We are developing a method to examine the location of specific tagged mRNAs in living adult C. elegans. An elegant technique has been used to examine the distribution of mRNAs in yeast (Bertrand et al. 1998, Beach et al. 1999) and in vertebrate neurons (Rook et al. 2000) in vivo. In this method, cells are simultaneously transformed with two constructs. One construct encodes a fusion between GFP and a viral RNA binding protein (MS2) with a nuclear localization signal (nls) (GFP-MS2-nls). The second construct includes a reporter gene (lacZ), several MS2 binding sites, and the RNA of interest (lacZ-MS2bs-RNAX). When both constructs are present, the GFP-MS2 fusion protein binds to the MS2 binding sites present on the mRNA of the lacZ-MS2bs-RNAX fusion. When the lacZ-MS2bs-RNAX message is present in the cytoplasm, it will bind the GFP-MS2-nls protein and the fluorescence will be detectable in the cytoplasm. If the tagged mRNA is non-uniformly distributed, then the fluorescence will be similarly non-uniform in distribution. Work in a number of organisms has identified a number of localized mRNAs. In most cases, the sequences necessary for sub-cellular localization are present in the 3 untranslated region (3-UTR). If these mRNAs are also localized in C. elegans, then one would expect conservation of the localization sequences in the 3UTR. We have used published information on subcellular mRNA localization as well as conservation of 3-UTR sequences in C. elegans vs. C. briggsae to identify 3UTRs that may be involved in mRNA localization. Initial constructs include lacZ-MS2bs-RNAX where RNAX is the 3-UTR from actin, myosin (
unc-54), Calcium-calmodulin dependent Kinase II (
unc-43), or, as a control,
let-858. Due to difficulties in expression and localization with our initial GFP constructs, we have also had to vary the location and number of nls sequences in our GFP-MS2-nls constructs. We will report on our progress in identifying and monitoring localized mRNAs using this system.