To express a transgene only in cells of interest at the time of interest, we add a smg-sensitive (RNA destabilizing) extension to a specific construct and transform that into a temperature-sensitive smg mutant. At 15 degrees C,
smg-1(
cc546) is active and degrades smg- sensitive transcripts (aberrant transcripts containing premature stop codons or long 3' UTRs), limiting translation of these transcripts. At 25 C,
smg-1(
cc546) is inactive, allowing accumulation and translation of these transcripts. A cell-specific promoter limits expression of the construct to the tissue of interest. (see also Getz et al., WBG 14#5,
p26, and Zwaal et al., IWM97,
p128.) We have tested this method by expressing GFP under two different promoters. The first, from
myo-2, is a strong pharyngeal muscle-specific promoter. Upon injection at a concentration of 4 ng/ul, worms show low but detectable fluorescence when grown at 15 C, and an estimated 8 to 10-fold increase in fluorescence at 25 C. With higher concentrations the difference is less obvious (at least by eye) because of bright staining even at 15 C. With lower concentrations not all the transgenic lines contained the GFP construct (probably the construct just wasn't incorporated by chance, because of the low copy number). The smg-tagged GFP was also expressed under the
mek-1 promoter, which drives expression predominantly in the gut, the pharyngeal muscles and some as yet unidentified neurons. At 25 ng/ul, there was faint fluorescence at both 15 C and 25 C. At 75 ng/ul, there was faint fluorescence at 15 C (although 3 or 4 neurons were consistently visible), but very bright fluorescence in the proper tissues at 25 C. Together these results show that this method can be used for stage- and tissue-specific expression of transgenes engineered to be smg-sensitive, but that one needs to fiddle around a bit to find the optimal concentra tion for injection. Visualization of GFP requires high levels of expression, whereas to study biologically active proteins (e.g. those involved in signal transduction) one might need much lower expression. Indeed, injection of a smg-tagged, dominant active form of
mek-1 under the
myo-2 promoter gave fewer transgenic F1s then expected, and transgenics were very starved and grew very slow (in concordance with a role for
mek-1 in pharyngeal pumping, see Koga et al., WBG 14#2,
p72). No lines containing the transgene could be obtained (presumably due to selection against presence of the transgene). One line was obtained with smg-tagged, dominant active
mek-1 under its own promoter, injected at 4 ng/ul (20-fold lower than the GFP construct). This transgene was subsequently integrated into the genome. Worms are healthy at 15 C, but die as embryos or early larvae at 25 C. At 20 C, most of the worms are fairly healthy, but at room temperature (a nice 23C down here in Texas), the worms are very sick and there is a high degree of larval lethality. This shows that transgenes with severe effects on the animal's general well-being can be used, and that intermediate temperatures can be helpful in obtaining intermediate phenotypes.