Carol J. Phelps

Pituitary hormones as neurotrophic signals: anomalous hypophysiotrophic neuron differentiation in hypopituitary dwarf mice.

Abstract Anterior pituitary hormones are known to exert dynamic negative feedback effects on their respective regulatory (“hypophysiotropic”) neurons in the hypothalamus. The purpose of this review is to present the evidence for a theory that the effect of pituitary hormones on these hypophysiotropic neurons is neurotrophic, extending beyond dynamic feedback to influence upon cell survival, phenotypic differentiation, and axonal connectivity. To that end, the adult condition and the development of hypophysiotropic neurons in mutant mice which lack pituitary growth hormone (GH) and prolactin (PRL) are presented as models of the effect of absent specific neurotrophic signals. The expression of the neurohormones which inhibit PRL and GH secretion, dopamine (DA) and somatostatin, respectively, is markedly reduced in the hypothalamus of the hypopituitary dwarf mouse, and this adult condition is the result of postnatal failure to develop or actual regression, which may include neuronal cell death. The deficit in DA may be reversed by PRL replacement, but only if initiated at an identified critical postnatal period. Conversely, expression of the stimulatory GH-releasing hormone (GHRH) is markedly increased in the dwarf mouse hypothalamus. The loss of DA and the increase in GHRH occur in the same hypothalamic area, suggesting neuronal phenotypic plasticity in response to absence of pituitary feedback signals. The axonal terminations of extant GH- and PRL-regulating neurons in external median eminence appear to be reduced, suggesting that pituitary signals are required for appropriate axonal guidance during development, even though an endocrine vascular route intervenes between these regulatory neurons and their target secretory cells. The collective observations indicate that GH and PRL may be regarded as neurotrophic factors for their respective regulatory neurons in the hypothalamus.

Prolactin Gene Disruption Does Not Compromise Differentiation of Tuberoinfundibular Dopaminergic Neurons

In mice with spontaneous mutations in transcription factors Prop-1 or Pit-1, the pituitary fails to produce prolactin (PRL), GH and TSH, and numbers of hypothalamic PRL-regulating dopaminergic (DA) neurons (area A12) are reduced by more than 50%. A normal neuronal population can be maintained in these mutants by PRL treatment of neonates, but not of adults. Targeted disruption of the PRL structural gene in mice provides a new model of isolated PRL deficiency to test the specificity of the PRL neurotrophic effect. The present study used morphological methods to assess hypophysiotropic tuberoinfundibular dopaminergic (TIDA) neurons in these mice, with the hypothesis that isolated PRL deficiency also would lead to reduction in TIDA neuron number. Brains of female and male homozygous PRL-null (–/–) mice and normal heterozygous (+/–) siblings were compared using formaldehyde-induced endogenous catecholamine fluorescence and tyrosine hydroxylase (TH) immunocytochemistry. Immunostaining intensity was quantified using computerized image analysis, and total numbers of TH-immunoreactive neurons were counted in three diencephalic DA brain regions. Intensity of DA fluorescence in A12 perikarya and median eminence (ME) was reduced in –/– mice; fluorescence in other brain areas was comparable for –/– and +/– mice. Immunostaining intensity of TH was significantly lower (p = 0.0001) in –/– than in normal mice in perikarya of A12, but not in cell bodies of nonhypophysiotropic area A13 (medial zona incerta). In external ME, TH immunostaining intensity was lower (p = 0.0001) in PRL-null than in normal mice. The decrease in TH intensity in both perikarya and in ME was significant for both female and male –/– mice. However, numbers of A12 neurons in the PRL-null mice were not lower than those of normal siblings. TH-immunoreactive cell number also did not differ between +/– and –/– mice in areas A13 and periventricular A14. The presence of a normal complement of A12 DA neurons in the PRL-null mice, despite greatly reduced DA and TH, emphasizes that steady-state content and differentiation of phenotype in individual neurons are very different assessments. The results suggest that, although absence of the stimulatory PRL feedback signal results in diminished activity of TIDA neurons, differentiation of these cells is not adversely affected.

