|Home | About | Journals | Submit | Contact Us | Français|
Objectives were to 1) characterize the relationship of third-ventricle (IIIV) cerebrospinal fluid (CSF) concentrations of growth hormone releasing hormone (GHRH) with concentrations of GH in the peripheral circulation, and 2) assess the influence of acute administration of appetite-regulating peptides, leptin (anti-orexigenic) and neuropeptide Y (NPY; orexigenic) on release of GHRH. Six mature beef cows fitted with IIIV and jugular vein cannulae were treated intracerebroventricularly with saline, and leptin (600 μg) and NPY (500 μg) in saline, in a replicated 3 x 3 Latin square design. Third-ventricle CSF and blood were collected 10 min before and continued 220 min after treatments. Mean concentrations of GHRH and frequency of pulses after treatments were 2.2 ± 0.13 ng/mL and 1.2 ± 0.15 pulses/220 min, respectively. These measures were not influenced by treatments. Concentrations of GHRH in CSF were weakly correlated (r = 0.15; P < 0.03) with serum concentrations of GH; however, 58% of the GH pulses were preceded by a pulse of GHRH and 90% of the GHRH pulses occurred within 20 min preceding a pulse of GH. Leptin tended (P < 0.10) to suppress GH area under the curve (AUC) compared to saline. Concomitantly, NPY tended (P < 0.10) to increase GH AUC, which appeared to be a consequence of increased (P < 0.05) pulse amplitude. Infusion of NPY also increased (P < 0.05) AUC of GHRH relative to saline. No differences were detected among treatments in serum concentrations of IGF-I or its AUC. Sampling CSF from the IIIV appears to be a viable procedure for assessing hypothalamic release of GHRH coincident with anterior pituitary gland secretion of GH in cattle. These data also demonstrate the differential responsiveness of the GH axis to appetite-regulating peptides.
Growth hormone (GH) is involved in many physiological processes, including regulation of cell growth and metabolism. Actions of GH can be direct in target organs or mediated by insulin-like growth factor I (IGF-I) released from hepatocytes in response to GH stimulation [1–2]. Secretion of GH from somatotrophs is regulated by growth hormone-releasing hormone (GHRH) secreted from neuronal terminals adjacent to capillaries of the hypothalamic-hypophyseal portal system in the infundibulum [3–4]. In cattle, the endocrine relevance of GHRH for anterior pituitary gland secretion of GH is demonstrated by observations that concentrations of GH in circulation are reduced in animals immunized against GHRH [5–7]. However, release of GH is also influenced by many other factors of peripheral and hypothalamic origin, including somatostatin, IGF-I, leptin, neuropeptide Y (NPY), ghrelin, and opioids [8–10]. Therefore, due to the complexity of the neuroendocrine regulation of GH release and difficulty of assessing temporal release of neuropeptides in the infundibular region, our understanding of hypothalamic GHRH stimulation of episodic release of GH is limited.
Growth hormone releasing hormone has been assayed in hypophyseal portal blood of sheep [11–13]. In addition, GHRH immunoreactive fibers have been observed to project towards the ependymal layer of the third cerebroventricle (IIIV) in several species [3–4], suggesting that GHRH can be released into the cerebroventricular system. Detection of neurohormones such as gonadotropin-releasing hormone (GnRH) in the cerebrospinal fluid (CSF) of the IIIV has been documented in several species and used as indicator of GnRH release into the hypothalamic-hypophyseal portal circulation [14–15]. The detection and measurement of GHRH in lumbar CSF collected from humans [17–19] is further evidence for potential hypothalamic release of GHRH into the CSF of the third ventricle. Because direct, temporal evaluation of neuroendocrine events is anatomically challenging, a technique involving surgical cannulation of the IIIV would be of great value for investigating hypothalamic-patterns of release of GHRH and anterior pituitary gland secretion of GH into peripheral circulation.
In cattle, leptin acts directly upon the anterior pituitary gland to regulate the release of GH and anterior pituitary responsiveness to GHRH ; however, hypothalamic effects of leptin on the regulation of GH should not be disregarded. More specifically, centrally-administered NPY stimulated GH release  and leptin attenuated NPY-induced increases in circulating GH . Whether the effects of leptin and NPY on GH release in cattle involve modulation of GHRH release is unknown. Using the IIIV cannulation model described above, objectives of this study were to 1) characterize the relationship between concentrations of GHRH in the third-ventricle CSF and GH in circulation and 2) assess the influence of acute administration of two appetite-regulating peptides, leptin (anti-orexigenic) and NPY (orexigenic) on release of GHRH.
