Worm Breeder's Gazette 16(4): 18 (October 1, 2000)

These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.

Snap-back RNAi in neurons

Kim Schuske, Fude Huang, Dan Williams, Erik Jorgensen

Department of Biology, University of Utah, Salt Lake City, UT 84112

Double stranded RNA interference (dsRNAi) is an effective method for suppressing the expression of a gene. However, genes expressed in the nervous system, in particular those that are involved in neuronal function rather than development, seem to be refractory to RNAi mediated gene disruption. For example, the construct PY38H6C.17:GFP is expressed in neurons, body muscles, pharynx and hypodermis. When we injected GFP-dsRNA, fluorescence was lost in all tissues except neurons. Failure of RNAi in the nervous system might be caused by an inability to transport dsRNA into neurons, or alternatively, by a lack of RNAi mechanisms in neurons. Recently, Tavernarakis et al. demonstrated that expressing dsRNA in neuronal cells resulted in robust RNAi phenotypes, even for genes required for neuronal function (1). In these experiments, dsRNA was expressed in the nervous system using a snap-back gene construct, that is, an inverted repeat of the open reading frame placed under the regulation of a heatshock promoter. We have further explored this method of gene disruption by targeting the unc-47 gene.

unc-47 encodes the vesicular GABA transporter and is required for GABA neuronal function. unc-47 mutants show a 'shrinking' phenotype when touched on the nose. Worms that were injected with unc-47 dsRNA did not exhibit a phenotype. We then designed several snap-back constructs with inverted repeats (IR) of 449 base pairs of the unc-47 coding DNA. However, the constructs were not stable in bacteria, perhaps due to the formation of cruciform DNA structures which would disrupt replication. For this reason we made snap-back constructs which were generated from two separate plasmids. One plasmid contained one of three different promoters driving unc-47 DNA in the coding orientation, the second plasmid contained unc-47 DNA in the non-coding orientation and the unc-54 terminator. These two plasmids were digested and ligated to form a fragment of the following composition: promoter:coding:noncoding:terminator. The ligated fragments were gel purified and injected into lin-15 (n765ts) worms at a concentration of 25ng/ul. A lin-15(+) rescuing plasmid (EKL15) was co-injected at a concentration of ~50ng/ul.

Three different promoters were used to drive expression of the snap-back construct, hsp16-2 a heat inducible promoter, Punc-119, which is expressed in all neurons, and Punc-47, which is expressed only in the GABA neurons. Four lines were obtained containing the hsp16-2: unc-47(IR) construct. None of these showed a phenotype after heat shocking L1-L2, L3-L4 or adult worms for one hour at 35 C. Five lines were obtained containing the Punc-119: unc-47(IR) construct, none of which showed any discernible phenotype. However, the three lines containing the Punc-47: unc-47 (IR) construct displayed the shrinker phenotype. Interestingly, the first shrinkers that were observed were found in one line in the F2 generation. In the other two lines shrinker animals were only observed in the F3 and later generations. Moreover, it appeared that the proportion of shrinker animals increased in later generations. By the F4 generation between 50-87% of the lin-15 rescued animals displayed a shrinker phenotype, although the severity of the phenotype was highly variable. By contrast, co-injection of plasmids with the unc-47 promoter expressing coding RNA and a second expressing noncoding RNA did not result in a shrinker phenotype. This result suggests that the single strands of RNA do not efficiently form a double stranded molecule in vivo. It also suggests that it is unlikely that the unc-47 mutant phenotype observed in animals containing the snap-back construct is due to titration of transcription factors, since worms containing the two plasmids did not show the unc-47 phenotype.

We can make three conclusions from this work. First, RNAi only worked when expressed from a snap-back construct. Second, the strength of the promoter used to drive expression of the snap-back RNA molecule may be critical for the generation of a phenotype. Third, the phenotype was variable and became more severe over multiple generations.

1. N. Tavernarakis, S. L. Wang, M. Dorovkov, A. Ryazanov, M. Driscoll, Nature Genetics 24, 180-183 (2000)