Growth Hormone-Releasing Hormone (GHRH) - Oxford Academic

0013-7227/03/$15.00/0 Printed in U.S.A.

Endocrinology 144(6):2728 –2740 Copyright © 2003 by The Endocrine Society doi: 10.1210/en.2003-0006

Growth Hormone-Releasing Hormone (GHRH) Neurons in GHRH-Enhanced Green Fluorescent Protein Transgenic Mice: A Ventral Hypothalamic Network N. BALTHASAR, P.-F. MERY, C. B. MAGOULAS, K. E. MATHERS, A. MARTIN, P. MOLLARD, I. C. A. F. ROBINSON

AND

Division of Molecular Neuroendocrinology (N.B., C.B.M., K.E.M., A.M., I.C.A.F.R.), National Institute for Medical Research, Mill Hill, London NW7 1AA, United Kingdom; and Institut National de la Sante´ et de la Recherche Medicale Unite´-469 (P.-F.M., P.M.), 34094 Montpellier Cedex 5, France The hypothalamic GHRH neurons secrete pulses of GHRH to generate episodic GH secretion, but little is known about the mechanisms involved. We have made transgenic mice expressing enhanced green fluorescent protein (eGFP) specifically targeted to the secretory vesicles in GHRH neurons. GHRH cells transported eGFP from cell bodies in the arcuate nucleus to extensively arborized varicose fiber terminals in the median eminence. Patch clamp recordings from visually identified GHRH cells in mature animals showed spontaneous action potentials, often firing in short bursts up to 10 Hz. GHRH neurons received frequent synaptic inputs, as demonstrated by the recording of abundant inward postsynaptic currents, but spikes were followed by large after-hyperpolarizations,

P

ULSATILE GH SECRETION IS regulated by two opposing hypothalamic hypophysiotropic systems, producing GHRH (1) and somatostatin (2). GHRH neurons, mostly in the hypothalamic arcuate (ARC) nuclei (3), project their terminals to the median eminence (ME) (4), from which they release GHRH episodically into portal blood to stimulate pituitary GH release (5, 6). Somatostatin neurons in the periventricular nuclei also project to the ME and release their peptide to inhibit GH release periodically. Much circumstantial evidence supports a model (7) in which episodic GH release is controlled by phasic secretion of GHRH and somatostatin. The mechanisms regulating pulsatility are important because different patterns of GH release have different effects on body growth (8 –10), signal transduction mechanisms (11, 12), and target gene expression (13). The regularity of spontaneous GH pulses implies an underlying pulse generator, either intrinsic to GHRH or somatostatin cells or extrinsic to them but driving their episodic activity in a coordinated fashion (14). Electrophysiological recordings have been made from unidentified ARC neurons (15), but it has not been possible to characterize the electrophysiological properties of identified GHRH neurons to know whether they show periodic increases in firing rates or bursting patterns of activity, as has been described for other hypothalamic neuroendocrine cells (16). To generate a GHRH pulse, there is likely to be some Abbreviations: ARC, Arcuate; DIG, digoxigenin; eGFP, enhanced green fluorescent protein; hGH, human GH; ME, median eminence; NT, nontransgenic; rGHRH, rat GHRH; RNase, ribonuclease; SV, secretory vesicle; T, transgenic.

which limited their firing rate. Because many GHRH neurons lie close to the ventral hypothalamic surface, this was examined by wide-field binocular epifluorescence stereomicroscopy. This approach revealed an extensive horizontal network of GHRH cells at low power and individual fiber projections at higher power in the intact brain. It also showed the dense terminal projections of the GHRH cell population in the intact median eminence. This model will enable us to characterize the properties of individual GHRH neurons and their structural and functional connections with other neurons and to study directly the role of the GHRH neuronal network in generating episodic secretion of GH. (Endocrinology 144: 2728 –2740, 2003)

temporal synchronization of episodic firing in a network of GHRH cells connected directly or indirectly via interneurons (14, 17). To investigate this, it is necessary to study structural and functional properties of populations of identified GHRH cells and monitor their connectivity and activity. Because peptide release from terminals can be regulated independently of firing activity in neuroendocrine cell bodies (18), an approach that could give some insight into the transport and release of GHRH from secretory vesicles (SVs) in the GHRH neuron terminals would also be useful. Recently, enhanced green fluorescent protein (eGFP) expressed from transgenes has been shown to be a useful way to first identify and then record from specific hypothalamic neurons (19 –22). We have now generated transgenic mice with eGFP targeted to GHRH cells, using a genomic GHRH promoter construct (23). By using an eGFP variant that is packaged in SVs (24), the transport of eGFP-tagged vesicles to terminals and exocytotic release events may also be visualized (22, 24, 25). In this study we present the first electrophysiological recordings of bursting activity from preidentified GHRH-eGFP neurons and use multiple approaches to image a network of GHRH cells and their terminals at the subcellular, single cell, and multicellular levels at the intact ventral surface of the brain. Materials and Methods The rat GHRH (rGHRH)-eGFP cosmid construct The rGHRH-eGFP transgene is illustrated in Fig. 1. The eGFP reporter cassette has been described (24); it contains genomic sequences for the signal peptide, first intron, and N-terminal 22 residues of human GH

2728

Balthasar et al. • GHRH-eGFP Transgenic Mice

Endocrinology, June 2003, 144(6):2728 –2740 2729

FIG. 1. Construction of the rGHRH-eGFP transgene. A, We used the Mlu1 fragment (described in Ref. 24) containing genomic sequences for the signal peptide (SP), first intron, and N-terminal 22 residues of hGH (dark bars) fused in frame via a 15-mer oligonucleotide linker to the coding sequence of eGFP (hatched bar). B, This fragment was subcloned into a 12-kb KpnI plasmid containing rGHRH genomic sequence via a single MluI restriction site previously introduced in the 5⬘untranslated region of the first hypothalamic exon of GHRH (23). C, This was then packaged into a rat GHRH cosmid (23), containing 16 kb of 5⬘- and 14 kb of 3⬘-flanking sequence and the insert released by digestion with NotI. (hGH) fused in frame via a 15-mer oligonucleotide linker to the coding sequence of eGFP (Fig. 1A). After cleavage of the GH signal peptide, the fluorescent product is eGFP with a short N-terminal peptide extension; for simplicity this will be referred to as eGFP. This cassette was cloned into a 12-kb KpnI plasmid containing rGHRH genomic sequence via a single MluI restriction site introduced in the 5⬘untranslated region of the first hypothalamic exon of GHRH (Fig. 1B) and then packaged (Gigapack III XL, Stratagene, Amsterdam, The Netherlands) into a rat GHRH cosmid, containing 16 kb of 5⬘- and 14 kb of 3⬘-flanking sequence (Fig. 1C), all as previously described (23).

