Worm Breeder's Gazette 13(2): 62 (February 1, 1994)

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.

About 70% of C. elegans mRNAs are trans-spliced and as many as 25% could be in polycistronic units

Niansheng Cheng, Diego Zorio, John Spieth, Tom Blumenthal

Department of Biology, Indiana University, Bloomington, IN 47405

Our lab has reported previously that genes that produce mRNAs trans-spliced to SL2 are downstream genes in polycistronic transcription units, which appear analogous to bacterial operons ( 1, 2). We have been investigating the mechanisms that determine the specificity of SL2 trans-splicing and that govern mature mRNA production from polycistronic precursors, and determining how widespread operons are in the C. elegans genome. In this report we describe two approaches we have undertaken to identify additional operons. We have also been able to estimate the percent of C. elegans genes that are trans-spliced and the percent that are co-transcribed with other genes.

In the first approach, we have identified new SL2 -containingmRNAs by making a cDNA library using RT-PCR in which the first strand is primed with oligo(dT) and the opposite strand with the SL2 sequence. We cloned several new cDNAs, including the worm homolog to ribosomal protein S16 by this approach. However most were novel (i.e. were not similar to any gene in Genbank). So far we have investigated only one of the novel genes, which happened to fall within a sequenced cosmid (ZK632). This gene turned out to be the last gene in a 3-gene cluster (all 3 novel). We determined that both of the downstream genes in this cluster receive SL2 ,implying it is a 3-gene operon. Including previously reported examples, we have now successfully predicted the existence of a closely-spaced upstream gene in four separate instances in which a gene's mRNA was known to be SL2 trans-spliced.

In the second approach, we scanned sequenced cosmids (thank you to the sequencing project!) for the presence of clusters of predicted genes in the same orientation, and subsequently tested the downstream genes for SL2 trans-splicing using RT-PCR. We have found numerous clusters and tested five of them. So far, all downstream genes have been found to be SL2 -acceptors.Some, but not all, were also found to receive SL1 in addition. We don't know why some pre-mRNAs receive just SL2 ,while others receive both spliced leaders, although the presence of a second promoter within the operon appears to be responsible for the one case investigated so far(1).

Several potentially interesting clusters have been identified. One 3-gene cluster is composed of a homolog of the TJ6 proton pump at the 5' end, followed by a novel gene, and finally a homolog of glutathione reductase. Nathalie Pujol has evidence suggesting the last gene of the three is unc-32 (3). In addition, Andrew Lynch and Ian Hope have shown (personal communication) that the promoter of the proton pump homolog fused to the lacZ gene results in an expression pattern remarkably similar to that expected on the basis of unc-32 'sphenotype. Our finding that both the middle and third genes are SL2 -acceptorssuggests the 3 genes share a single promoter; if this promoter is active only in certain neurons, as indicated by the lacZ results, loss of glutathione reductase function in those neurons could result in loss of those cells' ability to function, resulting in the unc-32 phenotype.

A second cluster contains two genes: K06H7 .4,encoding a homolog of the yeast sec-7 protein required for processing of secreted proteins through the golgi, and K06H7 .3,which is the worm homolog of isopentenyl-pyrophosphate isomerase, an enzyme in the pathway of cholesterol biosynthesis. Again the downstream gene is an SL2 -acceptor,implying that these two genes are co-transcribed.

Marc Perry drew our attention to another 3-gene cluster, the third gene of which encodes topoisomerase II. The middle gene, which Genefinder had mistakenly fused to topoII, is homologous to a gene from the human major histocompatibility complex (G9a) and to the fly trithorax gene. This is possibly significant because both topoII and trithorax are known to be chromatin-associated proteins. Their coordinated expression could be achieved by co-transcription from a single promoter. The first gene in the cluster is novel. Both downstream genes are SL2 -acceptors.

In sum, all 12 downstream genes in closely-spaced clusters in the same orientation we have tested have turned out to be SL2 -acceptinggenes. Thus it seems fair to conclude that the presence of a cluster indicates the presence of a likely polycistronic transcription unit, and conversely, that the finding that a gene product is trans-spliced to SL2 indicates it is part of an operon.

Interesting unresolved questions include: What is the mechanism by which being a downstream gene in a polycistronic transcription unit specifies SL2 ?How widespread are operons in nematodes? Are polycistronic units a way of co-regulating functionally-related genes in C. elegans?

1. Spieth J. et al., 1993 Cell 73:521-532

2. Spieth J. et al., 1993 C. elegans Meeting Abstracts, p. 340.

3. Pujol, N. 1993 Worm Breeder's Gazette 13 #1 p. 92