All animal experiments reported herein were approved by the Institutional Animal Care and Use Committee of the Texas A&M University System. Six well-fed, ovariectomized, and estradiol-implanted Braford cows (6.5 ± 0.4 yr of age) were each fitted with IIIV cannula using procedures of Gazal et al. . Estradiol implants were designed to produce serum concentrations of estradiol of 2–3 pg/mL . Cows were given at least 1 mo of rest after surgical procedures before initiation of the experiment. At the start of the experiment, cows weighed 489.2 ± 39.4 kg and were in body condition 5.6 ± 0.1 on a scale of 1 to 9 (i.e., 1 = emaciated and 9 = obese).
One day before experiments began, each cow was fitted with a jugular catheter for collection of blood. On each day of the experiment, each cow was placed in a stanchion and polyethylene tubing was inserted through the IIIV cannula to allow collection of CSF and delivery of treatments. Cows were then allowed 1 h of undisturbed rest before treatments were applied. Cows were treated with sterile saline, and NPY (500 μg) and leptin (600 μg) dissolved in saline, in a replicated 3 x 3 Latin Square design (n = 6 and 18 cells of data from the 3 treatments) using dosages and peptides described by Thomas et al. , Morrison et al. , and Amstalden et al. . The randomized blocks of the Latin square were applied at 2-d intervals. Treatments were delivered through IIIV cannula in a 100 μL volume and infused at a rate of 100 μL/min. Subsequently, tubing was flushed with 100 μL saline at this rate to ensure delivery of treatments. Blood (5 mL) and CSF (200 μL) from IIIV were collected simultaneously at 10-min intervals beginning 10 min before treatments and continuing for 220 min after treatments using procedures described by Gazal et al . After treatments were applied, no sampling from the IIIV was performed for a total of 20 min to allow treatments to diffuse into hypothalamic tissue; thus, the 10 min sample was not collected.
Serum concentrations of GH were determined in the samples collected every 10-min using the procedures described by Hoefler and Hallford . Serum concentrations of IGF-I were also determined in these samples using procedures described by Reecy et al. . Intra- and inter-assay coefficients of variation for the two assays were less than 5 and 10%, respectively.
To assay GHRH, a double antibody radioimmunoassay was developed which utilized hGHRH (> 90% homology with bovine; Phoenix Pharmaceuticals, Inc., Burlingame, CA) as the standard and for iodination, polyclonal sheep anti-hGHRH (OBT 1427, AbD Serotec, Oxford, UK) as the primary antiserum, and donkey anti-oIgG (5184-3004, AbD Serotec, Oxford, UK) as the second antibody. A 20-μg solution of hGHRH in deionized water was iodinated using the chloramine T/sodium metabisulfite procedure. The resulting 125I-hGHRH (tracer) was purified by column chromatography using Sephadex G-10 (Amersham Pharmacia Biotech, Uppsala, Sweden). A standard solution was prepared by suspending hGHRH (Phoenix 031-02) at 100 ng/mL in assay buffer (pH 7.5) containing 200 mg protamine sulfate, 4.14 g of sodium phosphate (monobasic), 2.5 g of bovine serum albumin, 200 mg of sodium azide, and 3.36 g of disodium EDTA per liter of deionized water. This stock standard solution was used to provide a standard curve of 0, 0.4, 0.8, 1.6, 3.2, and 6.4 ng hGHRH/tube. A 1:1000 dilution of primary antiserum was prepared in assay buffer. The radioactive tracer was also diluted in assay buffer such that 100 μL contained approximately 20,000 cpm. The second antibody was suspended at a 1:20 dilution in 0.01 M PBS containing 0.05 M EDTA. Briefly, the assay consisted of pipetting standards and CSF samples (usually 0.15 mL) into duplicate borosilicate glass tubes (12 x 75 mm) after which tubes were normalized to 0.4 mL using assay buffer. Each tube then received 0.2 mL of the diluted primary antiserum followed by 0.1 mL of the radioactive tracer. Tubes were then vortexed and incubated overnight at approximately 4º C. On the second day, tubes received 0.2 mL of the diluted second antibody and were again vortexed and incubated overnight at 4º C. On the third day, tubes were centrifuged at 2300 x g, the supernatant was decanted, and the bound fraction was counted for 1 min in a gamma counter. The concentration of GHRH was subsequently calculated using the four parameter logistic method. The anti-hGHRH did not cross-react significantly with the following compounds: corticotrophin releasing hormone, oxytocin, vasopressin, thyroid releasing hormone, GnRH, somatostatin, NPY, or leptin. Addition of 0.05, 0.1, and 0.2 mL of bovine CSF resulted in displacement parallel to the standard curve. Assay sensitivity was 0.3 ng/mL (95% depression). Specific binding of the tracer to the primary antiserum was 47%, and addition of 0.4 ng of hGHRH to the assay resulted in 7% of the tracer being displaced from the antibody. When 2.5 or 5 ng/mL hGHRH was added to a pool of bovine CSF, 95 and 94% were recovered, respectively. To increase sample volume of the CSF samples for this assay, two consecutive 10-min samples were combined; thus, samples were described as collected on 20- min intervals. Each sample was then assayed in duplicate and all samples evaluated within a single assay with a CV of 4%.