Generation and analysis of GHRH-eGFP transgenic mice The rGHRH-eGFP DNA insert was released by digestion with NotI, purified by ultracentrifugation in a 5–20% salt gradient, and brought to a concentration of 1–5 ng/␮l in 10 mm TrisHCl (pH 7.5), 0.1 mm EDTA (pH 8.0). Transgenic mice were generated by pronuclear microinjection of this construct into fertilized oocytes of superovulated (CBa/Ca ⫻ C57Bl/10 F1) mice followed by oviductal transfer into pseudopregnant recipients. Genomic DNA from tail biopsies was obtained and analyzed for transgene DNA by Southern blotting and PCR. Southern blots were performed on BglII-digested genomic DNA with a probe homologous to the hGH sequences and 65 bp of eGFP sequence in the reporter cassette, radiolabeled with [␣32P]dCTP and [␣32P]dATP] by random-prime labeling (Prime-a-gene kit, Promega Corp., Southampton, UK) and hybridized to 4.9-kb and 1.4-kb fragments in transgenic and wild-type DNA, respectively. Progeny were subsequently genotyped using a PCR assay (24) that amplifies across the first intron of hGH present in the transgene. All lines were maintained as hemizygous, with nontransgenic (NT) littermates serving as controls for the transgenic (T) animals. All T mice were generated and maintained at the National Institute for Medical Research, London; breeding pairs were then sent to establish a parallel colony of line 39 at Institut National de la Sante´ et de la Recherche Medicale Unite´ -469, Montpellier, for other experiments. All animal experiments were carried out strictly in accordance with the relevant institutional and national guidelines at both centers.

RT-PCR RNA was extracted using Trizol reagent (Life Technologies, Inc., Paisley, United Kingdom) and 500 ng transcribed with 200 U Moloney murine leukemia virus reverse transcriptase (Roche Diagnostics, Lewes, UK) in 1⫻ Moloney murine leukemia virus reverse transcriptase buffer (Roche Diagnostics) supplemented with 1 ␮g random primers (Invitrogen, Paisley, UK), deoxynucleotide triphosphates (0.3 mm, Amersham Pharmacia Biotech, Little Chalfont, UK), 40 U ribonuclease (RNase) inhibitor (Promega Corp.), and 5 mm dithiothreitol. The mixture was incubated at 37 C for 2 h and cDNAs amplified by PCR. For the transgene product, the primers spanned from hGH to eGFP sequences and were AACCACTCAGGGTCCTGTGGACAG (forward) and GAGGACGGCAACATCCTGGGGCA (reverse) to amplify a predicted fragment size of 950 bp. Mouse ␤-actin control transcripts were amplified using the primers TTGTAACCAACTGGGACGATATGG (forward) and GATCTTGATCTTCATGGTGCTAGG (reverse) to amplify a predicted fragment size of 764 bp.

RNase protection assay RNase protection assays were performed using the RNase protection assay III kit (Ambion, Inc., Huntingdon, UK). [␣32P]-uridine 5-triphosphate-labeled RNA probes were generated by in vitro transcription (see below) and purified by 5% acrylamide gel electrophoresis. For mGHRH, a PvuII digest of an mGHRH clone (I.M.A.G.E., 1496474, HGMP Resource Centre, Cambridge, UK) was transcribed to generate a 330-bp probe that protects a 300-bp 3⬘ fragment of mGHRH. For ␤-actin a 334-bp probe was generated to protect a 246-bp fragment of RNA. Labeled probes (1 ⫻ 105 cpm) were incubated with 10 ␮g hypothalamic RNA samples at 42 C overnight. For quantification, RNA samples from two hypothalami were pooled and four extracts each of T or NT mice were processed separately. After hybridization, samples were treated with RNase and protected fragments separated on 5% acrylamide gels. Gels were analyzed using ImageQuant (Molecular Dynamics, Inc., Sunny-

2730

Endocrinology, June 2003, 144(6):2728 –2740

vale, CA), and the amounts of protected sample RNA were normalized to ␤-actin RNA measured in the same samples.

In situ hybridization Antisense and sense riboprobes corresponding to full-length cDNAs for eGFP or mouse GHRH were labeled with either [␣35S]-uridine 5-triphosphate or digoxigenin (DIG) using SP6/T7 transcription kits (Roche Diagnostics), treated with Dnase I for 15 min at 37 C, and purified by gel filtration or phenol/chloroform extraction and ethanol precipitation. For radiolabeled probes, in situ hybridizations were performed as previously described (26). Slides were exposed to autoradiographic films that were scanned and analyzed using NIH Image. Slides were then dipped in photographic emulsion (G5, Ilford Imaging, Knutsford, UK) for several days, developed, and counterstained in cresyl-fast violet. For DIG-labeled probes, paraffin-embedded slides were rehydrated, permeabilized (10 ␮g/ml proteinase K, Sigma, Poole, UK), fixed in 4% paraformaldehyde, acetylated, and incubated in prehybridization buffer (4⫻ saline sodium citrate, 50% formamide). Sections were then hybridized overnight, washed, and incubated with RNase (40 mg/ml) at 37 C for 30 min. DIG labeling was visualized (27) using sheep anti-DIGalkaline phosphatase and Nitro-blue tetrazolium/Bromo-chloroindolyl-phosphate as substrate.

RIAs Tissues were homogenized and assayed for eGFP protein or GH using RIAs as previously described (24).

Confocal microscopy Coronal vibratome sections (150 ␮m) were cut from GHRH-eGFP mouse brains and placed in a recording chamber on a microscope stage, perfused with artificial cerebrospinal fluid (mm: 120 NaCl, 3 KCL, 1.2 NaH2PO4, 2 CaCl2, 1 MgSO4, 23 NaHCO3, 10 glucose; gassed with 95% O2 and 5% CO2 at 32 C). These were imaged using a Axioskop upright microscope (Carl Zeiss, Le Pecq, France), a ⫻60, 0.9 NA, long-working distance water immersion objective (Olympus Corp.), and an MRC 1000 scanning confocal microscope (Bio-Rad Laboratories, Inc., Hemel Hempstead, UK). For excitation, the 488-nm line of an argon ion laser was used, and emission filters were optimized for eGFP. Digital images were collected in COMOS (Bio-Rad Laboratories, Inc.) and analyzed offline in NIH Image.

Culture of hypothalamic neurons Brains from a litter of 1- to 2-d old pups were placed in ice-cold HDMEM (25 mm HEPES-buffered Dulbecco’s medium, pH 7.3, Life Technologies, Inc.) and the hypothalamus dissected into an ice-cold solution of 25 mm HEPES-buffered Ca2⫹ and Mg2⫹-free Earle’s BSS, pH 7.3 (Life Technologies, Inc.). Tissues were transferred into a solution containing 6 mg/ml papain, 1 mg/ml DL-cysteine, 250 ␮m EDTA, 50 ␮m ␤-mercaptoethanol in 25 mm HEPES-buffered Ca2⫹ and Mg2⫹-free Earle’s BSS, pH 7.3 (Life Technologies, Inc.), and incubated at 37 C in a shaking water bath for 20 –25 min. Tissue was washed and then incubated with 1 mg/ml chick egg white trypsin inhibitor, 0.1 mg/ml Dnase in HDMEM (Life Technologies, Inc.). After dilution with 5 ml culture medium (1% N2 medium supplement, 10% horse serum, 1% antibiotics, 1 mm sodium pyruvate, 20 mm glucose in glutamine-free basal medium Eagle, Life Technologies, Inc.), tissue was dissociated by trituration, centrifuged (200 ⫻ g for 2 min), and the cell pellet resuspended and layered on 13-mm coverslips coated with poly-d-lysine- and laminin. Coverslips were removed from culture every 24 h, fixed in 4% paraformaldehyde, mounted in Mowiol (Harco, Essex, UK), and examined by bright-field and fluorescence microscopy.