Growth Hormone-Releasing Hormone-Producing and Dopaminergic Neurones in the Mouse Arcuate Nucleus Are Independently Regulated Populations

Differentiation of hypophysiotropic neurones that regulate the secretion of growth hormone (GH) and prolactin is influenced by GH and prolactin. Genetic GH and prolactin deficiency in mutant rodent models such as the Ames dwarf (df/df) mouse results in an increase in the number of GH‐stimulatory GH‐releasing hormone (GHRH) neurones and a reduction of prolactin‐inhibitory tuberoinfundibular dopaminergic (TIDA) neurones in the arcuate nucleus during postnatal development. The present study tested the hypothesis that these concomitant changes in numbers of tyrosine hydroxylase (TH)‐ and GHRH‐immunoreactive neurones in df/df hypothalamus might represent a neuronal population of fixed number that undergoes a partial change in phenotype during postnatal development. To evaluate this possibility, the postnatal reduction of the df/df TIDA population was prevented by administering prolactin neonatally to preserve TH phenotype; dwarf and normal sibling mice were treated with daily injections of ovine prolactin or vehicle starting at postnatal day 12 and continuing for 30 days. Following this treatment, numbers of arcuate neurones containing GHRH or TH, or both, were quantified using immunocytochemistry. It was hypothesized that prolactin preservation of TH‐immunoreactive cell number would be accompanied by either a decrease in the GHRH‐producing population or an increase in numbers of cells producing both TH and GHRH. In prolactin‐treated normal (DF/df) mice, numbers of arcuate TH‐immunoreactive neurones were similar to those in vehicle‐treated normals. Numbers of TH‐positive neurones in prolactin‐treated dwarfs were higher than in vehicle‐treated dwarfs, and did not differ from numbers in DF/df. Numbers of GHRH‐immunoreactive cells in vehicle‐treated df/df were higher than in vehicle‐treated DF/df, and were not different in prolactin‐treated groups of either dwarf or normal mice. Neurones containing both TH and GHRH constituted 15% of the TH population, and 76% of the GHRH population, in control normal mice; in control dwarfs, double‐labelled cells were 9.3% of TH and 9.9% of GHRH. Numbers of cells immunoreactive for both TH and GHRH were not affected by prolactin treatment in either mouse type. These results demonstrate that the increase in number of GHRH‐expressing neurones in the df/df arcuate nucleus does not occur at the expense of the TH phenotype, and that this increase is not influenced by prolactin feedback. Although coexpression of TH and GHRH in a subpopulation indicates that TIDA and GHRH populations are not exclusive, they appear to be influenced independently by prolactin and GH signals during development.

Identification of Growth Hormone-Releasing Hormone and Somatostatin Neurons Projecting to the Median Eminence in Normal and Growth Hormone-Deficient Ames Dwarf Mice

In the spontaneous mutant Ames dwarf mouse, GH deficiency coincides with a dramatic increase in the expression of both mRNA and peptide for stimulatory GHRH and reduced expression of GH-inhibitory somatostatin (SRIH) mRNA and peptide. However, both GHRH and SRIH are markedly reduced in the dwarf median eminence (ME), suggesting that ME innervation by GHRH and SRIH neurons may be aberrant in the absence of GH. In order to test this hypothesis, the number of GHRH and SRIH ME-projecting neurons was evaluated in normal and dwarf mice using a combination of retrograde tract-tracing and neuron phenotype identification by immunocytochemistry (ICC). Adult animals were injected intraperitoneally with the fluorescent tract-tracer fluorogold (FG), which, in the brain, is taken up only by axons terminating in areas deprived of the blood-brain barrier, such as the ME. Visualization of FG was achieved by either UV illumination or ICC, and was combined as appropriate with fluorescence or bright-field ICC for GHRH or SRIH. Cells immunoreactive for GHRH or SRIH and labeled with FG were quantified at each 180-microns rostral-to-caudal level through the hypothalamus. As reported previously, the total number of hypophysiotropic GHRH neurons was markedly increased in dwarf compared with that in normal mice. However, a similar percentage of ME-innervating GHRH cells was estimated in dwarf (73 +/- 4%) and normal (76 +/- 3%) animals. Such a percentage in dwarfs thus represents a larger population of ME-projecting GHRH cells (749 +/- 53) than in normal mice (128 +/- 15). Increased numbers of FG-labeled GHRH neurons in dwarfs were located at the middle and posterior levels of the arcuate nucleus (2.08, 2.26 and 2.44 mm posterior to bregma). The percentage of FG-labeled SRIH neurons was also similar for dwarf (83 +/- 2%) and normal (87 +/- 2%) mice. Because the total SRIH-immunoreactive neuronal population in dwarfs is significantly reduced compared to that in normal animals, the similar FG-labeled percentage reflects a reduced number of SRIH cells projecting to ME in dwarf (1,376 +/- 104) compared with normal (3,192 +/- 267) mice. Fewer FG-labeled SRIH cells were found in dwarfs at every anterior-to-posterior level of the periventricular nucleus (p < 0.01 for comparisons at 0.28, 0.46, 0.64, and 1.0, and p < 0.05 for comparison at 1.18 mm posterior to the bregma). The present study indicates that the reduction in GHRH and SRIH immunoreactivity in the dwarf ME may result from different phenomena for each neuronal population. The reduction in GHRH immunostaining in the ME, despite a marked increase in the total ME-projecting GHRH neurons, may be interpreted as increased GHRH release, with consequent depletion of the ME stores. In contrast, the deficit in ME SRIH may be proportional to the deficit in the number of detectable SRIH periventricular nucleus neurons.