Hormone data for each cow on each sampling day were analyzed using Cluster 8 and Pulse algorithms of Pulse_XP software . Programs were parameterized with sampling interval (20 min for GHRH and 10 min for GH), number of points per peak or nadir (1 or 2), t-score for an increase in a pulse or a decrease for a nadir (1), and the half-life of the hormones, which was 10 min for GHRH and 20 min for GH [29–31]. Mean concentration, frequency and amplitude of pulses, and area under the curve (AUC) for GHRH and GH was determined for each treatment of each cow. Because the release of IGF-I is non-episodic, only mean concentration and AUC were evaluated. Two categorical variables were also created to describe pulse patterns of GHRH and GH. These variables were 1) a pulse of GHRH occurred coincident with a pulse of GH (1 = yes and 0 = no) and 2) a pulse of GH occurred within 20 min of a pulse of GHRH (1 = yes and 0 = no).
Statistical analyses were conducted using SAS (SAS Inst. Inc., V9.1, Cary, NC, USA). Hormone data for continuous/numeric response variables within treatments and sampling days were evaluated for heterogeneity of variance with Proc Gplot. These variances were normally distributed. Pearson’s correlations were used to evaluate the association among concentrations of GHRH, GH, and IGF-I. Frequency tables generated with Chi-Square tests were used to evaluate the two categorical variables among treatments (i.e., 1) percentage of GH pulse that were preceded by a pulse of GHRH and 2) pulse of GH occurred within 20 min of a pulse of GHRH). To evaluate the effects of treatment on the numerical variables describing tonic secretion patterns of GHRH, GH, and IGF-I, ANOVA procedures were conducted using Proc Mixed. The model was a prediction of each response variable with treatment. This simple model was derived from two efforts that yielded similar results. The first model included fixed effects of treatment, day, and replicate with the degrees of freedom adjusted with the Satterthwaite method. Day served as the repeated term with cow as the subject. Compound symmetry was determined to be the most appropriate covariance structure. The second model fitted treatment as a fixed effect and day and replicate as random effects. Neither day nor replicate contributed to the two models as significant sources of variation, so the terms were eliminated from the models. When treatment was found to be a significant source of variation, mean separations were conducted using pre-planned pairwise comparisons generated from least-squares procedure (PDIFF). Comparisons were NPY or leptin relative to saline.
Figure 1 illustrates endocrine profiles of GHRH and GH in the 6 cows treated with saline, leptin, and NPY, respectively. Concentrations of GHRH in CSF were weakly correlated (r = 0.15; P < 0.03) with serum concentrations of GH. However, of the GH pulses detected, 58% coincided with a pulse of GHRH and 90% of the GHRH pulses occurred within 20 min preceding a pulse of GH. However, no differences were detected among treatments for these two categorical variables. Note that cow #5 provides an example where pulses of GHRH and GH occurred coincidentally. Cow #653 provides an example where a pulse of GH occurred without a pulse of GHRH, whereas cows #4213 and 9532 provide examples where a pulse of GHRH did not occur coincidentally with a pulse of GH. Finally, no associations were detected between these two hormones and concentrations of IGF-I.
Table 1 depicts mean concentrations of GHRH, GH, and IGF-I and pulsatile characteristics of these hormones among treatment groups. Compared with saline, NPY enhanced (P < 0.05) AUC of GHRH release as well as pulse amplitude of GH. Leptin tended (P < 0.10) to suppress GH AUC relative to saline, whereas NPY tended (P < 0.10) to increase GH AUC. No differences were detected among treatments in serum concentrations of IGF-I or its AUC.
Leshin and co-workers  described immunostaining of GHRH and somatostatin peptides in cattle in the ventrolateral portion of the arcuate nucleus and the median eminence. This report by Leshin et al. , and research in rodents , provided evidence to suggest that GHRH immunoreactive fibers project towards the ependymal layer of the third ventricle, which could release GHRH into the CSF of the IIIV. Cannulating the IIIV in cattle as an approach to evaluate relationships of hypothalamic-releasing hormones and adenohypophysis secretion of hormones has been used to describe the relationship of GnRH and luteinzing hormone [15, 25, 32]. Results herein provide strong evidence to suggest a similar type of relationship for the growth endocrine axis and achieved the first objective of this study, revealing 1 to 2 coincident pulses of GHRH and GH during each 220-min sampling period.