Slice preparation for electrophysiological recording The procedure was essentially as described in (28) but adapted for adult mice. The GHRH-eGFP mice (6 –12 wk old, line 39) were anesthetized by isoflurane inhalation, killed by decapitation, and brains quickly removed into cold (0 – 4 C) solution-1 (in mm; 92 NMDG-Cl, 2.3 KCl, 1 CaCl2, 6 MgCl2, 26 NaHCO3, 1.2 KH2PO4, 25 glucose, 0.2 ascorbic


acid, 0.2 thiourea (pH 7.4), gassed with 95% CO2, 5% O2). Meninges and blood vessels were carefully dissected from the ventral hypothalamus, sagittal sections (300 ␮m) cut with a vibrating blade microtome (DSK, Kyoto, Japan) and stored at 34 C in solution-2 (mm; 115 NaCl, 2.5 KCl, 1 CaCl2, 4 MgCl2, 26 NaHCO3, 1.25 NaH2PO4, 25 glucose, 0.2 ascorbic acid, 0.2 thiourea, pH 7.4, gassed with 95% CO2, 5% O2) for at least 30 min.

Patch-clamp recordings Slices were immobilized with a nylon grid in a submersion chamber on the stage of an upright microscope (Axioskop FS, Carl Zeiss) and superfused continuously with solution-3 (mm; 125 NaCl, 2.5 KCl, 2 CaCl2, 1 MgCl2, 26 NaHCO3, 1.25 NaH2PO4, 5 glucose, pH 7.4, gassed with 95% CO2, 5% O2) at a rate of approximately 4 ml/min for at least 30 min at 25–27 C. They were viewed with a ⫻63 immersion objective (Achroplan, Carl Zeiss) and Nomarski differential interference contrast optics. Infrared differential interference contrast illumination was used to visualize neurons deeper in the slices and the images captured with an infrared camera (Hamamatsu Photonics, Massy, France), displayed, and stored on a computer controlled by Scion 1.62c (NIH Image). Bosilicate glass pipettes (5– 8 m⍀) filled with a standard intracellular solution (in mm; 100 K-gluconate, 40 KCl, 2 MgATP, 3 EGTA, 10 HEPES, pH 7.3 with KOH) were used for whole-cell patch clamp experiments in voltage- or current-clamp modes. Alexa 350 (0.1 mm, Molecular Probes, Inc., Eugene, OR) was added to the pipette solution as an intracellular marker dye (29). Pipettes were connected to the head stage of an EPC-9 amplifier (HEKA, Lambrecht, Germany) to acquire and store patchclamp data using Pulse 8.09 software (HEKA). Pipette and cell capacitances were fully compensated, and recorded voltages were corrected for a junction potential of ⫺10 mV (30). Spontaneous activity was recorded in current-clamp mode or at a fixed membrane potential (typically ⫺70 mV) in voltage-clamp mode. Passive and active membrane properties were examined in current-clamp mode by injecting triggered current pulses (50 msec or 200 msec duration) through the pipette. Recordings were analyzed with IgoPro version 3.0 (Wavemetrics, Lake Oswego, OR) and Axograph 4.0 (Axon Instruments Inc., Foster City, CA).

Intact brain imaging by epifluorescence stereomicroscopy Brains were removed as described above and kept in cold solution-2. Care was taken when cutting the pituitary stalk to avoid damaging the median eminence. Whole brains were transferred, ventral surface uppermost, to the stage of a modified stereomicroscope equipped with epifluorescence and ⫻1.6, ⫻2.5, or ⫻20 objectives (Carl Zeiss; Kramer, Valley Cottage, NY). The intensity of the fluorescent lamp (HBO 100 W; Osram, Munsen, Germany) was reduced to less than 50% by an AttoArc power supply (Carl Zeiss). Fluorescence images (16 bit) were acquired with a cooled CCD (Roper Scientific, Evry, France) controlled using Metamorph 4.0 (Universal Imaging, West Chester, PA). Image stacks were collected with a motorized Z-controller (LEP, Hawthorne, NY) at fixed step amplitudes (typically 200 nm) and deconvoluted with Huygens2 as previously described (31).

Data analysis Unless otherwise stated, data are shown as mean ⫾ sem. Differences between groups were analyzed by t test, with P less than 0.05 taken as significant (n.s. ⫽ nonsignificant).

Results

Of 23 pups born, four were transgenic for the rGHRHeGFP construct and bred with C57Bl/10 mice. One founder was sterile; from the other three founders, lines (39, 151, and 342) were established in which the transgenes transmitted 50:50% to both males and females. PCR and Southern blotting readily distinguished T from NT animals (Fig. 2, A and B), and RT-PCR detected an eGFP transgene product of the expected size in hypothalamic extracts from T but not NT



FIG. 2. Analysis of GHRH-eGFP transgenic mice. A, PCR of tail DNA was used to identify the transgenic founder in line 39 (F#39) and distinguish between its T and NT progeny using primers spanning the first GH intron (350 bp for transgene and 300 bp for mouse GH). B, Southern blot of tail DNA showed a strong 4.9-kb transgene band and a weaker 1.4-kb band corresponding to endogenous mouse GH. C, RT-PCR on hypothalamic RNA extracts amplified the expected 950-bp hGH/eGFP mRNA fragment in T but not NT progeny, and ␤-actin transcripts (764 bp) were amplified from both. Control reactions without reverse transcriptase (⫺RT) or input RNA (⫺RNA) gave no product.

mice (Fig. 2C). Many of the subsequent studies were performed on all three lines, but for clarity, data presented will be from line 39, unless stated otherwise. Animals were weighed weekly for 18 wk, and the growth curves of T and NT littermates were indistinguishable for all three lines. At 18 wk, body weights were 35.1 ⫾ 1.7 vs. 35.1 ⫾ 2.2 g, T vs. NT males, and 25.1 ⫾ 0.8 vs. 26.1 ⫾ 0.9 g, T vs. NT females (n ⫽ 5–9 per group, n.s.). Pituitary GH contents were also unaffected (42.0 ⫾ 5.2 vs. 37.1 ⫾ 5.6 ␮g, male T vs. NT mice, n ⫽ 5, n.s.). RIA of hypothalamic extracts showed eGFP-immunoreactive protein in T but not NT mice from each line (Fig. 3), with line 151 showing a 6- to 10-fold higher hypothalamic eGFP protein content than the other two lines. Extracts from groups of males and females from line 39 showed similar eGFP contents (9.7 ⫾ 2.2 vs. 8.1 ⫾ 1.5 ng eGFP/hypothalamic extract, n ⫽ 7). RNase protection assays, performed on hypothalamic extracts from adult NT and T mice from line 39 showed that expression of eGFP protein in GHRH neurons did not affect endogenous mouse GHRH mRNA abundance (2.61 ⫾ 0.4 vs. 2.45 ⫾ 0.5 arbitrary densitometer units, n ⫽ 4 pairs per group, n.s.). In situ hybridization was used to screen for eGFP mRNA in the central nervous system and a variety of peripheral tissues in line 39. Strong eGFP expression was seen mainly in the ARC nuclei (Fig. 4Aa), with a few positive neurons also apparent in more dorsomedial hypothalamic regions. No signal was observed in NT animals (Fig. 4Ab) or in T animals with a sense eGFP riboprobe. No eGFP expression was seen in any other brain area, pancreas, thymus, pituitary, stomach, testes, spleen, liver, or heart. Some eGFP signal was detected sporadically in a few glomerular podocyte cells in the kidney cortex, but the relevance of this was unclear because it was seen in all three lines and some but not other T littermates of

