GROWTH HORMONE SECRETING PITUITARY ADENOMAS ARE HETEROGENEOUS IN CELL CULTURE AND COMMONLY SECRETE GLYCOPROTEIN HORMONE α-SUBUNIT

Cell culture methods were used to assess whether human pituitary adenomas secreting GH and associated with clinical acromegaly also secreted the structurally unrelated glycoprotein hormone alpha-subunit. Thirty-two tumours, together with peri-adenomatous tissue from two of them and three normal pituitaries were studied. Anterior pituitary hormones were measured by radioimmunoassay and included PRL, TSH, LH, FSH, and ACTH, as well as GH and alpha-subunit. Normal pituitary tissues secreted all hormones assayed. All 32 tumours secreted GH ranging from 241 to 5556 ng/2 X 10(5) cells/24 h and 12 (37.5%) secreted alpha-subunit in amounts which could not be accounted for by cross-reaction of other hormones or contamination by normal pituitary tissue, and which ranged from 10.3 to 73.5 ng/2 X 10(5) cells/24 h. Ten other tumours also secreted alpha-subunit but in very small amounts, not exceeding 1.8 ng/2 X 10(5) cells/24 h. PRL was secreted from 21 tumours (66%), and small amounts of other hormones, chiefly LH and TSH, were occasionally secreted from tumours. These cell culture studies would suggest that pituitary adenomas causing acromegaly are hormonally heterogeneous and that PRL and glycoprotein alpha-subunit are commonly detected in addition to GH.     

Effect of hypothyroidism and thyroid hormone replacement in vivo on pituitary cytoplasmic concentrations of thyrotropin-beta and alpha-subunit messenger ribonucleic acids

We have studied the influence of hypothyroidism and thyroid hormone replacement in vivo on rat pituitary cytoplasmic concentrations of TSH beta and alpha-subunit mRNA, measured by cytoplasmic dot hybridization, as well as circulating TSH and pituitary TSH content. Cytoplasmic concentrations of GH, PRL, and LH beta-subunit mRNA were measured in parallel. Marked increases in serum TSH and TSH beta and alpha mRNA were found in hypothyroidism; these changes reversed 72 h after beginning T3 replacement. A more marked effect of hypothyroidism on TSH beta mRNA than on alpha mRNA was evident, suggesting differential control of regulation of the TSH genes. In contrast to a rapid fall in serum TSH after T3 administration, an increase in TSH beta and alpha mRNA was evident at 1 and 6 h, followed by a later fall in TSH mRNA to euthyroid values. Pituitary cytoplasmic concentrations of GH and PRL mRNA were reduced in hypothyroidism and increased after administration of T3; LH beta mRNA was unaffected by hypothyroidism or T3 replacement. The differential effects of thyroid status on TSH, GH, PRL, and LH mRNA indicate specificity of regulation of these anterior pituitary hormone genes.     

Effects of a novel hypothalamic peptide, pituitary adenylate cyclase-activating polypeptide, on pituitary hormone release in rats.

We have demonstrated that the novel hypothalamic peptide pituitary adenylate cyclase-activating polypeptide (PACAP-38; 0.1-100 nmol/l) caused an increase in the release of GH, ACTH, LH and alpha-subunit and accumulation of intracellular cyclic AMP from dispersed rat anterior pituitary cells in static culture for 24 h. There were no significant effects on TSH or prolactin release over the same time-period. PACAP-38 (10 nmol/l) increased the release of GH by 1.3-fold (P less than 0.05), ACTH by 1.9-fold (P less than 0.05), LH by 3.5-fold (P less than 0.001) and alpha-subunit by 2.0-fold (P less than 0.005) and the accumulation of intracellular cyclic AMP by greater than 2-fold (P less than 0.001) after 24 h. However, the time-course for the effect of PACAP-38 (1 mmol/l) on hormone release and intracellular cyclic AMP levels showed a temporal dissociation. The effect of PACAP-38 on GH and ACTH levels did not reach significance until 24 h whereas the effect of PACAP-38 on LH and alpha-subunit release reached significance after 4 h implying a different mechanism of action for their release. To investigate the PACAP-induced secretion of LH and alpha-subunit further, we examined the effects of PACAP after down-regulation of protein kinase C (PKC). PACAP-38 at a dose maximal for the stimulation of LH and alpha-subunit release (10 nmol/l) added together with the PKC activator, 12-O-tetradecanoyl-phorbol-13-acetate (TPA; 0.1 mumol/l) had no greater effect on LH and alpha-subunit release than TPA alone over a 4 h incubation period.(ABSTRACT TRUNCATED AT 250 WORDS)     