There are several methodologies that can be used to describe relationships of hypothalamic-releasing hormones and anterior pituitary secretion of hormones. These methodologies include hypophyseal portal blood sampling in sheep [11,33,34], intercavernous sinus vein cannulation and sampling in horses [35–36], lumbar sampling of CSF in humans [17–19], and IIIV sampling in goats  and cattle . Skinner and co-workers [14,16] reported that measurement of GnRH in IIIV CSF of sheep was a reliable indicator of neurosecretory activity of the hypophyseal portal system even though it was not as precise as portal blood sampling. Nonetheless, due to the complex anatomical architecture of the cranium, IIIV cannulation appears to be a viable procedure for whole-animal study of hypothalamo-pituitary endocrine relationships in cattle.
Push-pull perfusion of the infundibular region has been used to evaluate patterns of GHRH release in Holstein calves ; however, the study reported no clear correlation between GHRH concentration in perfusates and GH in circulation. Push-pull perfusion techniques pose risk of damaging tissue; thus, it appears that cannulation of the IIIV is an advantageous technique as it will avoid risk of interfering with functioning neuronal networks. In contrast to the push-pull perfusion study of Kasuya et al. , more than half of the GHRH pulses observed in the current study were followed by episodes of GH release. Figure 1 illustrates examples of direct responses as well as discrepancies that can occur between pulses of GHRH and GH. Review articles by MacGregor and Leng , Smith , and McMahon et al.  denote that somatotrophs must experience a stimulation of GHRH followed by an interaction with somastostatin before a subsequent pulse of GH occurs in response to GHRH.
Discrepancies among pulse patterns of GHRH, somatostatin, and GH could be a result of many complex events. Electrophysical regulation of hypothalamic secretion of hormones (i.e., pulse generator) is well documented, but lacks characterization [33,40–41]. Differential expression of genes as discerned by microarray analyses and digital display of transcriptome sequences provide convincing evidence of the need to further explore and categorize the numerous feedback signals and genes involved in tissues such as hypothalamus [42–43]. However, these approaches yield voluminous amounts of data, which require software to visualize gene networks within a system [44–45]. Two well-described proteins that influence the neuroendocrine axes that could be selected by such an approach are NPY and leptin [10,41,46]. In brief, NPY is an orexigenic neuropeptide that is expressed during body weight loss or diet deprivation, and leptin is the anti-orexigenic adipose-derived protein hormone that increases in circulation with adiposity. The physiologic conditions of weight loss or increased adiposity appear to have inverse effects on the growth endocrine axis. Specifically, weight loss increased serum concentration of GH whereas increased adiposity suppressed serum concentration of GH [47–49].
As a second objective of this study, we assessed the influence of acute administration of leptin and NPY on secretion of GHRH and GH. Measuring GHRH in CSF from the IIIV is a novel finding. Moreover, using this methodology as a means to understand physiological mechanisms regulating the growth endocrine axis is of even greater relevance. Specifically, central infusion of NPY into well-fed, ovariectomized, estradiol-treated cows (closed-loop model) stimulated GH secretion, similar to our previous findings . This treatment also increased release of GHRH in the current study. McMahon et al  and Luque et al.  suggested that the effect of NPY on the growth hormone axis may be mediated by interneuron activities other than GHRH (i.e., somatostatin, GABA, etc). Even though it remains unclear as to the exact mechanism through which NPY influenced hypothalamic-release of GHRH in the animal model used in this study, it appears that the site of action may be within the hypothalamus. In previous studies, NPY had no direct effect on GnRH-induced release of luteinizing hormone from bovine anterior pituitary cell cultures .
In contrast to the effect of NPY, leptin decreased area under the GH curve in the current study. This response is in agreement with observations of Zieba et al. , who used adenohypophyseal explants to reveal that a portion leptin action occurs at the level of the anterior pituitary. Also, since leptin pretreatment attenuated NPY-induced increase in GH release in cows , it is possible that leptin’s effects are mechanistically downstream from NPY. Leptin’s influence on hypothalamic-secretion of somatostatin have also been described [52–53]. Additional study, however, is needed to delineate the mechanisms by which NPY and leptin influence the growth endocrine axis (i.e., GHRH versus somatostatin) as well as the impacts and interactions of stimulating neuronal pathways that regulate appetite. In conclusion, results from the current study provide strong evidence to suggest that measurement of bovine GHRH in CSF from IIIV, coincident with GH in the peripheral vasculature, is a viable technique to monitor endocrine communication among the hypothalamus and anterior pituitary gland. Results also suggest that the growth endocrine axis in cattle is responsive to appetite-regulating peptides.
Research supported by the NIH-SCORE (GMO 8136-26) and RISE (GM 61222) programs and the New Mexico Agricultural Experiment Station (Hatch project 180674). Appreciation is expressed to Dr. Marta Remmenga (Department of Experimental Statistics, New Mexico State University) for assistance with statistical analyses. Cattle used in the study were made available through support of USDA-NRI grant 00-35203-9132 and Texas AgriLife Research.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.