCR products corresponding to the germline mRNAs are digested completely by AlwNI, demonstrating that they have been edited. In contrast, most (if not all) of the 1269-bp product is not cleaved, and so is not edited.

FIGURE 3.

Putative cytidine deaminases acting on RNA in C. elegans. An unrooted tree is presented consisting of nine putative cytidine deaminases of C. elegans and the E. coli cytidine deaminase. Transcript numbers of all nine C. elegans genes are given. EcCDA1–150, amino acid 1–150 of E. coli cytidine deaminase. First a multiple sequence alignment of the cytidine deaminase domains was generated by the Superfamily alignment program (http://supfam.org/SUPERFAMILY/). The tree was then built with a phylogenetic tree program (http://www.ebi.ac.uk/clustalw/) using the neighbor joining method, setting on Kimura correction of distances. 0.1, 0.1 substitutions per 100 residues.