The effect of gene localization on RNA processing

Several lines of evidence suggest that the nucleus of mammalian cells is compartmentalized. It is possible that the position of a gene either within the 3-dimensional space of the nucleus or within a local chromatin domain could play an important role in the efficiency with which its transcripts are spliced or polyadenylated, or its mRNA is transported from the nucleus. To examine this hypothesis, I electroporated a 40 kb cosmid fragment that contained the 25 kb dhfr gene together with a pSV2gpt gene into the CHO deletion mutant DG44 which lacks the dhfr locus. Nine clones that had received a single copy of the dhfr gene were examined by fluorescence in situ hybridization (FISH). The transgenes had integrated into a different location in each case. Steady-state dhfr mRNA levels measured by RNase protection and quantitative RT-PCR varied from 1/5 to 3 times the level found in a CHO hemizygote carrying a single dhfr gene in its natural chromosomal location (wild type). To test the idea that dhfr transcripts were differentially processed in independent integrants, I measured the ratio of dhfr pre-mRNA to dhfr mRNA by RT-PCR. All of the integrants exhibited a higher ratio than wild type cells; in the most extreme case, the ratio was 20 times higher than wild type. A higher ratio was found for transfectants that exhibited mRNA levels both higher and lower than wild type. This indicated that some dhfr RNA processing step or steps were slower in the integrants. Splicing is one of the steps that may be sensitive to gene position. I therefore measured the in vivo splicing rate of selected introns, by stopping RNA synthesis with Actinomycin D and detecting the disappearance of intron-exon joints. I found that the transfectant with the lower dhfr mRNA level had a slower splicing rate, while that with the higher mRNA level showed a faster splicing rate than the wild type (which had a t 1/2 of 6 minutes). We conclude that splicing could be affected by gene position, but the 2-fold differences observed were insufficient to account for the greater differences in pre-mRNA:mRNA ratio. I also examined the length of dhfr poly(A) tail in the transfectants, using a poly(U)-Sepharose affinity spin chromatography assay and an RT-PCR-based assay. Similar lengths of poly(A) tails were found for the dhfr RNAs of wild type and the two transfectants studied. Transport of dhfr mRNA from the nucleus to cytoplasm is presently being examine: the preliminary data show equal nuclear/cytoplasmic ratios of mRNA in the wild type and 2 transfectants. This measurement of steady-state levels does not rule out an effect on transport, however. We conclude that gene position can influence post-transcriptional RNA metabolism, with the major effect possibly being at the level of transport and/or nuclear mRNA degradation.