Combined pituitary stimulation test: interactions of hypothalamic releasing hormones in man

The use of hypothalamic releasing hormones for the clinical assessment of anterior pituitary function is both simple and free of severe side effects. Tests with the recently discovered substances GRF and CRF as well as with combinations of several releasing hormones are therefore used in many clinics. A reliable interpretation of such combined tests, however, is only possible when positive or negative interactions between these releasing hormones are known. After a rest of 2 h to reach basal cortisol levels, 7 groups of 5 male volunteers each received an iv bolus injection consisting of either: A): GRF (100 micrograms) + CRF (50 micrograms) + TRH (200 micrograms) + LHRH (100 micrograms), B): CRF + TRH, C): GRF + TRH, D): LHRH + TRH, E): TRH, F): GRF, G): CRF. During the following 2 h, GH, TSH, cortisol, LH, FSH and prolactin were measured every 15 min. The TSH response after the injection of all 4 releasing hormones was significantly higher (delta TSH = 16.5 +/- 2.0 microU/ml, x +/- SE) compared to the injection of TRH alone (delta TSH = 9.3 +/- 1.4 microU/ml; p less than 0.025). This increment in TSH secretion was confirmed when 2 groups of 5 female volunteers were studied with the TRH-test (delta TSH = 9.9 +/- 1.8 microU/ml) or the combination of all four releasing hormones (delta TSH = 16.8 +/- 2.9 microU/ml; p less than 0.05). This exaggerated TSH-response to TRH was demonstrated to be entirely due to simultaneous administration of GRF, whereas CRF and LHRH in combination with TRH had no additional effect on TSH release.(ABSTRACT TRUNCATED AT 250 WORDS)     

Effects of intraventricular growth hormone-releasing factor on growth hormone release: further evidence for ultrashort loop feedback

We examined the effects of cerebroventricular injection of synthetic human GH-releasing factor [hGRF-(1-44)] on regulation of GH release in conscious male rats. These results were compared with the direct effects of hGRF on hormone released from dispersed anterior pituitary cells. Administration of two higher doses of hGRF (200 and 2000 ng) into the third ventricle (3V) produced a dose-related increase in plasma GH levels (P less than 0.001). Injection of hGRF into the 3V at two lower doses actually reduced GH release. Infusion of 20 ng (5 pmol) hGRF reduced plasma GH from 5-60 min (P less than 0.005), with a maximum suppression of 66%. The 2-ng (0.5-pmol) dose decreased GH secretion by 45% (P less than 0.05). hGRF stimulated a significant and dose-dependent release of GH from dispersed pituitary cells at concentrations of 10(-10) and 10(-9) M (P less than 0.025). The specificity of GRF for GH control, whether stimulatory or inhibitory, was seen by the failure of GRF to modify PRL, TSH, or LH release. Our results indicate that injection of larger doses of GRF into the 3V produce GH release, but at lower doses, 3V GRF may exert an action centrally to inhibit GH release. We propose that hypothalamic GRF may decrease its own neurosecretion by negative ultrashort loop feedback.     

Regulation of growth hormone (GH) secretion by GH-releasing factor, somatostatin, and insulin-like growth factor I in ovine fetal and neonatal pituitary cells in vitro