FIG. 3. Hypothalamic content of eGFP protein in three lines of GHRH-eGFP mice. Hypothalami were dissected from groups of six transgenic animals from litters from the three different lines (39, 151, 342) of GHRH-eGFP transgenic mice and from six NT mice, homogenized and assayed for eGFP protein by RIA. EGFP was not detectable (n.d.) in NT mice but present in all three transgenic lines, with line 151 showing significantly larger amounts than in the other two lines. ***, P ⬍ 0.001.

each line, and mouse GHRH mRNA expression was not detected in these cells. Emulsion dipping and counterstaining the sections confirmed that eGFP transcripts were expressed in a distinct population of ventral ARC neurons (Fig. 4, Ac–Af) spreading from medial to lateral areas of the ARC. The overall pattern matched that of GHRH expression examined separately by in situ hybridization (Fig. 4Ag), and simultaneous double in situ hybridization using an 35S-labeled riboprobe for eGFP, and a DIG-labeled riboprobe for GHRH confirmed that eGFP and GHRH mRNAs were colocalized in the ARC neuronal cell bodies when examined under high power (Fig. 4Ah). Of 108 labeled ARC neurons examined in sections from GHRHeGFP transgenic mice of line 39, both mRNAs were colocal-

2732



FIG. 4. Expression and distribution of eGFP in arcuate GHRH neurons. A, Cryostat sections of hypothalamus were taken for in situ hybridization with a 35S-labeled eGFP riboprobe and exposed to x-ray film. EGFP expression was detected predominantly in the ARC (arrow) in T (a) but not NT (b) mice. Sense control probes gave no signal (not shown). After dipping in photoemulsion and counterstaining, the distribution of eGFP-positive cells could be observed at low power (⫻10) in the bilateral ARC (c, d) on either side of the third ventricle, concentrated in the ventral portion of ARC. Individual cell bodies were resolved at higher power (⫻40), under bright field (e), and the silver grains visualized as bright spots in dark-field images of the same sections (f). A 35S-labeled riboprobe for mouse GHRH confirmed its expression in a similar pattern in ARC (g), and double in situ hybridization with a DIG-labeled GHRH riboprobe (blue) and a 35S-labeled eGFP riboprobe (silver grains, black) showed coexpression in most ARC GHRH cells when examined individually (h) at high power (⫻100) after emulsion dipping and development. B, Freshly prepared hypothalamic slices (150 ␮m) from GHRH-eGFP mice (line 39) were placed in a perfusion chamber, maintained in artificial cerebrospinal fluid, and regions of the arcuate nucleus (a, b) and ME (c– e) examined by confocal microscopy. Clusters of brightly fluorescing eGFP-positive neurons could be seen in ARC, which at higher power (b) showed a punctate appearance of eGFP in their axons. Numerous eGFP-filled varicose were evident, coursing toward the ME (c), which showed a very bright accumulation of eGFP-filled terminals. Individual fibers, beaded varicosities, and terminals could be resolved (d, e) in the external zone of the ME. 3v, Third ventricle. C, Dissociated cells from neonatal hypothalami of GHRH-eGFP mice (line 39) were placed in tissue culture and examined by confocal microscopy 4 h (top panels) and 72 h (bottom panels) later. Left panels, eGFP fluorescence; right panels, bright-field illumination. Scale bars, 10 ␮m.

ized in 90% of the cells, and in 6% and 4%, only eGFP or GHRH mRNAs were detected, respectively. Similar results were obtained for transgenic mice from line 342. However, animals from line 151 showed much more widespread and intense eGFP expression with many more

eGFP-labeled cells in many other hypothalamic areas in addition to the ventral ARC, consistent with the higher amounts of eGFP protein detected in this line (Fig. 3). In the ARC of this line, colocalization of eGFP and GHRH was observed in only 81% of the cells. Because eGFP was clearly


not a specific reporter for GHRH cells in line 151, these animals were not studied further. Fluorescence microscopy

Fluorescence microscopy on vibratome slices (150 ␮m) of hypothalamus from transgenic mice from lines 39 and 342 showed fluorescent neurons in ARC and a stronger fluorescent signal in the ME. No other regions of the hypothalamus showed eGFP fluorescence in these lines. Using confocal microscopy, individual GHRH cell bodies and their primary axons were resolved, which showed a punctate pattern of eGFP fluorescence (Fig. 4, Ba and Bb). In the ME, the terminal fields of these neurons were brightly fluorescent (Fig. 4Bc), and fluorescent beaded varicosities were evident along the entire length of these projections (Fig. 4, Bd and Be) coursing throughout both internal and external zones of the ME. These images were consistent with the packaging of the eGFP product into SVs, and their transport along the axons, to be stored in the ME terminals of GHRH neurons. We tested whether GHRH-eGFP neurons could be identified after in vitro culture of dissociated hypothalamic cells from newborn mice from line 39. Because the hemizygous mice could not be genotyped before pooling for cell culture, the cells originated from both T and NT pups. Very few eGFP-fluorescent rounded cells were visible after 4 h of culturing (Fig. 4C, upper panels). After 72 h, only one to two eGFP fluorescing neurons per coverslip could be detected, but these were bright and had clearly discernible projections containing fluorescent material at their tips (Fig. 4C, lower panels). Electrophysiological recordings from GHRH-eGFP neurons

