No abstract

Some internal adenosine residues in messenger RNA are methylated posttranscriptionally in the nucleus. Most of the methylated adenosine residues in prolactin mRNA are in the 3' untranslated region. The site of methylation in the 3' end of prolactin mRNA was determined. This methylation reaction is highly specific; of the three adenosine residues in consensus sequences located in the 3' end, only one is methylated. An in vitro methylation system was developed in which bovine prolactin mRNA, synthesized in vitro with T7 RNA polymerase, was accurately methylated in a HeLa cell nuclear extract. The adenosine residue that was methylated in vitro was the same as the one methylated in vivo. This cell-free system, which accurately methylates the N6-position of adenosine residues in mRNA, will allow further study of the mechanism of adenosine methylation.

N6-methyladenosine (m6A) residues occur at internal positions in most cellular and viral RNAs; both heterogeneous nuclear RNA and mRNA are involved. This modffication arises by enzymatic transfer of a methyl group from S-adenosylmethionine to the central adenosine residue in the canonical sequence G/AAC. Thus suggested following the precise localization of m6A in Rous sarcoma virus (18). In that molecule, the modified residues were found bracketing the src splice acceptor site and therefore may influence in some way the accuracy and/or efficiency of processing. Methylation of internal adenosine residues of eucaryotic cell RNA also takes place in the nucleus, as evidenced by the demonstration of m6A in eucaryotic cell heterogeneous nuclear RNA (24). However, previous analysis indicated a conservation of m6A residues during processing and transport to the cytoplasm, suggesting that this modification occurs mainly within exon regions (19). It was therefore of great interest to determine whether m6A occurs in intron-specific regions of eucaryotic precursor RNA. This analysis would identify the temporal relationship between transcription, methylation, and nuclear processing and may assist in developing a concept for the role of this posttranscriptional modification in eucaryotic RNA biogenesis.Attempts have been made to determine the influence of m6A on processing of gene transcripts. Earlier studies with cycloleucine, an inhibitor of S-adenosylmethionine synthetase, suggested that m6A affects the splicing of avian sarcoma virus genomic RNA (34) and processing of CHO cell mRNA (2). This hypothesis is supported by the similarity between the m6A consensus sequences in both eucaryotic and viral RNA and the weakly conserved branch point sequence YNCU(G/A)AC used in intron removal (18,27,40). Camper et al. (5) used the mRNA methylation inhibitor S-tubercidinyl homocysteine to measure the effects of undermethylation on the half-life and cytoplasmic appearance of poly(A)+ RNA in HeLa cells. A significant effect on processing and/or transport of undermethylated mRNA occurred with a delay in its appearance in the cytoplasm, whereas no significant alteration in the cytoplasmic half-life of undermethylated mRNA was observed.

To study the effects of premature and chronic ligand-mediated luteinizing hormone receptor (LHR) activation on reproductive development, we have generated transgenic mice expressing a genetically engineered, constitutively active yoked hormone-receptor complex (YHR), in which a fusion protein of human chorionic gonadotropin (hCG) is covalently linked to the N-terminus of rat LHR. YHR-expressing mice (YHR + ) were analyzed at pre-and post-pubertal ages. Relative to wild type (WT) controls, male mice exhibited prepubertal increases in testosterone levels and seminal vesicle weights, and decreases in serum FSH, serum LH, testes weight, and the size of the seminiferous tubules. In adult male YHR + mice, testosterone and LH levels are not significantly different from WT controls. However, FSH levels and testes weights remain decreased. Female YHR + mice undergo precocious puberty with early vaginal opening, accelerated uterine development, enhanced follicular development, including the presence of corpora lutea, and an increase in serum progesterone. At 12 weeks of age, the ovary exhibits a relative increase in the amount of interstitial tissue, comprised of cells that are hypertrophic and luteinized, as well as follicles that are degenerating. Additionally, hemorrhagic cysts develop in approximately 25% of the transgenic mice. These degenerative changes are consistent with an aging ovary suggesting that CG-induced LHR activation in female mice leads to precocious sexual development and ovarian lesions. Taken together, these data indicate that the single chain YHR is functional in vivo and demonstrate that YHR + mice provide a novel system to further understand the reproductive consequences of aberrant LHR activation.

The methylation of internal adenosine residues in mRNA only occurs within GAC or AAC sequences. Although both of these sequence motifs are utilized, a general preference has been noted for the extended sequence RGACU. Not all RGACU sequences in an mRNA are methylated and the mechanisms that govern the selection of methylation sites in mRNA remain unclear. To address this problem we have examined the methylation of transcripts containing sequences of a natural mRNA, namely, bovine prolactin mRNA. In this mRNA, a specific AGACU sequence in the 3' untranslated region is the predominant site of methylation both in vivo and in vitro. The degree to which N6-adenosine methyltransferase recognizes the sequence context of the consensus methylation site was explored by mutational analysis of the nucleotides adjacent to the core sequence as well as the extended regions in which the core element was found. Our results indicate that efficient methylation depends on the extended five nucleotide consensus sequence but is strongly influenced by the context in which the consensus sequence occurs within the overall mRNA molecule. Furthermore, consensus methylation sites present in an RNA duplex are not recognized by the methyltransferase.