Serum GH concentrations in the ovine fetus are much higher than those in the neonate, and the maximal GH response induced by GRF is 5-fold greater in the fetus than in the neonate. To clarify these in vivo observations further, we studied the effects of GRF, somatostatin (SRIF), and insulin-like growth factor I (IGF-I) on primary cultures of fetal and neonatal ovine pituitary cells. GH secretion from fetal ovine pituitary cells increased from 148 +/- 34 to 950 +/- 130 ng/10(5) cells.3 h in response to 1 nM GRF, whereas GH secretion from neonatal pituitary cells rose from 113 +/- 26 to 1221 +/- 129 ng/10(5) cells.3 h, a significantly greater response (P less than 0.001). This greater GRF-induced GH response in neonatal than fetal cells differs from the response in vivo and suggests that the increased in vivo response in the fetus is not due to inherently increased sensitivity of pituitary cells to GRF. SRIF (10 nM) decreased maximal GRF-induced GH secretion by 37 +/- 3% in fetal cells compared with 59 +/- 8% in neonatal cells (P less than 0.01). This may explain in part the decreased in vivo sensitivity to SRIF in the ovine fetus compared to that in the neonatal lamb. In fetal pituitary cells, 10 nM GRF increased ovine (o) GH mRNA from 100 +/- 14% to 145 +/- 40%, SRIF decreased oGH mRNA to 84 +/- 3%, and GRF and SRIF in combination increased fetal oGH mRNA to 126 +/- 24%. Values in neonatal pituitary cell cultures were similar (control, 100 +/- 17%; GRF, 132 +/- 6%, SRIF, 85 +/- 15%; GRF plus SRIF, 105 +/- 26%). Pretreating fetal cells with 100 nM IGF-I for 3 days reduced GRF-stimulated GH secretion from 1049 +/- 38 to 232 +/- 8 ng/10(5) cells.3 h (P less than 0.001). Similarly, IGF-I pretreatment of neonatal cells reduced GRF-stimulated GH secretion from 810 +/- 18 to 419 +/- 16 ng/10(5) cells.3 h (P less than 0.001). The mean secreted IGF-I was 0.58 U/ml (36 nM) in culture medium from neonatal cells and was unchanged by incubation for 3 days with 5 micrograms/ml hGH. Secreted IGF-I in medium from fetal cells was 0.87 U/ml (54 nM) without GH and 0.81 U/ml (51 nM) after incubation with human GH. IGF-I mRNA was present in neonatal pituitary and brain.(ABSTRACT TRUNCATED AT 400 WORDS)     

Endocrine, biochemical, and morphological studies of a pituitary adenoma secreting growth hormone, thyrotropin (TSH), and alpha-subunit: evidence for secretion of TSH with increased bioactivity

A 40-yr-old man who had acromegaly and hyperthyroidism due to a GH/TSH-secreting pituitary adenoma is described. Serum free T4 was 2.8 ng/dl, free T3 was 1.1 ng/dl, and TSH was 1.2-1.5 microU/ml; the latter was measured in an immunoradiometric assay with a sensitivity of 0.07 microU/ml. Serum TSH was immunologically identical to standard TSH and did not decrease during a T3 suppression test. Serum free alpha-subunit and the molar alpha-subunit to TSH ratio were high (6.1 ng/ml and 31.2, respectively). TRH administration induced significant increases in both GH (+129%) and alpha-subunit (+156%) levels. Conversely, dopamine infusion resulted in a decrease in serum GH (-66%) and alpha-subunit (-43%) levels, and subsequent administration of the dopamine antagonist sulpiride induced significant increases in both GH and alpha-subunit (+393% and +106%, respectively). Similarly, somatostatin infusion inhibited GH (-43%) and alpha-subunit (-61%) secretion. Serum TSH levels were not affected by TRH, dopamine, or somatostatin. The biological to immunological activity ratio of serum TSH purified by immunoaffinity chromatography and measured in an adenylate cyclase assay was significantly increased compared to that in serum from hypothyroid or euthyroid subjects [biological to immunological activity ratio, 6.9 +/- 0.2 (+/- SD) vs. 4.4 +/- 1.1; P less than 0.001]. In gel chromatography, the apparent mol wt of the patient's TSH was smaller than that of the controls. After adenomectomy, all of the altered parameters of pituitary function became normal. Double gold particle immunostaining of the adenomatous tissue showed that all of the cells contained secretory granules positive for GH and alpha-subunit, while very few cells were positive for TSH beta as well as GH and alpha-subunit. These data indicate that in this patient serum TSH had an apparent mol wt smaller than that of normal TSH and an increased biological activity which, along with the autonomous TSH secretion, account for hyperthyroidism in the presence of low normal TSH levels; alpha-subunit originated from the same adenomatous cells that secreted GH but not TSH, thus explaining the in vivo observation that alpha-subunit responses to several agents were dissociated from TSH responses and parallel to GH responses; and TSH and GH were colocalized in a minority of the neoplastic cells.     

Effect of clonidine on growth hormone, prolactin, luteinizing hormone, follicle-stimulating hormone, and thyroid-stimulating hormone in the serum of normal men.