Hypothalamic Dopaminergic Neurons in Prolactin‐Deficient Ames Dwarf Mice: Localization and Quantification of Deficit by Tyrosine Hydroxylase Immunocytochemistry

Hypothalamic tuberoinfundibular prolactin‐inhibiting neurons show decreased levels and synthesis of dopamine in two types of genetically prolactin‐deficient dwarf mice (Snell, Ames) which arise from separate mutations. A reduction to 2% of normal in this neuronal population has been quantified for Snell dwarfs. The present study was undertaken in order to quantify morphometrically the deficit and its distribution in Ames dwarf mice, including comparisons of sex and adult age. The brains of dwarf (df/df) and normal phenotypic (DF/?) sibling mice of both sexes from 4 to 16 months of age were immunostained for tyrosine hydroxylase, the rate‐limiting enzyme in dopamine synthesis; neuronal perikarya were counted in coronal sections of tuberoinfundibular arcuate nucleus (area A12), medial zona incerta (A13) and anterior periventricular (A14) hypothalamic areas at 180 μm rostral‐to‐caudal intervals. Normal (DF/?) mice exhibited no differences in neuron numbers, with regard to age or sex, in any of the three dopaminergic areas. In dwarf mice, a tendency toward decreased neuron numbers with age was statistically significant for area A14 only, and the size of the neuronal population in A12 was reduced in males compared with females. Total A12 neuron number in dwarfs was 48% of that in normal mice (P< 0.001). Periventricular (A14) perikaryal numbers were reduced slightly (P<0.05) in dwarfs compared with normals. Numbers of A13 neurons were comparable for DF/? and df/df. The morphometric distribution of tyrosine hydroxylase‐immunoreactive neurons in A12 showed that the decrease in neuronal number in dwarfs was distributed throughout the rostral‐to‐caudal length of the nucleus, with significant decrease of total perikarya (P<0.05) at each 180 μm sampling interval. Thus, lifelong absence of prolactin in Ames dwarf mice is accompanied by a significant decrease in hypothalamic tyrosine hydroxylase‐immunoreactive neurons, which is restricted to hypophysiotropic areas, uniformly distributed within A12, and is less severe than the reduction in the phenotypically similar Snell dwarf.

Prolactin replacement must be continuous and initiated prior to 21 d of age to maintain hypothalamic dopaminergic neurons in hypopituitary mice