Patch-clamp experiments were performed on eGFP-positive GHRH neurons visualized in acutely prepared parasagittal brain slices from adult GHRH-eGFP mice (Fig. 5). Two eGFP-positive neurons were seen in the field shown in Fig. 5Ab and the left one brought into focus. Infrared illumination was used to monitor the shape of this neuron before (Fig. 5Aa) and during (Fig. 5Ac) patch-clamp recording. The patch pipette was loaded with a blue fluorescent dye (Alexa 350, Molecular Probes, Inc.), which labeled the cytoplasm of the eGFP-positive neuron being recorded (Fig. 5Ad). All eGFP-positive neurons of rGHRH-eGFP mice displayed significant spontaneous membrane activity when studied in the current-clamp mode. For the recording shown in Fig. 5B, the mean action potential firing rate was 0.83 Hz, although the firing rate was frequently higher in brief bursts of activity, when the instantaneous firing rate could reach up to 10 Hz. Subthreshold depolarizations were also observed, especially between bursts of action potentials. Action potential spikes were always preceded by brief depolarizations and followed by sharp hyperpolarizations. Triggered current injection was performed to resolve passive membrane properties and action potential kinetics of GHRH-eGFP neurons. Constant negative current injection was often required to minimize spontaneous activity under steady-state conditions, upon which 50-msec-duration pulses of varying amplitude were delivered. In Fig. 5C, the membrane potential was driven in a linear manner by sub-


threshold current injections (⫺15 to 40 pA). Current injections above 40 pA elicited action potential firing as membrane potential exceeded threshold. Note again the large afterhyperpolarization phase following repolarization. The responses of GHRH-eGFP neurons to longer (200 msec) current pulses were also examined (Fig. 5D). Moderate current injections elicited action potentials that shared almost identical threshold, peak, and afterhyperpolarization characteristics. However, increasing current amplitude from 40 (top panel) to 55 pA (bottom panel) increased both the depolarization rate near threshold and action potential firing. On average, the action potential threshold was ⫺47.7 ⫾ 1.6 mV (n ⫽ 8), and the action potential peaked at 10.1 ⫾ 3.8 mV (n ⫽ 8), followed by an afterhyperpolarization (⫺62.8 ⫾ 2.2 mV, n ⫽ 8), well below the resting potential (⫺54.1 ⫾ 1.89 mV, n ⫽ 10). Spontaneous ionic currents were always observed in the voltage-clamp mode (15 experiments). In the recording shown in Fig. 5E, inward synaptic currents are shown in an eGFP-positive neuron held at ⫺70 mV. These inward currents were often clustered or even superimposed, and their fast kinetics are shown in Fig. 5F. The maximal amplitude of the inward currents increased as the holding potential was lowered to ⫺90 mV and decreased when the potential was held at ⫺50 mV (data not shown). No outward currents were observed under these experimental conditions. Imaging the GHRH ventral neuronal population in intact brains from GHRH-eGFP mice

Because many of the GHRH cells are very close to the ventral hypothalamic surface, we examined the ventral surface of the intact brain of a GHRH-eGFP mouse with an epifluorescence binocular stereomicroscope. Figure 6 shows that a large population of eGFP-positive cells, their fibers, and terminals is readily visible with this technique. At low magnifications (Fig. 6, A and B), an intense eGFP signal was evident in the ME. As magnification was progressively increased (Fig. 6, C–E), numerous eGFP-positive neuronal cell bodies became evident in the vicinity of the ME, with more scattered eGFP-positive neurons in lateral areas of the hypothalamus. At the highest magnifications (Fig. 6, E and F), individual rGHRH-eGFP neurons could be distinguished and the punctate subcellular distribution of eGFP fluorescence could be resolved in discrete areas of the cells and their processes (Fig. 6, E and F). Intriguingly, many bright eGFPcontaining beaded processes were seen apparently coursing between nearby rGHRH-eGFP neurons (Fig. 6F). The distribution of GHRH-eGFP-positive neurons in the ARC was also examined in the Z axis, moving into the brain from the ventral surface in 800-nm steps (Fig. 7, A–F). The GHRH-eGFP neurons could be resolved up to 80 ␮m deep within the brain, and several layers of eGFP-positive neurons could be seen in closely apposed clusters in a parasagittal region close to the ME. In more lateral regions, GHRH-eGFP neurons were localized in deeper areas of the hypothalamus. The intact ME was also examined at higher magnifications. Figure 8, A–F, shows a similar gallery of images moving from the ventral surface of the ME inward. Myriad small highly arborized brightly eGFP fluorescent bundles of fiber termi-

2734



FIG. 5. Electrophysiology of eGFP-positive neurons in GHRH-eGFP mice. Whole-cell patch-clamp recordings from eGFP-positive neurons were performed in brain slices from GHRH-eGFP mice from line 39 (A). The tissue was visualized with a ⫻63 objective under infrared light before (Aa) and during (Ac) the patch-clamp recording of an eGFP-positive neuron (Ab). Diffusion of Alexa 350 from the pipette confirmed that the recordings had been obtained from the GHRH-eGFP neuron. B, Spontaneous membrane potential variations are evident in an GHRH-eGFP neuron. C, Active and passive membrane potential variations were elicited by current injections. Pulses (50-msec duration) of varying amplitudes were delivered as indicated in the upper panel. D, Same neuron as in C. Representative recordings of the bursting activity, elicited by injection of 40 pA (top) or 55 pA (bottom) for 200 msec as indicated by the solid line. E, Spontaneous membrane currents recorded in an GHRH-eGFP neuron at a holding potential of ⫺70 mV. F, Enlarged view of part of the membrane current recording in E indicated by the star. In B–D, the 0 mV level is indicated by dotted lines.



FIG. 6. Epifluorescence visualization of the ventral GHRH-eGFP neuronal population in the intact brain. The whole brain of an GHRH-eGFP mouse from line 39 was removed and immersed ventral side uppermost in a cold Ringer solution on the stage of a binocular epifluorescence stereomicroscope. A–F, Images of the ventral surface of the hypothalamus were taken sequentially from the lowest (⫻1.5 in A) to the highest magnification (⫻80 in F). The box (D) indicates the location of the image F, Images were acquired and processed as detailed in Materials and Methods. Scale bars (A), 1 mm; (B), 400 ␮m; (C), 200 ␮m; (D and E), 100 ␮m; (F), 20 ␮m. Cb, Ventral side of the cerebellum; OC, optic chiasm. The bright central mass (A) is the median eminence.

nals of varying lengths were apparent as well as small dots (500 – 800 nm apparent diameter), which reflect the axons, varicosities, and nerve terminals of the GHRH-eGFP neurons; no GHRH-eGFP neuronal cell bodies were visible.

Discussion

There are two main requirements for using transgene reporters to identify and study specific neuronal cell groups (32).

2736



FIG. 7. Distribution of GHRH-eGFP neurons within the ARC in the intact brain. Images were obtained from the intact brain of a GHRH-eGFP mouse from line 39 using a binocular epifluorescence stereomicroscope. A series of images (A–F) were acquired from the ventral surface of the hypothalamus just lateral to the border of the ME from inside outward in 800-nm steps and processed as described in Materials and Methods. Note the clusters of closely apposed GHRH cell bodies throughout the region of ARC just parasagittal to the ME. Scale bar, 50 ␮m.

The first is that the construct must contain appropriate regulatory sequences to drive high-level expression, specifically restricted to the cells of interest. This is straightforward for neuroendocrine neurons because one can use the strong promoter and regulatory sequences of the genes coding for the major

peptide secretory products that define these cells. The second requirement is that the reporter must be readily imaged in living cells and not affect cell function. This is readily satisfied by eGFP and its variants (33), which have been expressed in a wide variety of cells without deleterious effects (34).



FIG. 8. Distribution of GHRH-eGFP neuron fibers and terminals within the ME in the intact brain. Images were obtained from the intact brain of a GHRH-eGFP mouse from line 39 using a binocular epifluorescence stereomicroscope. A series of images (A–F) were acquired from the ME from inside outward in 800-nm steps and processed as described in Materials and Methods. They show an extensive network of eGFP-containing processes, varicosities, and terminals at all depths within the ME. Scale bar, 20 ␮m.