Clonidine (0.15 mg iv), a selective noradrenergic receptor agonist, increased serum growth hormone (GH) levels (greater than 6 ng/ml) on 8 out of 12 administrations to 6 normal men. This increase was independent of the hypotensive effects of the drug and unrelated to changes in serum cortisol. Clonidine induced a hyperglycemic effect in all subjects which was greatest 15 min after commencint the injection. No changes in blood sugar or GH occurred after placebo injection. Apomorphine, a selective dopamine receptor agonist, elevated GH in each of these 6 subjects (greater than 10 ng/ml). Clonidine had no effect on serum prolactin (PRL), luteinizing hormone (LH), follicle-stimulating hormone (FSH), or thyroid-stimulating hormone (TSH). These data are compatible with a dual dopaminergic and noradrenergic mechanism modulating GH secretion in normal men and with the absence of a noradrenergic mechanism in the regulation of PRL, LH, FSH, or TSH.     

Differential effects of opiate peptides and alkaloids on anterior pituitary hormone secretion

In order to investigate the opiate receptors involved in the control of anterior pituitary hormone secretion, five different opioid drugs were administered intravenously to groups of 6 normal male subjects. Morphine (10 mg), methadone (10 mg), pentazocine (30 mg), nalorphine (10 mg) and 0.25 mg of the met-enkephalin analogue, DAMME, all caused similar increases in circulating prolactin with falls in serum LH and cortisol. Methadone and DAMME also elevated GH and TSH; morphine elevated TSH but not GH, nalorphine GH but not TSH. After pentazocine neither GH nor TSH changed. FSH failed to change significantly after any drug. All these changes, except serum cortisol, were antagonised by 4 mg naloxone. Taking into account the known receptor subtypes preferentially activated by each opiate, it is suggested that prolactin secretion is modulated by epsilon-receptors and TSH by mu-receptors. The control of ACTH probably involves delta-or kappa-receptors, that for LH kappa-or epsilon-receptors. It is not possible on present data to allocate a specific receptor mediating the opioid control of GH.     

Abnormal hypothalamic-pituitary function in polyostotic fibrous dysplasia

Hypothalamic-pituitary dysfunction was found in two patients with polyostotic fibrous dysplasia and leontiasis ossea. Both patients probably had McCune--Albright syndrome with early sexual development, disfiguring craniofacial bone lesions, cutaneous hyperpigmentation and gigantism in childhood. Endocrinological studies revealed measurable plasma growth hormone throughout a 24-h sampling period with preservation of sleep augmented rises. The mean 24-h values were 329 and 7 ng/ml, respectively. Both their mean plasma prolactin concentrations, 385 and 45 ng/ml, and the 24-h secretory patterns of prolactin were abnormal. One patient had an elevated mean plasma cortisol concentration of 8.5 micrograms/dl but the 24-h pattern of cortisol secretion was normal. Basal plasma LH, FSH, TSH, T4 and T3 concentrations were normal but neither patient had a TSH response to TRH. Prolactin and growth hormone secretions were, however, increased after TRH. Oral glucose resulted in partial suppression of GH but an exaggerated insulin response in both patients. Post-mortem in one patient revealed a thick calvarium with bony encasement of the pituitary gland. The pituitary and hypothalamus appeared normal on gross, light microscopic and electronmicroscopic examination. These data strongly suggest the presence of hypothalamic-pituitary dysfunction in these two patients.     

Evidence of a circulating growth hormone stimulating factor other than growth hormone releasing hormone in a patient with pituitary tumour and acromegaly

This study is a report on the growth hormone (GH) stimulatory effect of serum and plasma from a patient with notably active acromegaly due to a GH producing pituitary adenoma. Pituitary adenomatous tissue from 7 patients with GH producing adenomas, one with a prolactin (Prl) producing adenoma, one with a TSH producing adenoma, and one with a non-secreting adenoma, were cultured in vitro for 8-10 days. Media were changed every 48-72 h and contained Neumann Tytell buffer with the addition of 1) foetal calf serum, 2) patients' own serum or plasma, 3) serum or plasma from the patient with notably active acromegaly. GH release expressed as microgram GH/1/48-72 h between day 6 and 8 in culture did not differ when adenomatous tissue was cultured in buffer, foetal calf serum or the patients' own serum or plasma. In contrast, GH release was increased in 9/10 patients, when media contained serum or plasma from the patient with notably active acromegaly. This GH stimulatory effect was demonstrated in vitro in human pituitary adenomatous tissue from patients with pathological as well as normal GH secretion in vivo. Furthermore, this GH releasing plasma in a concentration of 10% increased GH release in cultures of dispersed rat anterior pituitary cells. In the same system, synthetic growth hormone-releasing hormone (GRF)-44 stimulated the release of GH in a dose-dependent manner. However, at all dose levels including maximally stimulating doses of GRF, an additive effect on GH release was seen with 10% of the GH releasing plasma.(ABSTRACT TRUNCATED AT 250 WORDS)     