The prolactin (PRL) deficit in mice homozygous for the spontaneous Ames dwarf (df) mutation coincides with a marked reduction in the number of PRL-regulating tuberoinfundibular dopaminergic (TIDA) neurons. The TIDA deficit develops after 14–21 d postnatally and may be prevented by PRL replacement initiated at 12, but not at 60, d of age. The present study was designed to define further the developmental period during which PRL can prevent the deficit in the number of TIDA neurons in df/df mice, as well as to evaluate whether exposure to PRL neonatally affects the response to PRL by TIDA neurons in later development. To address the first aim, litters of df/df and normal (DF/df) mice were treated daily with ovine PRL (50 µg intraperitoneally), starting at 12, 21, or 30 d of age. To address the second aim, DF/df and df/df animals treated with PRL for 30 d starting at 12 d of age were subjected to PRL withdrawal (15 d of saline vehicle treatment), followed by PRL retreatment. All brains were evaluated using both catecholamine histofluorescence and tyrosine hydroxylase (TH) immunocytochemistry. Total numbers of TH-immunostained cells were counted in area A12 (TIDA neurons) and in A13 (medial zona incerta). Qualitatively, catecholamine fluorescence in A12 perikarya and terminals in df/df mice was enhanced by PRL treatment initiated at 12 or 21, but not at 30, d of age. TH immunostaining intensity was enhanced in all df/df PRL-treated groups, compared with saline treatment. However, total numbers of TH-positive TIDA neurons were reduced significantly in df/df mice treated with PRL beginning at 21 or 30 d, as well as with saline at 12 d, compared with similarly treated DF/df groups and with df/df animals treated with PRL beginning at 12 d (p<0.01 for all comparisons). Among dwarf mice treated with PRL beginning at 12 d, followed by PRL withdrawal, the numbers of TH-positive TIDA neurons were greater than those of saline-treated dwarfs, but less than those in DF/df mice (p<0.05 for both comparisons). In dwarfs retreated with PRL after withdrawal, the TIDA population was also smaller than that in normal animals (p<0.05), although it was larger than in vehicle-treated dwarfs of the same age (p<0.05). No effect of PRL treatment on TIDA cell numbers in normal mice, or of treatment or mouse phenotype on the number of TH-positive cells in zona incerta, occurred in either experiment. These results indicate that the effect of PRL on preventing the reduction in the TIDA population in df/df mice is limited to a developmental period prior to 21 d postnatally. In addition, this study provides evidence that continuous PRL feedback is required to maintain normal numbers of TIDA neurons. These findings extend the evidence for a critical role of PRL feedback in the differentiation and preservation of phenotype in TIDA neurons.

Mouse growth hormone transcription factor Zn-16: unique bipartite structure containing tandemly repeated zinc finger domains not reported in rat Zn-15.

Rat Zn-15 is a transcription factor activating GH gene expression by synergistic interactions with Pit-1, named for 15 DNA-binding zinc fingers, including fingers IX, X, and XI that are responsible for GH promoter binding. In this study, a mouse cDNA for Zn-15 was characterized. The predicted 2192-amino acid mouse protein is 89% identical to rat (r) Zn-15 overall, and is 97% similar in the C-terminal domain necessary for binding the GH promoter. However, the mouse cDNA encodes 16 zinc fingers, and sequences of rZn-15 pituitary cDNAs were the same as the mouse (m) Zn-16; the rat sequence in GenBank has a one nucleotide offset of a 17-bp segment in the finger V region. The mouse and corrected rat sequences contain four tandemly repeated fingers in the N-terminus, each separated by seven amino acids, typical of zinc finger proteins of the transcription factor IIIA-type. Analysis of mZn-16 expression by RT-PCR showed that the mRNA is, produced at similar levels in normal and GH-deficient Ames dwarf (Prop-1 ) mouse pituitaries at postnatal day 1. Mouse Zn-16 mRNA also was detected by ribonuclease protection assay in the pre-somatotrophic mouse cell line GHFT1-5. The Zn-16 protein is bipartite in that the N-terminal half displays tandem spacing typical of most zinc finger proteins, while the C-terminal portion contains long linkers between fingers that cooperatively bind to a DNA response element. Expression in early postnatal pituitary and in pre-somatotrophic cells suggests that Zn-16 could play a role in pituitary development prior to somatotroph differentiation.

Median eminence-afferent vasoactive intestinal peptide (VIP) neurons in the hypothalamus : localization by simultaneous tract tracing and immunocytochemistry

Retrograde tract tracing and immunocytochemistry were used to investigate the CNS source of the VIP that is present in high concentrations in the hypophysial portal blood and has been shown to have a stimulatory effect on pituitary prolactin secretion. Fluoro-gold (FG), which enters the CNS through areas devoid of the blood-brain barrier, such as median eminence, was injected peripherally. Brain sections from FG-treated animals were immunostained for VIP. A small population of VIP-containing cell bodies in the parvocellular and periventricular parts of the paraventricular nucleus (PVN) was also labeled with FG. Vasoactive intestinal peptide-immunoreactive perikarya not labeled with FG were also observed in the PVN, as well as FG-labeled cells that did not contain VIP. The results suggest that some VIP-producing neurons in the PVN project to the median eminence and are, therefore, functionally related to pituitary regulation; the function of other VIP neurons in the PVN is unknown.

