Worm Breeder's Gazette 12(5): 80 (February 1, 1993)
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.
We have isolated two ubiquitin genes from C. elegans. One is the polyubiquitin locus described by Roger Graham et al. (Mol. Cell. Biol. (1989)9: 268-271) and the other gene encodes a fusion of ubiquitin with a ribosomal protein (WBG 12(3) 63). We reasoned that the latter gene should have a "universal" promoter since the ribosomal portion of the gene product is presumably required for ribosome function. To date, three constructs have been made and tested. It should be noted that the gene has two small introns, both in the ubiquitin coding region, and that the message is trans-spliced. The constructs were: [See Figure]
No rolling transformants (stable or transient) were obtained with pPCUbil. In addition, we observed that injected animals laid several dead oocytes, suggesting that pPCU bi1 is a lethal construct. The fusion of the ribosomal protein with ß-Galactosidase may impair ribosome function or assembly. pPCU bi2 lacks all of the ubiquitin fusion coding sequence. In addition, 120 bp of the leader were removed, which eliminated the trans-splice acceptor sequence. This construct used an unc-54 ,rather than a ubiquitin, 3' end. pPCU bi2 showed minimal expression in five transgenic lines with only the occasional neuron staining. Probably removal of the trans-splice acceptor site knocked out transcription or processing.
In pPCU bi3 ,we fused ubiquitin to ß-Galactosidase, but removed the ribosomal coding region. This fusion should be cleaved in vivo at the ubiquitin/NLS junction. This construct showed reasonable expression in the three extrachromosomal strains obtained. One strain expressed ß-galactosidase only in the intestine, another in the intestine and nerve ring, and the third stained extensively in neurons, including the nerve ring and other head ganglia, the ventral cord, and tail ganglia. Pharyngeal muscle and the pharyngeal-intestinal valve also stain in this strain. At high magnification, faint staining in the occasional cell of the lateral hypodermis and in uterine and/or vulval muscle was also detectable. Both of these latter patterns were not overwhelming however. While no expression was observed in the germ line or early embryos, a few pretzel staged embryos did stain.
This pattern is still far more limited than we anticipated. Southern analysis suggests that the portion of ubiB encoding the ribosomal protein is present as a single copy in the genome; therefore, we expected that every cell and certainly embryos would constitutively express ubiB. It is possible that some important sequence elements involved in expression or in processing of the transcripts are still missing from the test fusion. We are currently trying more constructs to see if this is the case. Alternatively, we could be witnessing negative interactions between the roller plasmid, pRF4 ,and the ubiB-lacZ fusion, especially since the three strains obtained vary so much in their spatial expression patterns.
Despite our initial optimism, the ubiquitin promoters have so far not demonstrated universal tissue expression. Previous analysis of the 5' sequences of the polyubiquitin gene ubiA, also failed to yield the expected tissue general pattern of expression (Stringham et al., Gene (1992) 113: 165-173). We conclude that the promoters of the ubiquitin genes in C. elegans are rather complicated and, unfortunately, are not neatly contained within a small region upstream from the start site. Supported by the Medical Research Council of Canada.