Effects of a growth hormone releasing factor in man

The effects of the 44-amino acid growth hormone releasing factor (GRF-44) were tested in normal adult men and women. At a dose of 1 microgram/kg, intravenous boluses of GRF-44 stimulated prompt elevations of plasma GH, which in 5 men reached maximum levels of 34 +/- 28 (S.D.) ng/ml, and in 3 women in the mid-follicular phase, 53 +/- 10 ng/ml. The action of GRF was highly selective; there were no changes in plasma PRL, LH, FSH, TSH, or cortisol at this dose level. Side effects, mostly flushing and a sense of warmth of the face and chest, were mild and occurred only in a minority of subjects.     

Diminished pituitary responsiveness to growth hormone-releasing factor in aging male rats

The pattern of GH secretion undergoes substantial changes in the aging rat, resulting in decreased daily secretion of GH. In this study, the pituitary responsiveness to GH-releasing factor (GRF) was examined in young (2- to 5-month old) and aging (14- to 18-month old) male rats. In vivo studies were performed under sodium pentobarbital anesthesia. After injection of 250 ng GRF/100 g BW, young rats experienced more GH secretion [peak level, 544.5 +/- 209.5 (+/- SEM) ng/ml] than did 18-month-old rats (89.3 +/- 13.7 ng/ml). To investigate the locus of this insensitivity to GRF, anterior pituitary cells from young and aging rats were dispersed and placed in primary culture. While basal GH secretion from the cultured pituitary cells was similar in the two groups (49.7 +/- 3.5 vs. 47.8 +/- 2.7 ng/ml X 4 h for the 2- and 18-month old rats, respectively), the GH-releasing ability of GRF was partially but significantly impaired in cells derived from both 14- and 18-month old rats; 100 nM GRF stimulated the release of 96.7 +/- 1.8 ng/ml X 4 h in the 18-month old rats as opposed to 115.0 +/- 6.0 (P less than 0.05) ng/ml X 4 h in the 2-month-old rats. Since GRF stimulates GH release through the activation of adenylate cyclase, intracellular cAMP levels were measured in the cultured pituitary cells. GRF stimulated 65% less intracellular cAMP accumulation in the 18-month-old rats. In 14-month-old rats, the ability of forskolin and (Bu)2 cAMP to release GH was impaired, while phorbol ester-elicited GH secretion was unchanged. In conclusion, the GH response to GRF is blunted in aging rats. While much of the insensitivity to GRF may be mediated by the increased somatostatin tone reported in aging rats, a diminished pituitary cAMP response to GRF may also be an important etiological factor in the hyposomatotropinemia of the aging male rat.     

The effect of altered thyroid status on pituitary hormone messenger ribonucleic acid concentrations in the rat

To study the effects of altered thyroid status on pretranslational control of pituitary hormones, adult male rats were given propylthiouracil for 6 weeks and underwent the following studies. 1) Rats were injected with T3 at 10 micrograms/100 g BW daily for 10 days. 2) Rats were given T3 injections at 0, 0.01, 0.1, 1.0, or 10 micrograms/100 g BW for 10 days. 3) Rats were killed 0, 1, 6, or 24 h after a single injection of T3 at 10 micrograms/100 g BW or after 5 or 10 days of daily T3 injections. Pituitary mRNA concentrations of TSH beta, alpha-subunit, PRL, GH, POMC, FSH beta, and LH beta were determined for individual animals. Marked increases in TSH beta and alpha-subunit mRNAs occurred after PTU treatment, and these changes were reversed by 1.0 microgram/100 g BW T3 and within 24 h of a single T3 injection of 10 micrograms/100 g BW. Further increases in the dose or time course of T3 administration led to a relatively greater suppression of TSH beta mRNA levels than alpha-subunit mRNA levels. In contrast, GH and PRL mRNA levels were low in hypothyroid animals, and both rose toward control levels with 0.1 microgram/100 g BW T3 and by 24 h after a single T3 dose. Induction of hyperthyroidism did not further increase GH mRNA levels above control, but increased PRL mRNA levels 2-fold over control. No changes were seen in FSH beta, LH beta, or POMC mRNA levels with any treatment. Thus, studies of altered thyroid status in the rat reveal dose-response and time-course variability in the pretranslational control of TSH beta, alpha-subunit, GH, and PRL by thyroid hormone.     