Altered Growth Hormone-Releasing Hormone mRNA Expression in Transgenic Mice with Excess or Deficient Endogenous Growth Hormone

Hypothalamic expression of growth hormone-releasing hormone (GHRH) mRNA was examined in two transgenic mouse models displaying excess or deficient endogenous GH. Transgenic dwarf mice bore a gene construct consisting of the rat growth hormone (GH) promoter fused to a diphtheria toxin A chain structural gene (DT-A); the GH promoter restricted DT-A expression to endogenous GH-producing cells, which were destroyed. GH was undetectable in either the pituitary or the peripheral circulation (Behringer et al., Genes Devel. 2: 453-461, 1988). Transgenic giant mice carried a construction joining the metallothionein promoter to the human GHRH structural gene, which stimulated endogenous pituitary GH production (Hammer et al., Nature 315: 413-417, 1985). In situ hybridization to GHRH mRNA in transgenic dwarf, giant, and nontransgenic controls was performed using single-stranded RNA probes generated from cloned mouse GHRH cDNA. Hybridization to GHRH mRNA was limited to the neurons of the hypothalamic arcuate nucleus (ARC). Autoradiographic densities on X-ray films were quantified by computerized image analysis. There was an increase in GHRH signal intensity in the dwarfs (282 +/- 20 units; mean +/- SEM) relative to that measured in control animals (107 +/- 8 units; P < 0.001), while giant mice had decreased signal (42 +/- 7 units; P < 0.001) in the ARC. The present studies demonstrate that increase in GHRH mRNA expression accompanies GH deficiency, while a decrease in GHRH mRNA accompanies GH excess, suggesting both positive and negative feedback upon steady-state mRNA levels in hypophysiotropic neurons by target pituitary hormone.

Transcript abundance in mouse pituitaries with altered growth hormone expression quantified by reverse transcriptase polymerase chain reaction implicates transcription factor Zn-16 in gene regulation in vivo.

The correlation of growth hormone (GH) mRNA abundance and expression of specific transcription factors was studied in pituitaries of panhypopituitary (Ames df/df and Snell dwJ/dwJ dwarf), isolated GH-deficient (lit/lit), and GH-overproducing (growth hormone-releasing hormone [GHRH] transgenic) mice compared with normal littermates. A fluorescence-based reverse transcriptase polymerase chain reaction assay was developed for seven target mRNAs: GH, prolactin (PRL), proopiomelanocortin (POMC), α-subunit of the glycoprotein hormones (αSU), Pit-1, Prop-1, and Zn-16. Amplification parameters for each of these primer pairs were determined in order to calculate initial mRNA transcript number. The reproducibility of the assay was found to be ±10% for either Pit-1 or Zn-16 mRNAs measured in characterized murine GHFT1-5 somatotroph precursor cells. The cell extracts also showed an increased abundance of both Zn-16 and Pit-1 mRNAs when compared with whole pituitary extracts. Measurement of copy number in normal pituitaries showed that for every 106 GH or PRL mRNAs, there were 3 × 105 POMC, 4 × 104 αSU, 2 × 103 Pit-1, and only 70 Zn-16 or Prop-1 transcripts. Transcript abundance in GH-altered mice as a percentage of copy number per normal gland showed that POMC was significantly reduced in dwJ/dwJ (p<0.01) and df/df (p<0.05) mice. αSU mRNA was reduced in df/df (p<0.05), dwJ/dwJ (p<0.05), and lit/lit (p<0.05) mice, but not in GHRH-excess mice. PRL mRNA was not detected in dwarf mice, reduced to 52% of normal in lit/lit (p<0.05), and unchanged in GHRH-excess animals. GH mRNA was not detected in dwarf mice, reduced to 1.3% in lit/lit (p<0.005), and increased to 242% in GHRH-excess mice (p<0.05). Pit-1 mRNA was not detected in dwarf mice, was 2.9% of normal in lit/lit (p<0.005) mice, and increased to 200% in GHRH-excess mice (p<0.05). Prop-1 was not present in dwarf mice, was decreased to 1.4% in lit/lit (p<0.01), and increased to 223% in GHRH-excess mice (p<0.05). Zn-16 abundance in df/df mice was significantly reduced (p<0.05) to 4.8% of normals, to 6.3% of normals in dwJ/dwJ (p<0.005), to 6.1% of normals in lit/lit (p<0.005) mice, and significantly elevated in GHRH-excess mice to 197% (p<0.05). Altered pituitary mRNA abundance was found for several products not previously measured, or thought not to be affected by these mutations. Correlation of GH mRNA abundance with transcription factor copy number showed a significant correlation for Pit-1, Prop-1, and Zn-16. These quantitative analyses provide the first in vivo evidence that Zn-16 mRNA abundance correlates with GH expression.