Regulated recovery of pulsatile growth hormone secretion from negative feedback: a preclinical investigation

Abstract

Although stimulatory (feedforward) and inhibitory (feedback) dynamics jointly control neurohormone secretion, the factors that supervise feedback restraint are poorly understood. To parse the regulation of growth hormone (GH) escape from negative feedback, 25 healthy men and women were studied eight times each during an experimental GH feedback clamp. The clamp comprised combined bolus infusion of GH or saline and continuous stimulation by saline GH-releasing hormone (GHRH), GHRP-2, or both peptides after randomly ordered supplementation with placebo (both sexes) vs. E(2) (estrogen; women) and T (testosterone; men). Endpoints were GH pulsatility and entropy (a model-free measure of feedback quenching). Gender determined recovery of pulsatile GH secretion from negative feedback in all four secretagog regimens (0.003 ≤ P ≤ 0.017 for women>men). Peptidyl secretagog controlled the mass, number, and duration of feedback-inhibited GH secretory bursts (each, P < 0.001). E(2)/T administration potentiated both pulsatile (P = 0.006) and entropic (P < 0.001) modes of GH recovery. IGF-I positively predicted the escape of GH secretory burst number and mode (P = 0.022), whereas body mass index negatively forecast GH secretory burst number and mass (P = 0.005). The composite of gender, body mass index, E(2), IGF-I, and peptidyl secretagog strongly regulates the escape of pulsatile and entropic GH secretion from autonegative feedback. The ensemble factors identified in this preclinical investigation enlarge the dynamic model of GH control in humans.

Fig. 1.

Joint regulation of pulsatile growth hormone (GH) secretion during negative feedback in 25 adults (14 men, 11 women). Subjects received a single bolus of GH intravenous to enforce feedback and constant intravenous infusion of saline, GH-releasing hormone (GHRH), GH-releasing peptide (GHRP-2), or both peptides (each, 1 μg·kg⁻¹·h⁻¹) to test feedback escape. Pl, placebo; T, testosterone, E₂, estrogen. Pulsatile GH secretion was estimated by deconvolution analysis. Three-way ANCOVA was applied to evaluate main effects of peptide type (top) (P < 0.001), gender (P < 0.001), and sex steroid treatment (P = 0.006), as well as gender × peptide (upper middle), gender × treatment (lower middle) and treatment × peptide (bottom row) interactions. Overall, P was < 0.001 and R² = 0.73 for the model with covariate P < 0.001. Within each row, means with no shared letters differ by post hoc Tukey's multiple-comparison honestly significant difference test. Thus, A differs from B and C, but not from AB or AC. Data are the geometric means ± SE.

Fig. 2.

Dual control of basal (nonpulsatile) GH secretion during feedback recovery in 14 men and 11 women by peptide type (P < 0.001) and sex steroid treatment (P = 0.022) (overall ANCOVA model and covariate P < 0.001, R² = 0.82). The descending rank order of secretagog effects was combined GHRH/GHRP-2>GHRP-2>GHRH>saline (top). There were significant interactions between gender and peptide (P < 0.001, upper middle) and a strong interactive trend between gender and treatment (P = 0.053, lower middle). Basal GH secretion was similar in women and men independently of peptide type across both sex hormone milieus (upper middle), higher in women given E₂ compared with placebo independently of peptide type (lower middle), and comparable in subjects given E₂/T and placebo independently of peptide type and gender (bottom). Superscripts are described in Fig. 1.

Fig. 3.

Triple control of total GH secretion during escape from GH's negative feedback by peptide, gender, and sex steroid treatment. Data are presented as described in the legend of Fig. 1.

Fig. 4.

Stepwise backward elimination multivariate regression analysis of tripartite determination of pulsatile GH secretion by E₂, body mass index (BMI), and IGF-I (overall R² = 0.495, P = 0.006) under placebo/saline stimulation during negative feedback. The three-dimensional plots depict the joint influences of BMI and E₂ (top), BMI and IGF-I (middle), and IGF-I and E₂ (bottom). Standardized regression coefficients (std coeff) are the ratio of the multivariate regression slope to the variable's SD. E₂ and IGF-I positively, and BMI negatively, determined pulsatile GH recovery.

Fig. 5.

Top: joint modulation of GH secretory burst mass by E₂ (positively, P = 0.006) and BMI (negatively, P = 0.022) in 25 adults during negative feedback in the placebo/saline setting (overall model, P = 0.011, R² = 0.34). Bottom: 2-variable control of GH secretory burst number by BMI (negatively, P = 0.039) and IGF-I (positively, P = 0.0043) (overall model, P = 0.013, R² = 0.37). Data are presented as stated in Fig. 4.

Fig. 6.

Gender and E₂ concentrations codetermine GH secretory burst mode time delay to maximal GH release within a burst (overall model, P = 0.002, R² = 0.45). The mode is higher in men than women and increases with E₂ concentrations under GHRP-2 drive of feedback escape. Top: a single statistical outlier (P < 0.001) is marked by a box. Bottom: IGF-I represses and E₂ heightens basal GH secretion in men and women during feedback escape in the E₂/T/GHRH context ) (overall model, P = 0.001, R² = 0.50 with standardized coefficients as noted). IGFBP-1, insulin-like growth factor-binding protein-1.

Fig. 7.

Peptide secretagog (P < 0.001) and E₂/T treatment (P < 0.001) but not gender determine approximate entropy (ApEn) of GH secretion patterns in the feedback-enforced setting. The E₂ and T effects were significant in women (P = 0.003) and men (P = 0.020). During sex steroid supplementation, peptide effects were GHRP-2>GHRH. Paired (men/women) columns differ significantly when no alphabetic characters are shared (e.g., a differs from bc but not from ab).