Although neuroendocrine promoter transgenes driving GFP have proved successful in targeting other hypothalamic neuronal systems (19 –22), short GHRH promoter sequences have not succeeded (35) (Robinson, I. C. A. F., unpublished observations), probably because of the presence of several upstream promoters in this locus (36, 37)

whose enhancer and regulatory elements are poorly defined. However, we have engineered a GHRH cosmid that does drive transgene expression specifically in hypothalamic GHRH neurons in most lines (23, 38, 39), although low-level transgene expression can be detected in other tissues, reflecting low-level production of endogenous

2738


GHRH (40), other products from the GHRH precursor (41), or ectopic expression. We generated three GHRH-eGFP transgenic lines, none of which showed any overt phenotypic differences from their NT littermates. Hypothalamic GHRH mRNA, pituitary GH contents, and growth rates were all normal, suggesting that the rGHRH-eGFP transgene had no deleterious effect on the GHRH/GH axis. All three lines expressed a fluorescent eGFP product in ARC, but the specificity and intensity of expression differed between the lines. Two lines showed transgene expression highly restricted to the hypothalamic GHRH cells, mainly in ARC (42, 43) but with few scattered eGFPand GHRH-positive neurons in more dorsal hypothalamic regions (4). Colocalization studies showed that GHRH and eGFP mRNAs were coexpressed in at least 90% of the labeled ARC neurons in these two lines. This is probably an underestimate because GHRH signal intensity varies markedly from cell to cell, and the sensitivities of the probes are lower when used in double-label studies. Because only 6% of the ARC eGFP cells did not show detectable GHRH expression in line 39, eGFP identifies the vast majority of ARC GHRHexpressing neurons in this line. Although 4% of the GHRH cells that did not show eGFP expression probably reflect differences in RNA abundance and detection, it is possible that they activate their GHRH expression from a promoter element other than that into which the eGFP reporter was cloned. Other tissues showed no eGFP expression, with the exception of a few eGFP-positive cells in the kidney in some T individuals. The cosmid might contain cryptic kidneyspecific sequences but why these would be active in some animals but not others is unclear, and no GHRH transcripts were detected in these cells. However, because transgenic line 151 showed more widespread eGFP expression in other parts of the central nervous system, it is clear that even our 38-kb rGHRH cosmid does not contain all the locus control sequences (44) necessary for position-independent, tissuespecific GHRH expression in every line, and it was necessary to assess several lines of mice to identify one (line 39) in which eGFP was a reliable reporter for GHRH neurons. Because we wished to visualize GHRH terminals as well as cell bodies, we used an eGFP variant targeted to the secretory pathway (24) and obtained a distribution of eGFP fluorescence very similar to that of GHRH in mice (42, 43). Fluorescence and confocal microscopy showed ARC neurons with bright perinuclear fluorescence in the soma, from which axons could be traced to the ME, filled with intensely fluorescent varicose terminals. That the eGFP accumulates in the ME is expected because the vast majority of GHRH neurons project to this site, storing their peptide in terminal varicosities for release into hypophysial portal blood (43). In the outer zone of the ME, eGFP was clearly contained in vesicular varicosities, typical of those identified in beaded strands in GHRH neurons and their terminal arborizations by immunostaining and EM (45). Whether this eGFP variant is costored with GHRH in the same SVs will require a doubleimmunogold EM study, but in other studies in GH cells (24), we showed that this eGFP variant was copackaged with GH in SVs and coreleased on stimulation. It was possible to culture hypothalamic neurons from neo-


natal GHRH-eGFP mice. Only few eGFP-positive cells were detected, but these formed projections and maintained eGFP expression for up to 3 d. The use of homozygous litters or fluorescence-activated cell sorting of cells before culture may improve the efficiency, but yields are likely to remain low because only a small proportion of neonatal hypothalamic neurons express GHRH and at levels much lower than in adults (46). It may be more useful to combine this eGFP approach with suspension or explant culture techniques (47, 48). We present here the first electrophysiological recordings from preidentified GHRH cells in hypothalamic slices. All GHRH neurons showed spontaneous action potential firing, similar to what was reported from other GFP-tagged neuroendocrine neurons (20), although GHRH neurons might be distinguished from proopiomelanocortin neurons (20) and cholecystokinin-sensitive neurons (49) by their lower resting membrane potential. GHRH neurons also showed clusters of spike activity in bursts. We believe the bursting behavior is physiologically relevant because GH release induced by electrical stimulation of the ARC is markedly potentiated with increases in the duration of bursts of firing (50). The GHRH neurons exhibited high input resistance, and internal dialysis with Alexa 350 always stained the single neuron under study, so the soma of GHRH neurons are unlikely to be tightly coupled via gap junctions. In these isolated slices, the clustering of bursts was apparently randomly distributed within and between recordings in individual neurons. However, when held at hyperpolarized potentials, GHRH neurons exhibited frequent and robust inward currents, reflecting intense synaptic activity, and spontaneous depolarizations were routinely observed, triggering action potentials. Further experiments are in progress to elucidate the ionic nature and pharmacological properties of these synaptic events. Although this provided new information about the electrophysiological properties of individual GHRH neurons, relating this to global GHRH episodic secretion requires monitoring of populations of GHRH cells. Because so many of the GHRH cells lie close to the ventral surface of the hypothalamus and send their projections superficially to the outer zone of the ME, we attempted to image a large ventral population of GHRH-eGFP cells in situ simply by external observation of the exposed hypothalamic surface of an intact brain, using an epifluorescence stereomicroscope. A population of individually identifiable eGFP-tagged GHRH cell bodies could readily be imaged in this way, and their fluorescent fibers followed as they coursed to their terminal projection fields in the ME. This should make it possible to perform optically guided recordings of multiunit activity and seek direct evidence for functional connections between identified GHRH neurons in known positions in this network. This imaging approach showed the existence of clusters of GHRH cells with closely apposed cell bodies and that many other more separated GHRH cells seemed to be locally interconnected with visible projections containing GHRHeGFP coursing between them. This suggests that this ventral population of GHRH neurons may form a directly connected homotypic network. Axosomatic contacts between GHRH


cells have been described previously (45), and there is ultrastructural evidence for direct connections between GHRH cells (51), suggesting that network activity could be regulated by GHRH itself (52) or a related product of these cells (53). Other morphological, stimulation, and receptor colocalization studies suggest other candidate neurotransmitters and neuropeptides that could directly affect GHRH cell activity (15, 48, 54 –58). Because it is possible to expose and record from the ventral ARC surface in anesthetized animals, we can now realistically contemplate measuring functional activity in networks of identified GHRH cells in vivo for the first time in the intact brain to gain an understanding of the mechanisms generating episodic GH secretion. Acknowledgments We are extremely grateful to Drs. Jean-Marc Israel, Ste´ phane Oliet, Michel Desarmenien, David Robbe, Robert Gardette, Jacques Epelbaum, Chris Magnus, and Nigel Emptage for considerable help and advice. Received January 3, 2003. Accepted February 21, 2003. Address all correspondence and requests for reprints to: Professor Iain C. A. F. Robinson, Division of Molecular Neuroendocrinology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, United Kingdom. E-mail: [email protected]. Current address for C.B.M.: Department of Neurosurgery, Barts, and The London School of Medicine and Dentistry, Queen Mary, University of London, United Kingdom. Current address for N.B.: Beth Israel Deaconess Medical Center, Harvard Medical School Division of Endocrinology, Boston, Massachusetts. This work was supported by the United Kingdom MRC (to N.B., C.B.M., K.E.M., I.C.A.F.R.), Institut National de la Sante´ et de la Recherche Medicale, and Fondation pour la Recherche Me´ dicale (to P.F.M., A.M., P.M.).

