Recent studies have revealed that TRH-like immunoreactivity (TRH-LI) in human serum is predominantly pGlu-Glu-ProNH2 (< EEP-NH2), a peptide previously found in, among others tissues, the pituitary gland of various mammalian species. In the rat pituitary, < EEP-NH2 is present in gonadotrophs and its pituitary content is regulated by gonadal steroids and gonadotrophin-releasing hormone (GnRH). Hence, we reasoned that < EEP-NH2 in human serum may also arise, at least in part, from the pituitary, and that its secretion may correlate with that of gonadotrophins. Therefore, blood was simultaneously sampled from both inferior petrosal sinuses, which are major sites of the venous drainage of the pituitary gland, and a peripheral vein from seven patients with suspected adrenocorticotrophin-secreting pituitary tumours. In addition, in six postmenopausal and six cyclic women, peripheral vein blood was collected at 10-min intervals for 6 h, then a standard 100 micrograms GnRH test was performed. In the sera, TRH-LI was estimated by RIA with antiserum 4319, which binds most tripeptides that share the N- and C-terminal amino acids with TRH (pGlu-His-ProNH2). In addition, LH and FSH were measured in these sera by RIA. In the blood samples taken at 10-min intervals, an episodic variation in serum TRH-LI was noted and pulses of TRH-LI were detected at irregular intervals (from one to six pulses per 6 h) in five postmenopausal and six cyclic women. In general, these pulses did not coincide with those of LH and FSH, suggesting that TRH-LI is not co-secreted with gonadotrophins. Moreover, unlike LH and FSH, serum TRH-LI did not increase during the menopause or after exogenous administration of GnRH. Whereas gonadotrophin concentrations were significantly greater in the inferior petrosal sinus than in peripheral serum, there were no differences in TRH-LI concentrations between these serum samples. In conclusion, serum TRH-LI in humans seems not to be regulated by gonadal steroids or GnRH. Moreover, serum derived directly from the pituitary contained no more TRH-LI than did peripheral serum, which suggests that the human pituitary gland does not secrete significant amounts of < EEP-NH2, and therefore does not contribute significantly to serum TRH-LI concentrations. Further research is required to identify the site of origin of < EEP-NH2 in human serum.

Abstract

We investigated the effects of diabetes mellitus on the hypothalamo-hypophysial-thyroid axis in male (R×U) F₁ and R-Amsterdam rats, which were found to respond to streptozotocin (STZ)-induced diabetes mellitus with no or marked increases, respectively, in plasma corticosterone. Males received STZ (65 mg/kg i.v.) or vehicle, and were killed 1, 2 or 3 weeks later. At all times studied, STZ-induced diabetes mellitus resulted in reduced plasma TSH, thyroxine (T₄) and 3,5,3′-tri-iodothyronine (T₃). Since the dialyzable T₄ fraction increased after STZ, probably as a result of decreased T₄-binding prealbumin, plasma free T₄ was not altered during diabetes. In contrast, both free T₃ and its dialyzable fraction decreased during diabetes, which was associated with an increase in T₄-binding globulin. Hepatic activity of type I deiodinase decreased and T₄ UDP-glucuronyltransferase increased after STZ treatment. Thus, the lowered plasma T₃ during diabetes may be due to decreased hepatic T₄ to T₃ conversion.

Median eminence content of TRH increased after STZ, suggesting that hypothalamic TRH release is reduced during diabetes and that this is not caused by impaired synthesis or axonal transport of TRH to the median eminence. Hypothalamic proTRH mRNA did not change in diabetic (R×U) F₁ rats during the period of observation, but was lower in R-Amsterdam rats 3 weeks after STZ. Similarly, pituitary TSH and TSHβ mRNA had decreased in R-Amsterdam rats by 1 week after STZ treatment, but did not change in (R×U) F₁ rats. The difference between the responses in diabetic R-Amsterdam and (R×U) F₁ rats may be explained on the basis of plasma corticosterone levels which increased in R-Amsterdam rats only. Hypothalamic TRH content was not affected by diabetes mellitus, but the hypothalami of diabetic rats released less TRH in vitro than those of control rats. Moreover, insulin had a positive effect on TRH release in vitro.