Endocrine and immunohistochemical studies on thyrotropin (TSH)-secreting pituitary adenomas: responses of TSH, alpha-subunit, and growth hormone to hypothalamic releasing hormones and their distribution in adenoma cells.

Endocrine and immunohistochemical studies were performed in two cases of TSH-secreting pituitary adenomas. The patients had elevated serum TSH and alpha-subunit concentrations despite high serum thyroid hormone levels. In addition, one patient (no. 1) had elevated serum GH levels with clinical evidence of acromegaly. GH-releasing hormone infusion increased serum levels of TSH, alpha-subunit and GH in the two patients. TRH injection increased serum TSH levels in both patients and, concomitantly, serum alpha-subunit and GH levels in patient 1. Basal TSH levels and their responses to TRH changed reciprocally to changes in serum thyroid hormone levels, although TRH-induced GH release did not. The administration of GnRH also increased serum TSH, alpha-subunit, and GH levels in patient 1. In accordance with these in vivo results, pituitary adenoma cells in culture obtained from patient 1 responded to GH-releasing hormone, TRH, or GnRH to secrete TSH, alpha-subunit, and GH. Incubation of cells with dexamethasone resulted in inhibition of TSH and stimulation of GH secretion without a significant change in alpha-subunit secretion. On the basis of light microscopic and electron microscopic double gold immunohistochemistry, the tumor from patient 1 was a bimorphous adenoma composed of two separate cell types: cells with TSH beta-subunit (TSH beta) and alpha-subunit, and those with GH and alpha-subunit. The remainder consisted mainly of cells with TSH beta and alpha-subunit. The coproduction of the unusual combination of two hormones such as GH and alpha-subunit in a single-type of adenoma cell and the coexistence of thyrotrophs and somatotrophs in one pituitary adenoma along with the aberrant responses of TSH beta, alpha-subunit, and GH to multiple hypothalamic hormones suggest the dedifferentiation of pituitary cells to multipotential progenitor cells by neoplastic transformation.     

EFFECTS OF GRF(1–40) AND DOMPERIDONE ON GH SECRETION IN NORMAL MAN

In eight normal adult men pituitary secretion following GRF(1-40) was studied. GRF administration (50 micrograms i.v.) was followed by an increase in GH release with a peak value between the 15 and 60 min. No effects were noticed on LH, FSH, PRL, TSH and ACTH secretion. GH and PRL release was also studied after domperidone (DOM) (5 mg i.v./h), and GRF plus DOM. PRL increased significantly after DOM and GRF plus DOM. During GRF plus DOM a more marked GH release was observed in comparison with the hormone response to GRF alone at 15-45 and 120 min (P less than 0.05). This phenomenon was found in in six out of eight subjects studied. Mean peak and secretory area was greater (P less than 0.05) after GRF plus DOM than after GRF alone. These data suggest that GRF(1-40) at the dose used is a useful tool in the study of GH secretion. The GH pattern during GRF plus DOM seems to indicate that dopaminergic tone may play a direct inhibitory role on GH secretion in man.     

Differential appearance of the subunits of glycoprotein hormones (LH, FSH, and TSH) in the pituitary of bullfrog (Rana catesbeiana) larvae during metamorphosis.

The ontogeny of gonadotrophs and thyrotrophs in the bullfrog pituitary was examined immunohistochemically using monoclonal antibodies against bullfrog lutropin beta-subunit (LH beta), follitropin beta-subunit (FSH beta) and its alpha-subunit, and polyclonal anti-human thyrotropin beta-subunit (TSH beta) serum. Immunoreactive alpha-subunit- and TSH beta-, but not FSH beta- and LH beta-containing cells were observed at embryonic stage 24 (Shumway's classification). Immunoreactive FSH beta cells first appeared at Taylor-Korllos stage V, and immunoreactive LH beta cells at stage X. Throughout metamorphosis, several gonadotrophs containing both FSH and LH were found in the ventrocaudal region, but most gonadotrophs contained only FSH. Immunoreactive alpha-subunit cells were always more frequent than the sum of immunoreactive beta-subunit cells, which was confirmed by quantitative studies using immunohistochemical and RIA techniques.