Mini-Oct and Oct-2c: two novel, functionally diverse murine Oct-2 gene products are differentially expressed in the CNS.

We report that two novel alternatively spliced products of the murine Oct-2 gene encode Mini-Oct (Oct-2d), a protein consisting of almost only the POU domain, and Oct-2c, a protein lacking the last 12 amino acids of Oct-2a. Ectopic expression in HeLa cells shows that Oct-2c is a transactivator, whereas Mini-Oct fails to transactivate if the octamer motif is in a promoter position next to TATA box. Mini-Oct can repress the transcriptional signal generated by endogenous octamer factors in F9 cells. It seems that Mini-Oct has the potential to serve as a transcriptional modulator for genes regulated by different octamer-binding factors. In situ hybridization reveals that Mini-Oct expression follows the general pattern of other known Oct-2 transcripts. However, it is absent from the Purkinje cell layer in the cerebellum of adult mice, and strong expression is observed in the developing nasal neuroepithelium and primary spermatids. Differential expression patterns of the Oct-2 transcripts with different transactivation/repression capacities of the encoded proteins may have a specific role in gene expression in the developing nervous system and in adult brain.

Structure and expression of the mouse Oct2a and Oct2b, two differentially spliced products of the same gene.

A large family of tissue-specific nuclear proteins interact with the octamer motif ATTTGCAT, a transcriptional regulatory element found in the promoter and enhancer sequences of many genes. As a step towards elucidating the mechanism of this regulation, cDNA clones of the mouse Oct2 protein were isolated. One, called here Oct2b, encodes a larger variant of the previously described Oct2a proteins. The Oct2b cDNA has an insertion of 74 bp close to the 3' end which creates an open reading frame distinct from Oct2a. As a result, the Oct2b protein has a carboxy end which is similar to that of the ubiquitous octamer-binding protein Oct1. Analysis of the Oct2 gene shows that Oct2a and Oct2b are differentially spliced products of the same gene. The insertion in the Oct2b cDNA results from the inclusion of an additional exon in the mRNA which would otherwise reside in an intron sequence of the Oct2a transcript. RNA analysis demonstrates that both Oct2a and 2b mRNAs are most abundant in B-cells but they are also expressed in a variety of tissues including brain, intestine, testis, kidney, as well as in embryos. Interestingly, the ratio of Oct2a and 2b varies among tissues. In situ hybridization studies during mouse embryogenesis show that the Oct2 gene is widely expressed in the developing nervous system. In contrast, expression in the adult brain is confined to very specific areas which include the suprachiasmatic and medial mammillary nuclei, hippocampus, olfactory tract and the olfactory bulb. Oct2 proteins are present in both neuronal and oligodendroglial cells, although they are more abundant in glial cells.

Ribosome biogenesis and nucleolar ultrastructure in neuronal and oligodendroglial rat brain cells.

The absolute amounts of precursor to ribosomal RNA (pre-rRNA) and ribosomal RNA (rRNA) in isolated rat brain neuronal and oligodendroglial nuclei were determined. The amount of the major pre-rRNA and rRNA species in neuronal nuclei was about twofold higher than in oligodendroglial nuclei. The relative rate of pre-rRNA synthesis in vivo was 2.3- to 2.7-fold higher in neuronal as compared with oligodendroglial nuclei. This corresponds to a 2.7-fold higher activity of the "template-bound" RNA polymerase I in isolated neuronal nuclei, whereas the activity of the "free" enzyme in both neuronal and glial nuclei was almost identical. The higher transcription rates of rRNA genes correlated with the markedly more prominent fibrillar component in neuronal nucleoli. The turnover times of the major pre-rRNA and rRNA species in neuronal and oligodendroglial nuclei were similar, except for 45S pre-rRNA, which turned over at an approximately 1.5-fold slower rate in neuronal nuclei. The relative rates of processing of pre-rRNA and of nucleocytoplasmic transport of rRNA in neuronal cells were approximately 2.7-fold higher than in oligodendroglial cells and corresponded to the differences in rRNA gene transcription rates. The established ribosome formation features correlated with an abundant (neurons) or exceedingly scarce (oligodendrocytes) nucleolar granular component. The turnover rate of cytoplasmic ribosomes in rat brain neurons was twofold slower than in oligodendrocytes, largely because of the about fivefold higher amount of ribosomes in the cytoplasm of neurons. We conclude that ribosome formation and turnover in neuronal and oligodendroglial cells are adapted to the protein synthetic levels in these two types of brain cells.

Different rates of synthesis and turnover of ribosomal RNA in rat brain and liver.

The kinetics of in vivo labeling of cellular free UMP and nucleolar, nucleoplasmic, and cytoplasmic rRNA with [14C]orotate in rat brain and liver were investigated. Evaluation of the experimental data shows: (a) The rate of nucleolar precursors of ribosomal RNA (pre-rRNA) synthesis and the deduced rate of ribosome formation in brain is about fivefold lower than in liver and corresponds to 220-260 ribosomes/min/nucleus. (b) The lower rate of in vivo pre-rRNA synthesis is correlated with a lower activity of RNA polymerase I in isolated brain nuclei. (c) The half-lives of nucleolar rRNA in brain and liver are 210 and 60 min, respectively, thus showing a slower rate of processing of pre-rRNA in brain nucleoli. (d) The nucleo-cytoplasmic transport of ribosomes in brain is also markedly slower than in liver and reflects the lower rates of synthesis and processing of pre-rRNA. (e) Cytoplasmic ribosomes in brain and liver turn over with half-lives of about 6 and 4 days, respectively. It is concluded that the markedly lower rate of ribosome biogenesis in brain is specified mainly at the level of transcription of rRNA genes.

Quantitative analysis of rat liver nucleolar and nucleoplasmic ribosomal ribonucleic acids.

rRNA from detergent-purified nuclei was fractionated quantitatively, by two independent methods, into nucleolar and nucleoplasmic RNA fractions. The two RNA fractions were analysed by urea/agar-gel electrophoresis and the amount of pre-rRNA (precursor of rRNA) and rRNA components was determined. The rRNA constitutes 35% of total nuclear RNA, of which two-thirds are in nucleolar RNA and one-third in nucleoplasmic RNA. The identified pre-rRNA components (45 S, 41 S, 39 S, 36 S, 32 S and 21 S) are confined to the nucleolus and constitute about 70% of its rRNA. The remaining 30% are represented by 28 S and 18 S rRNA, in a molar ratio of 1.4. The bulk of rRNA in nucleoplasmic RNA is represented by 28 S and 18 S rRNA in a molar ratio close to 1.0. Part of the mature rRNA species in nucleoplasmic RNA originate from ribosomes attached to the outer nuclear membrane, which resist detergent treatment. The absolute amount of nuclear pre-rRNA and rRNA components was evaluated. The amount of 32 S and 21 S pre-rRNA (2.9 x 10(4) and 2.5 x 10(4) molecules per nucleus respectively) is 2-3-fold higher than that of 45 S, 41 S and 36 S pre-rRNA.