The nematode Caenorhabditis elegans is an important model animal for biological research. Currently, transgenic C. elegans strains are mainly generated by injecting DNA encoding a gene of interest, in combination with a reporter gene, into the gonad. With this approach, the interpretation of negative results, such as the failure to observe reporter expression, is frequently required. Single, selectable vectors are urgently required. Internal ribosome entry site (IRES) elements are known to bind the eukaryotic ribosomal translation initiation complex and independently promote translation initiation. Bioinformatic analysis predicted an IRES motif upstream of the start codon of the C. elegans Hsp-3 gene. While this sequence has a Y-shaped double-hairpin secondary structure characteristic of IRES elements, it was unclear if it could function as an IRES. In the present study, this predicted Hsp-3 IRES was incorporated into a bicistronic vector driven by the myo-3 promoter, which allowed co-expression of RFP and GFP genes in the muscle tissue of C. elegans and thereby demonstrated that this IRES element is functional. This vector provides a novel, powerful tool for C. elegans research.

Caenorhabditis elegans is an important model organism in many areas of biology, including development, cell biology, and neurobiology. An essential aspect of C. elegans research is the generation of transgenic lines. The most commonly used method for transformation is microinjection of two vectors, one encoding a gene of interest and the other a reporter gene, such as rol-6 (roller phenotype), into the gonad of a C. elegans hermaphrodite (1). However, there are several disadvantages to this method of engineering transgenic worms. First, it is difficult to control the ratio of the two different vectors for injection. In addition, various events can lead to preferential silencing, such that expression of the reporter gene cannot be observed (2). Single selectable vectors are therefore urgently needed. A large number of single plasmid selectable vectors from the Fire Vector Kit are widely used. These vectors involve fusing a gene of interest with the GFP reporter gene. However, this approach may disrupt the topology and function of the target protein or inhibit GFP fluorescence.

Within the past decade, many studies have reported the use of internal ribosome entry site (IRES)-dependent vectors in viruses, yeast, bacteria, and mammalian cells (3-6) for bicistronic expression of genes. IRES elements were originally identified in virus RNA and were found to hijack translation initiation in the host cell. They bind the eukaryotic ribosomal translation initiation complex and independently promote translation initiation (7). They have also been discovered in the 5′-untranslated regions of some eukaryotic mRNAs (8). The structural features of IRES elements remain largely unknown (8,9); however, Le and Maizel predicted a common IRES motif that has a Y-shaped double-hairpin structure and is found upstream of the start codon of the C. elegans gene Hsp-3(10). There was, however, no experimental evidence for the functional role of this putative IRES.

The standard method of defining the activity of an IRES sequence is to test its ability to initiate translation from the second open reading frame (ORF) of a bicistronic construct. According to this premise, we constructed a bicistronic vector containing the predicted Hsp-3 IRES sequence and driven by the myo-3 promoter. We then examined whether the construct was able to co-express RFP and GFP in the muscle tissue of C. elegans.

Materials and methods

Plasmid construction

The predicted Hsp-3 IRES (M26604) was amplified from C. elegans genomic DNA using HiFi DNA polymerase (Transgen, Beijing, China) and the primers IRES-F: 5′-TGCTCTCCCTT-CACCACTCCCATCG-3′ and IRES-R: 5′-GCCCAATAAGAA-TAAGGTCTTCATA-3′. This fragment contains 285 bp upstream of the Hsp-3 start codon and was ligated into the T1 vector using the pEASY-T1 Cloning kit (Transgen) (Figure 1, step 2) to generate sufficient material for subsequent manipulations. The entire RFP ORF was amplified from pDsRed1-C1 using the primers RFP-F: 5′-GAGCTCATGGCCTCCTCCGAGAACGTCATCA-3′ and RFP-R: 5′-GGATCCCTACAGGAACAGGTGGTGGCGGCCC-3′, which contain SacI and BamHI sites, respectively (shown in italics). The entire GFP open reading frame was amplified from L3786 using the primers GFP-F: 5′-GATATCATGAGTAAAGGAGAAGAACTTTTCA-3′, which contains an EcoRV site, and GFP-R:5′-TCTAGATTACT-TGTATGGCCGGCTAGCGAAT-3′, which contains XbaI and NheI sites to enable ligation into pPD118.25 (L3786). PCR was performed under the following reaction conditions: initial denaturation at 94°C for 5 min, 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min, and a final 7-min extension at 72°C. After digesting each of the purified products with the appropriate restriction enzymes, GFP and RFP sequences were ligated into pEasy-T1/IRES (Figure 1, step 3). pEasy-T1/RFP-IRES-GFP was then digested with KpnI and NheI, and the RFP-IRES-GFP fragment was cloned into pPD118.25 (L3786), which contains an apparent internal enhancer from let-858, the upstream sequences of the let-858 translation start site, a multiple cloning site, a coding sequence for gfp, and the let-858 3′ sequence (Figure 1, step 4). Since the GFP sequence in pPD118.25 contained an upstream reverse T7 promoter with a stop codon, it was replaced with the RFP-IRES-GFP cassette. The RFP-IRES-GFP-let858.3′ fragment was amplified using primers sxRIGF: 5′-GAGCTCTCTAGAATGGCCTCCTCCGAGAAC-3′, which contains SacI and XbaI sites to enable ligation of a myo-3 tissue-specific promoter, and RIG3′ R: 5′-TTAGATTTGGATTGAATTAATTTTT-3′. The PCR product was ligated into the pEasy T1 simple vector (Transgen). The myo-3 promoter was cloned using the primers MyoF: 5′-GAGCTCTGTGTGTGATTGCTTTTTCACAATC-3′ and MyoR: 5′-TCTAGATGGATCTAGTGGTCGTGGG-3′, and the sequence was ligated into the T simple/RFP-IRES-GFP-let858.3′ construct. A control plasmid was constructed where we replaced the IRES fragment with a filler sequence (285 bp) obtained by amplifying the Hemonchus contortus 24-kDa excretory/secretory protein (24ES; accession no. U64793.1, GenBank) sequence using primers 24ES-F: 5′-GGATCCTACGCTCCTAAGGCAGCCAGGATGT-3′ and 24ES-R: 5′-GATATCACTACACCACTCTACTGCGCATCCG-3′ and then ligating the PCR fragment into T1 simple/Pmyo3:RFP-IRES-GFP:let858.3′ using the BamHI and EcoRV sites, which replaces the IRES fragment with 24ES. All cloning steps were verified by DNA sequencing.