Systemic endotoxin increases steady-state gene expression of hypothalamic nitric oxide synthase: Comparison with corticotropin-releasing factor and vasopressin gene transcripts

The enzyme responsible for nitric oxide (NO) formation, NO synthase (NOS), is found in hypothalamic neurons that control ACTH secretion. This led to the hypothesis that brain NO may modulate the response of the hypothalamic-pituitary (HP) axis to various stimuli. We tested this hypothesis by measuring changes in constitutive (c) NOS mRNA levels in the hypothalamus of rats systemically injected with endotoxin, a lipopolysaccharide (LPS) that releases endogenous cytokines, and analyzed these results in the context of the appearance of ACTH-releasing secretagogues such as corticotropin-releasing factor (CRF) and vasopressin (VP), as well as CRF receptors type A (CRF-R(A)). We purposefully chose doses of LPS thought to only minimally disrupt the blood-brain barrier and not be accompanied by an endotoxin shock, so that the results we obtained did not primarily stem from abnormal passage of compounds into the brain, or non-specific stress. Three to four hours following LPS injection (100 mu g/kg, i.v.), cNOS mRNA levels increased in the paraventricular nucleus (PVN) of the hypothalamus. LPS treatment also upregulated PVN CRF gene transcription (measured by CRF heteronuclear RNA) and increased steady-state gene expression of the immediate early genes (LEG) c-fos and NGFI-B, with the first changes noted 1-2 h after treatment. Transcripts of CRF receptors type A were present in the hypothalamus 6 h after endotoxin treatment. On the other hand, no alterations in cytoplasmic VP mRNA levels were noted in rats injected with LPS. Because the dose of LPS we used stimulates ACTH secretion within 30 min, our results suggest that systemic LPS acts first within the median eminence, where it stimulates peptidic nerve terminals. Neuronal activation of hypothalamic cell bodies takes place later, and whether this phenomenon is due to the production of brain neurotransmitters and/or cytokines, or whether it primarily results from increased demand on the synthetic machinery, remains to be established. These studies extend prior work showing that systemic LPS increases the neuronal activity of hypothalamic regions known for their involvement in the responses of the HP axis, and bring forth two important additional points. First, increases in CRF primary nuclear transcripts are delayed with regard to the temporal release of ACTH. This suggests, though it does not demonstrate, that under the experimental conditions we used, the first site of action of LPS is the median eminence. Second, the observation of increased cNOS gene expression following LPS treatment, and the presence of this enzyme in neurons that regulate ACTH secretion, bring support to the hypothesis that this gas plays an important function in mediating the HP axis response to an immune challenge.