Search

Arginine vasopressin (AVP) and oxytocin (OT) mRNAs are targeted to the axonal compartment of rat hypothalamic magnocellular neurons. Salt-loading results in a considerable rise in hypothalamic and axonal AVP mRNA but only a moderate increase for axonal OT mRNA. Here we report that hypoosmolality gives rise to a rapid decrease of axonal AVP encoding transcripts to undetectable levels after 2 weeks. The levels of OT mRNA in the axonal compartment did not change significantly. In the hypothalamus the mRNA for AVP also decreased. The size of the poly(A) tract of AVP encoding transcripts appeared to be strictly correlated with plasma osmolality. In contrast, the amount and size of OT encoding mRNAs were only moderately or not influenced by hypoosmolar stimuli.

Most magnocellular hypothalamic neurons synthesize the precursor for either vasopressin (AVP) or oxytocin (OT). The AVP precursor is cleaved to give AVP, AVP-associated neurophysin (AVP-NP) and a glycopeptide (GP), whereas the OT precursor gives OT and OT-NP. In Brattleboro rats a frame-shift mutation in the AVP-NP-encoding region of the gene prevents the secretion of AVP by the cells and, in most AVP neurons, AVP itself is virtually undetectable. A small number of magnocellular neurons in homozygous Brattleboro rats contain very large accumulations of peptide in distended saccules of rough endoplasmic reticulum (RER), and this peptide is immunoreactive for AVP and C-terminal OT-NP, but not for OT, AVP-NP or GP (Pow et al., 1992). We have now shown that this results from somatic non-homologous crossing over of the AVP and OT genes, resulting in the production of hybrid mRNA molecules with the 5'end of the AVP sequence and the 3' end of the OT sequence (AVP/OT transcripts). In most cases, the crossing over occurs within the highly homologous B exons (Mohr et al., 1994). In addition to the production of AVP/OT hybrid transcripts, polymerase chain reaction (PCR) amplification of mRNA from the hypothalami of homozygous rats also reveals OT/AVP hybrid transcripts, with 5' OT sequences and 3' AVP sequences. Furthermore, both types of hybrid transcript are not restricted to homozygous Brattleboro rats but can also be found in normal Long Evans animals. To date, we have not been able to locate cells in which the OT/AVP hybrids are produced; all the magnocellular neurons with hybrid peptide accumulations in the RER so far studied have been shown by immunocytochemistry to be of the AVP/OT type. In both normal and homozygous Brattleboro rats large accumulations of peptide do occur in the RER of OT-producing neurons but the peptide is immunoreactive for OT and OT-NP but not for AVP, AVP-NP or GP. Such cells increase in number 10-fold after injection of 20 micrograms estradiol daily for 7 days (Pow et al., 1991). Why this apparently normal gene product accumulates within the RER remains to be determined.

Molecular biological and immunocytochemical data demonstrate nonhomologous crossing-over between the closely linked vasopressin (VP) and oxytocin (OT) genes in rat hypothalamic neuroendocrine neurons. Reverse transcription of hypothalamic total RNA from wild-type or homozygous Brattleboro aged rats combined with polymerase chain reaction (PCR) amplifications in the presence of appropriate 5' forward and 3' reverse primers deduced from the VP and OT cDNA sequences yielded PCR products that, upon cloning and sequencing, revealed several hybrid transcripts. They encode the N-terminal part of the VP precursor fused to the C-terminal part of the OT precursor (VP/OT transcripts) and vice versa (OT/VP transcripts). VP/OT hybrid precursor proteins have been identified immunocytochemically in enlarged cisternae of the rough endoplasmic reticulum, yet there is no evidence that the products can be secreted from affected cells. Recombination appears to be a rather frequent genetic event affecting about 0.06-0.1% of the rat vasopressinergic magnocellular neurons in aged rats.

Hypothalamic vasopressin and oxytocin transcripts have been detected in the posterior pituitary suggesting either transcription of the respective genes in pituicytes or axonal mRNA transport from the hypothalamus to the nerve terminals of the posterior pituitary. The concept of axonal mRNA transport is supported firstly, by Northern blot and in situ hybridization analysis indicating that vasopressin and oxytocin mRNAs are also present in the neural stalk; secondly, by intron analysis and transcription run on experiments demonstrating the absence of primary vasopressin and oxytocin transcripts in non-neuronal cells of the posterior pituitary; thirdly, by embryonic developmental studies showing that appearance of vasopressin transcripts in the hypothalamus and the pituitary anlage is correlated. Furthermore, during axonal transport the respective mRNAs are subject to specific modification at the poly(A) tails.

