The divergent 5' ends of DPM2 mRNAs originate from the alternative splicing of two adjacent introns: characterization of the hamster DPM2 gene

Mammalian dolichol-phosphate-mannose (DPM) synthase has three subunits, DPM1, DPM2, and DPM3. In this report, an analysis of the gene and cDNAs of hamster DPM2 is presented. The CHO DPM2 gene has two special features. First, the initiation codon ATG is separated from the remainder of the coding region by intron sequences. Second, within these intron sequences the DPM2 gene contains an adjacent 3' splice site (acceptor) and a 5' splice site (donor), suggestive of a deleted exon between the first and second codons. In fact, these sites overlap by four nucleotides (nt) of AGGT. Splicing intermediates using both of these alternative splice sites were observed. This latter feature appears unique and is particularly unusual considering the relatively small size of the gene (2.7 kb) and of introns a (123 bp) and b (152 bp).     

Spatial constraints on polyadenylation signal function

Efficient cleavage and polyadenylation of eukaryotic messenger RNAs require at least two signal elements: an AAUAAA or closely related sequence located 7-30 base pairs (bp) upstream of the site of processing, and a G/U- or U-rich sequence located 3' to the cleavage site. The herpes simplex virus type 1 thymidine kinase (tk) gene contains two copies of the AATAAA hexanucleotide and a GT-rich region. We have shown that the first AATAAA and the GT-rich region are essential for efficient processing, both in vivo and in vitro, whereas the second AATAAA does not appear to play any role in the formation of tk mRNA 3' ends. The failure of a signal containing only the second AATAAA and the GT-rich element to signal cleavage and polyadenylation suggested that these two elements might be too close together to constitute a functional polyadenylation signal. The experiments described in this report were directed at determining the effects on mRNA 3' end formation of alterations in spacing between signal elements. Wild-type tk contains 19 bp between these two elements. Constructs were made in which an AATAAA and the GT-rich region were separated by various distances ranging from 7 to 43 bp. The quantity and location of 3' ends of the tk mRNA produced by these constructs in Cos-1 cells were measured by S1 nuclease protection analysis. Signal efficiency was gradually reduced as the separation between the two signal elements was increased; with a separation of 43 bp, the signal functioned at approximately one-eighth the efficiency of the parental construction. Bringing the two signals closer together resulted in decreased signal efficiency; with a separation of 7 or 9 bp, no tk mRNA polyadenylated within the normal region was produced. Altering the sequences between these two elements without changing the distance had small effects on processing efficiency.     

Alternative polyadenylation sites of human endothelial nitric oxide synthase mRNA

The mRNA 3'-untranslated region (3'-UTR) has been shown to have important roles in the regulation of mRNA function. In this study, we investigated the human endothelial nitric oxide synthase (eNOS) 3'-UTR to evaluate its potential regulatory role. 3'-RACE analysis revealed that the human eNOS mRNA has multiple alternative polyadenylation sites. Apart from the proximal site (418bp downstream of the stop codon), we identified two additional distal sites approximately 770 and 1478bp downstream of the stop codon. In addition, Northern analysis showed that the usage of these sites differed among human tissues. Further, amounts of these eNOS mRNAs were changed during growth of cultured human aortic endothelial cells; mRNAs with long 3'-UTRs decreased more rapidly than total mRNA, as cells approached confluency. Thus, the 3'-UTRs of human eNOS results from alternative polyadenylation sites and differ across tissues and during cell growth.