In reverse genetics RNAi that is suitable for gene-disruption is an outstanding method for knockdown of gene function. In Caenorhabditis elegans feeding RNAi is the most convenient. However, when knockdown of two genes (double knockdown) is carried out injection RNAi or soaking RNAi is usually adopted with a mixture of dsRNAs produced in vitro. That is due to less reliability of traditional double feeding directed to two genes as follows: in some cases, only one gene’s function may be significantly inhibited, or both genes may be only slightly knocked down by feeding with a mixture of two sorts of RNAi bacteria at the same time.

Therefore, we attempted to establish an efficient method of feeding RNAi for multiple-knockdown. We produced bacteria yielding distinct dsRNAs bound to each other. We selected oma-1 and oma-2 genes that are functionally redundant to each other. Each cDNA was amplified, sub-cloned into a T-vector (pGEM-T EASY, Promega), and then separated from the vector by using EcoRI. The resulting fragments were introduced into the RNAi vector L4440, yielding three sorts of RNAi plasmids: oma-1, oma-2, and both of them. We carried out double feeding for oma-1 knockdown and oma-2 knockdown, and single feeding for oma-1/oma-2 knockdown. Quantitative RT-PCR and observation of phenotypes revealed that our method for double knockdown by single feeding was much more effective than the traditional double feeding. This result has been describe in our report (Gouda et al., 2010).

Next, we applied our method to triple knockdown. We chose gfp gene in addition to oma-1 and oma-2 genes. We constructed an RNAi plasmid including cDNAs for these three genes as described above and carried out single feeding on C. elegans TJ356 strain, which expresses GFP constitutively. Observation of phenotype and GFP fluorescence revealed that knockdown of these three genes was quite effective. Min and coworkers have also reported the same result (Min et al., 2010).

In addition, we asked if our method might be applicable to genes which function in nerve cells. We selected rrf-3 gene in addition to gfp gene and carried out single feeding on the TJ356 strain. GFP fluorescence in nerve cells could be seen, while almost all fluorescence of GFP in somatic cells disappeared. Continuation of the feeding on descendants could not decease GFP fluorescence in nerve cells. Anyhow, our convenient method for multiple-knockdown is quite useful for investigating gene’s function in C. elegans.