Sequence analysis of the rat vasopressin and oxytocin gene family reveals that the two genes are linked by a long interspersed repeated DNA element (LINE) giving rise to seven long open reading frames encoding hypothetical proteins of 99 to 556 amino acid residues. Furthermore, although both DNA strands of LINEs serve as templates for transcription, transcripts initiated at the 3' end are more abundant than those started from the 5' end. The LINEs are transcribed preferentially in brain tissues as analyzed by Northern blot, in situ hybridization, and RNase protection experiments. The data show that most LINEs are transcribed at their entire length and that a major fraction of respective RNAs does not enter the cytoplasm but remains in the cell nucleus.

Expression of the gene encoding the oxytocin precursor occurs in the hypothalamus and, to a lesser extent, in a number of peripheral organs, the tissue-specific regulatory mechanisms of which are largely unknown. By DNA sequence analysis several elements upstream of the transcriptional start point of the rat oxytocin gene were identified matching the consensus sequence of enhancers inducible by estrogen or glucocorticoids, respectively. Their general transactivating capacities were investigated using heterologous gene constructs and revealed that the rat oxytocin gene harbours two functional estrogen responsive elements near the transcription initiation site, one of which is conserved in the respective human gene. In addition, one enhancer conferring glucocorticoid responsiveness to a reporter gene is located at nucleotide residues -2449 to -2464. These data might indicate a steroid hormone-mediated influence on oxytocin gene expression in the central nervous system and/or the periphery.

In rats, vasopressin- and oxytocin-encoding mRNAs are present in the posterior but absent in the anterior lobe of the pituitary gland. RNase protection experiments indicate that in the posterior pituitary and hypothalamus identical transcriptional start points are used. Furthermore, the two transcripts from posterior pituitary and hypothalamus show identical nucleotide sequences. Animals operated by paired electrical lesions in such a way that connections between the supraoptic nucleus (SON) and paraventricular nucleus (PVN) of the hypothalamus and the posterior pituitary lobe are destroyed continue to express the vasopressin and oxytocin gene in the hypothalamus but not in the posterior pituitary. Operated animals subjected to chronic intermittent salt loading for 6 days similarly contain vasopressin and oxytocin encoding transcripts in the hypothalamus but not in the posterior pituitary.

Poly(A)+ RNA isolated from post-mortem human hypothalami has been used to characterize the poly-protein precursors to vasopressin and oxytocin. Translation in a cell-free system and subsequent immuno-precipitation with antibodies raised against either vasopressin or neurophysin identified a product of Mr 19000 (prepro-vasopressin). A second less intense product of Mr 16500 was tentatively identified as prepro-oxytocin. A cDNA library derived from the human hypothalamic poly(A)+ RNA was screened for vasopressin and oxytocin-encoding cDNA using heterologous probes; clones encoding the two precursors were identified and found to be organized as their rat and bovine counterparts. Northern blot analysis shows that the mRNAs for the two prepro-hormones consist of approximately 840 (AVP) and approximately 700 (OT) nucleotides.

An 18 kb DNA fragment, containing the genes encoding both the vasopressin and oxytocin polyprotein precursors, has been isolated from a rat genomic library. The two genes are linked by approximately 11 kb of intervening sequence and transcribed from opposite DNA strands.

The genes for the hypothalamic hormones vasopressin and oxytocin are located in close proximity to each other within the rat genome. They are separated by only approx. 11 kbp of DNA sequence and oriented in such a way that their transcription occurs on opposite DNA strands. Although the two genes are structurally very similar including common potential regulatory elements in their putative promotor regions, they are expressed in discrete populations of magnocellular neurons of the hypothalamus. In rats placed under osmotic stress, the vasopressin gene is upregulated; concomitantly transcription of the oxytocin gene is also stimulated. To address the question of whether this coordinated rise in oxytocin-encoding mRNA is the result of switching on oxytocin gene transcription in vasopressinergic neurons, in situ hybridization with double labelled cRNA probes was carried out. Biotinylated and [alpha-35S]CTP labelled antisense cRNA probes specific for either vasopressin or oxytocin mRNA were constructed and hybridized to hypothalamic sections from salt-loaded rats. The results demonstrate that upregulation of oxytocin gene transcription is restricted solely to oxytocinergic cells; no oxytocin gene transcripts can be detected in vasopressinergic neurons.

Complete cDNA sequences for the vasopressin and oxytocin precursor polyproteins have been determined for the rat, calf, human and pig (vasopressin only), indicating the essential conservation of the precursor structures throughout mammals. DNA probes specific for vasopressin or oxytocin mRNAs have been used to identify both classic (hypothalamic) and novel (thymus, corpus luteum, phaeochromocytoma) sites of hormone expression. Semiquantitative DNA/RNA hybridization suggests that in rats expression of the vasopressin and oxytocin genes is positively effected by osmotic stress, negatively by a systemically applied excess of vasopressin; in the latter experiment a reduction in the hypothalamic levels of vasopressin and oxytocin mRNAs in normal and Brattleboro rats have been observed. This suggests a feedback regulation by the hormone as a possible element in controlling the transcription of the vasopressin gene.