References 1. Rivier J, Spiess J, Thorner M, Vale W 1982 Characterization of a growth hormone-releasing factor from a human pancreatic islet tumour. Nature 300: 276 –278 2. Brazeau P, Vale W, Burgus R, Ling N, Butcher M, Rivier J, Guillemin R 1973 Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 179:77–79 3. Rodier PM, Kates B, White WA, Phelps CJ 1990 Birth dates of the growth hormone releasing factor cells of the rat hypothalamus—an autoradiographic study of immunocytochemically identified neurons. J Comp Neurol 291:363– 372 4. Niimi M, Takahara J, Sato M, Kawanishi K 1989 Sites of origin of growth hormone-releasing factor-containing neurons projecting to the stalk-median eminence of the rat. Peptides 10:605– 608 5. Frohman L, Downs T, Clark I, Thomas GB T 1990 Measurement of growth hormone-releasing hormone and somatostatin in hypophysial portal plasma of unanaesthetized sheep: spontaneous secretion and response to insulininduced hypoglycaemia. J Clin Invest 86:17–24 6. Mayo KE, Miller T, DeAlmeida V, Godfrey P, Zheng J, Cunha SR 2000 Regulation of the pituitary somatotroph cell by GHRH and its receptor. Recent Prog Horm Res 55:237–266 7. Tannenbaum GS, Ling N 1984 The interrelationship of growth hormone (GH)releasing factor and somatostatin in the generation of the ultradian rhythm of GH secretion Endocrinology 115:1952–1957 8. Jansson JO, Albertsson-Wikland K, Eden S, Thorngren KG, Isaksson OGP 1982 Effect of frequency of growth hormone administration on longitudinal bone growth and body weight in hypophysectomised rats. Acta Physiol Scand 114:261–265 9. Gevers EF, Wit JM, Robinson ICAF 1996 Growth, growth hormone (GH)binding protein, and GH receptors are differentially regulated by peak and trough components of the GH secretory pattern in the rat. Endocrinology 137:1013–1018 10. Robinson ICAF, Hindmarsh PC 1998 Importance of the GH secretory pattern for statural growth. In: Kostyo JL, ed. Hormonal control of growth. New York: Oxford University Press; 329 –395 11. Choi HK, Waxman DJ 2000 Plasma growth hormone pulse activation of hepatic JAK-STAT5 signaling: developmental regulation and role in malespecific liver gene expression. Endocrinology 141:3245–3255


12. Davey HW, Xie T, McLachlan MJ, Wilkins RJ, Waxman DJ, Grattan DR 2001 STAT5b is required for GH-induced liver IGF-I gene expression. Endocrinology 142:3836 –3841 13. Mode A, Norstedt G, Simic B, Eneroth P, Gustafsson JA 1981 Continuous infusion of growth hormone feminizes hepatic steroid metabolism in the rat. Endocrinology 108:2103–2108 14. Robinson ICAF 2000 Control of growth hormone (GH) release by GH secretagogues. Novartis Found Symp 227:206 –220 15. Dickson SL, Leng G, Robinson ICAF 1993 Systemic administration of growth hormone-releasing peptide activates hypothalamic arcuate neurons. Neuroscience 53:303–306 16. Poulain DA, Wakerley JB 1982 Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience 7:773– 808 17. Jourdain P, Israel JM, Dupouy B, Oliet SH, Allard M, Vitiello S, Theodosis DT, Poulain DA 1998 Evidence for a hypothalamic oxytocin-sensitive patterngenerating network governing oxytocin neurons in vitro. J Neurosci 18:6641– 6649 18. Bicknell RJ, Leng G 1982 Endogenous opiates regulate oxytocin but not vasopressin secretion from the neurohypophysis. Nature 298:161–162 19. Spergel DJ, Kruth U, Hanley DF, Sprengel R, Seeburg PH 1999 GABA- and glutamate-activated channels in green fluorescent protein-tagged gonadotropin-releasing hormone neurons in transgenic mice. J Neurosci 19:2037–2050 20. Suter KJ, Song WJ, Sampson TL, Wuarin JP, Saunders JT, Dudek FE, Moenter SM 2000 Genetic targeting of green fluorescent protein to gonadotropinreleasing hormone neurons: characterization of whole-cell electrophysiological properties and morphology. Endocrinology 141:412– 419 21. Cowley MA, Smart JL, Rubinstein M, Cerdan MG, Diano S, Horvath TL, Cone RD, Low MJ 2001 Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature 411:480 – 484 22. Zhang BJ, Kusano K, Zerfas P, Iacangelo A, Young 3rd WS, Gainer H 2002 Targeting of green fluorescent protein to secretory granules in oxytocin magnocellular neurons and its secretion from neurohypophysial nerve terminals in transgenic mice. Endocrinology 143:1036 –1046 23. Flavell DM, Wells T, Wells SE, Carmignac DF, Thomas GB, Robinson ICAF 1996 Dominant dwarfism in transgenic rats by targeting human growth hormone (GH) expression to hypothalamic GH-releasing factor neurons. EMBO J 15:3871–3879 24. Magoulas C, McGuinness L, Balthasar N, Carmignac DF, Sesay AK, Mathers KE, Christian H, Candeil L, Bonnefont X, Mollard P, Robinson IC 2000 A secreted fluorescent reporter targeted to pituitary growth hormone cells in transgenic mice. Endocrinology 141:4681– 4689 25. McGuinness L, Magoulas C, Sesay AK, Mathers K, Carmignac D, Manneville J-B, Christian H, Phillips JA, Robinson ICAF 2003 Autosomal dominant growth hormone deficiency disrupts secretory vesicles in vitro and in vivo in transgenic mice. Endocrinology 144:720 –731 26. Bennett PA, Levy A, Sophokleous S, Robinson IC, Lightman SL 1995 Hypothalamic GH receptor gene expression in the rat: effects of altered GH status. J Endocrinol 147:225–234 27. Komminoth P 1992 Digoxigenin as an alternative probe labelling for in situ hybridization. Diagn Mol Pathol 1:142–150 28. Hewson A, Viltart O, McKenzie D, Dyball R, Dickson S 1999 GHRP-6induced changes in electrical activity of single cells in the arcuate, ventromedial and periventricular nucleus neurones of a hypothalamic slice preparation in vitro. Neuroendocrinology 11:919 –923 29. Panchuk-Voloshina N, Haugland R, Bishop-Stewart J, Bhalgat MK, Millard PJ, Mao F, Leung WY, Haugland RP 1999 Alexa dyes, a series of new fluorescent dyes that yield exceptionally bright, photostable conjugates. J Histochem Cytochem 47:1179 –1188 30. Neher E 1992 Correction for liquid junction potentials in patch clamp experiments. Methods Enzymol 207:123–131 31. Fauquier T, Lacampagne A, Travo P, Bauer K, Mollard P 2002 Hidden face of the anterior pituitary. Trends Endocrinol Metab 13:304 –309 32. Spergel DJ, Kruth U, Shimshek DR, Sprengel R, Seeburg PH 2001 Using reporter genes to label selected neuronal populations in transgenic mice for gene promoter, anatomical, and physiological studies. Prog Neurobiol 63: 673– 686 33. Heim R, Tsien RY 1996 Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr Biol 6:178 –182 34. Tsien RY 1998 The green fluorescent protein. Annu Rev Biochem 67:509 –544 35. Giraldi FP, Mizobuchi M, Horowitz ZD, Downs TR, Aleppo G, Kier A, Wagner T, Yun JS, Kopchick JJ, Frohman LA 1994 Development of neuroepithelial tumors of the adrenal medulla in transgenic mice expressing a mouse hypothalamic growth hormone-releasing hormone promoter-simian virus-40 T-antigen fusion gene. Endocrinology 134:1219 –1224 36. Nogues N, Del Rio JA, Perez-Riba M, Soriano E, Flavell RA, Boronat A 1997 Placenta-specific expression of the rat growth hormone-releasing hormone gene promoter in transgenic mice. Endocrinology 138:3222–3227 37. Suhr ST, Rahal JO, Mayo KE 1989 Mouse growth hormone-releasing hormone precursor structure and expression in brain and placenta. Mol Endocrinol 3:1693–1700 38. Balthasar N, Carmignac DF, Mathers K, Bennett PA, Le Tissier PR, Magoulas