In conclusion, the reduced hypothalamic TRH release during diabetes is probably not caused by decreases in TRH synthesis or transport to the median eminence, but seems to be due to impaired TRH release from the median eminence which may be related to the lack of insulin. Inhibition of proTRH and TSHβ gene expression in diabetic R-Amsterdam rats is not a primary event but appears to be secondary to enhanced adrenal activity in these animals during diabetes.

Journal of Endocrinology (1997) 153, 259–267

Abstract

TRH-like immunoreactivity (TRH-LI) was estimated in methanolic extracts of rat tissues and blood by RIA using antiserum 4319, which binds most peptides with the structure pGlu-X-ProNH₂, or antiserum 8880, which is specific for TRH (pGlu-His-ProNH₂). TRH-LI (determined with antiserum 4319) and TRH (determined with antiserum 8880) contents were 8 and 8 ng/g in brain, 216 and 222 ng/g in hypothalamus, 6·5 and 6 ng/g in pancreas, 163 and 116 ng/g in male pituitary, 105 and 77 ng/g in female pituitary, 1 and 0·1 ng/g in salivary gland, 61 and 42 ng/g in thyroid, 12 and 3 ng/g in adrenal, 3 and 0·3 ng/g in prostate, and 11 and 0·8 ng/g in ovary respectively. Blood TRH-LI (antiserum 4319) and TRH (antiserum 8880) levels were 31 and 18 pg/ml in male rats, and 23 and 10 pg/ml in female rats respectively. Unextracted serum obtained from blood kept for at least 1 h at room temperature no longer contained authentic TRH but still contained TRH-LI (males 20·3 ± 3·1, females 15·9 ± 3·0 pg/ml; means ± s.e.m.). Isocratic reverse-phase HPLC showed that TRH-LI in serum is largely pGlu-Glu-ProNH₂ (<EEP-NH₂), a peptide previously found in prostate and anterior pituitary.

In urine, TRH-LI (antiserum 4319) and TRH (antiserum 8880) levels were 3·21 ± 0·35 and 0·32 ± 0·04 ng/ml in male rats and 3·75 ± 0·22 and 0·37 ± 0·04 ng/ml in female rats respectively (means ± s.e.m.). Anion-exchange chromatography on QAE-Sephadex showed that urine of normally fed rats contains both basic/neutral TRH-LI (b/nTRH-LI) and acidic TRH-LI (aTRH-LI) in a ratio of ≈ 40:60, and further analysis by HPLC indicated that aTRH-LI represents <EEP-NH₂. Analysis of food extracts and urine from fasted rats demonstrated that b/nTRH-LI is derived from food particles spilled by the rats during urine collection, while aTRH-LI is endogenously produced. While urinary aTRH-LI levels were higher in female than in male rats (2·99 ± 0·41 vs 2·04 ± 0·20 ng/ml), the daily urinary excretion was similar in both sexes (females 15·6 ± 1·4, males 19·5 ± 2·0 ng/day). Intravenously injected <EEP-NH₂ disappeared from serum with a half-life of ≈ 1 h, and was recovered unchanged and quantitatively in urine. In contrast, when <EEP-NH₂ was administered with food, only ≈ 0·5% was recovered in urine. The urinary clearance rate of serum TRH-LI amounted to 0·52 ± 0·10 ml/min in males and 0·34 ± 0·05 ml/min in females.

In view of the presence of <EEP-NH₂ in the anterior pituitary gland, and the regulation of its content in parallel with gonadotrophins, we examined the possibility that serum <EEP-NH₂ is of pituitary origin and correlates with gonadotrophin secretion. However, treatments that alter pituitary <EEP-NH₂ content and gonadotrophin release had no effect on serum TRH-LI or urinary aTRH-LI.

In conclusion, the TRH-like peptide <EEP-NH₂ is present in rat serum and is excreted into the urine. Moreover, <EEP-NH₂ in serum and urine is not derived from rat food and is probably not of pituitary origin.

Journal of Endocrinology (1997) 153, 411–421