2740

39.

40. 41.

42. 43.

44. 45.

46. 47.

48.


C, Robinson ICAF, Growth hormone responses and ghrelin expression in growth hormone secretagogue receptor (GHS-R) transgenic mice. Program of the 82nd Annual Meeting of The Endocrine Society, Toronto, Canada, 2000, p 160 (Abstract 645) Le Tissier P, Magoulas B, Mathers K, McGuinness L, Sesay AK, Carmignac DF, Robinson ICAF, Ablating GRF neurones in transgenic mice using a novel viral ion channel transgene. Program of the 84th Annual Meeting of The Endocrine Society, San Francisco, CA, 2002, p 106 (Abstract OR31-1) Matsubara S, Sato M, Mizobuchi M, Niimi M, Takahara J 1995 Differential gene expression of growth hormone (GH)-releasing hormone (GRH) and GRH receptor in various rat tissues. Endocrinology 136:4147– 4150 Fang S, Steinmetz R, Walker King D, Zeng P, Vogelweid C, Cooper S, Hangcoc G, Broxmeyer HE, Pescovitz OH 2000 Development of a transgenic mouse that overexpresses a novel product of the growth hormone-releasing hormone gene. Endocrinology 141:1377–1383 Phelps CJ, Dalcik H, Endo H, Talamantes F, Hurley DL 1993 Growth hormone-releasing hormone peptide and mRNA are overexpressed in GH-deficient Ames dwarf mice. Endocrinology 133:3034 –3037 Romero MI, Phelps CJ 1997 Identification of growth hormone-releasing hormone and somatostatin neurons projecting to the median eminence in normal and growth hormone-deficient Ames dwarf mice. Neuroendocrinology 65:107–116 Festenstein R, Kioussis D 2000 Locus control regions and epigenetic chromatin modifiers. Curr Opin Genet Dev 10:199 –203 Ibata Y, Okamura H, Makino S, Kawakami F, Morimoto N, Chihara K 1986 Light and electron microscopic immunocytochemistry of GRF-like immunoreactive neurons and terminals in the rat hypothalamic arcuate nucleus and median eminence. Brain Res 370:136 –143 Hurley DL, Wee BE, Phelps CJ 1998 Growth hormone releasing hormone expression during postnatal development in growth hormone-deficient Ames dwarf mice: mRNA in situ hybridization. Neuroendocrinology 68:201–209 Gyevai A, Makara GB, Stark E, Palkovits M 1985 Long-term suspension culture of isolated hypothalamic nuclei of the rat: morphological differentiation and release of substances influencing corticotropin and growth hormone secretion. Neuroscience 14:519 –533 Lanneau C, Peineau S, Petit F, Epelbaum J, Gardette R 2000 Somatostatin


49. 50. 51. 52. 53.

54. 55. 56.

57.

58.

modulation of excitatory synaptic transmission between periventricular and arcuate hypothalamic nuclei in vitro. J Neurophysiol 84:1464 –1474 Burdakov D, Ashcroft FM 2002 Cholecystokinin tunes firing of an electrically distinct subset of arcuate nucleus neurons by activating A-type potassium channels. J Neurosci 22:6380 – 6387 Dickson SL, Leng G, Robinson ICAF 1993 Growth hormone release evoked by electrical stimulation of the arcuate nucleus in anaesthetized male rats. Brain Res 623:95–100 Horvath S, Palkovits M 1988 Synaptic interconnections among growth hormone-releasing hormone (GHRH)-containing neurons in the arcuate nucleus of the rat hypothalamus. Neuroendocrinology 48:471– 476 Poulain P, Carette B 1986 Effects of GRF on paraventricular neurons in slices of guinea-pig hypothalamus. Brain Res 371:355–359 Bloch B, Baird A, Ling N, Guillemin R 1986 Immunohistochemical evidence that growth hormone-releasing factor (GRF) neurons contain an amidated peptide derived from cleavage of the carboxyl-terminal end of the GRF precursor. Endocrinology 118:156 –162 Kiss J, Kocsis K, Csaki A, Gorcs TJ, Halasz B 1997 Metabotropic glutamate receptor in GHRH and beta-endorphin neurones of the hypothalamic arcuate nucleus. Neuroreport 8:3703–3707 Daikoku S, Hisano S, Kawano H, Chikamori-Aoyama M, Kagotani Y, Zhang RJ, Chihara K 1988 Ultrastructural evidence for neuronal regulation of growth hormone secretion. Neuroendocrinology 47:405– 415 Liposits Z, Merchenthaler I, Paull WK, Flerko B 1988 Synaptic communication between somatostatinergic axons and growth hormone-releasing factor (GRF) synthesizing neurons in the arcuate nucleus of the rat. Histochemistry 89:247–252 Bertherat J, Dournaud P, Berod A, Normand E, Bloch B, Rostene W, Kordon C, Epelbaum J 1992 Growth hormone-releasing hormone-synthesizing neurons are a subpopulation of somatostatin receptor-labelled cells in the rat arcuate nucleus: a combined in situ hybridization and receptor light-microscopic radioautographic study. Neuroendocrinology 56:25–31 Delemarre-van de Waal HA, Burton KA, Kabigting EB, Steiner RA, Clifton DK 1994 Expression and sexual dimorphism of galanin messenger ribonucleic acid in growth hormone-releasing hormone neurons of the rat during development. Endocrinology 134:665– 671