GH Signaling in Skeletal Muscle and Adipose Tissue in Healthy Human Subjects: Impact of Gender and Age

P F Vestergaard and others Impact of age and gender 171:5 623–631 Clinical Study on GH signaling

GH signaling in skeletal muscle and adipose tissue in healthy human subjects: impact of gender and age

Poul F Vestergaard, Mikkel H Vendelbo, Steen B Pedersen, Anders Juul1, Steffen Ringgard2, Niels Møller, Niels Jessen and Jens O L Jørgensen

The Medical Research Laboratories, Department of Endocrinology and Internal Medicine, Faculty of Health Correspondence Sciences, Institute of Clinical Medicine, Aarhus University Hospital, Aarhus University, Nørrebrogade 44, DK-8000 should be addressed to Aarhus C, Denmark, 1Department of Growth and Reproduction, University Hospital of Copenhagen, Rigshospitalet, P F Vestergaard Blegdamsvej 9, 2100 Copenhagen Ø, Denmark and 2Department of Clinical Medicine, MR Research Centre Aarhus Email University Hospital, Skejby, Brendstrupgaardsvej 100, DK-8200 Aarhus N, Denmark [email protected]

Abstract

Objective: The mechanisms underlying the impact of age and gender on the GH–IGF1 axis remain unclear. We tested the hypothesis that age and gender have impacts on GH signaling in human subjects in vivo. Design: A total of 20 healthy non-obese adults (‘young group’ !30 years (5F/5M) and ‘old group’ O60 years (5F/5M)) were studied after: i) an i.v. GH bolus (0.5 mg) and ii) saline. Methods: Muscle and fat biopsies were obtained after 30 and 120 min. Total and phosphorylated STAT5B proteins, gene

expression of IGF1, SOCS1, SOCS2, SOCS3 and CISH, body composition, VO2max, and muscle strength were measured. Results: In the GH-unstimulated state, women displayed signiﬁcantly elevated levels of CISH mRNA in muscle (PZ0.002) and fat (PZ0.05) and reduced levels of IGF1 mRNA in fat. Phosphorylated STAT5B (pSTAT5b) was maximally increased in all subjects 30 min after GH exposure and more pronounced in women when compared with men (PZ0.01). IGF1, SOCS1, SOCS2, SOCS3, and CISH mRNA expression increased signiﬁcantly in muscle after 120 min in all subjects with no impact of age and gender. GH-induced pSTAT5b correlated inversely with lean body mass (LBM; rZK0.56, PZ0.01) and positively with the CISH European Journal of Endocrinology mRNA response (rZ0.533, PZ0.05). Conclusion: i) GH signaling in muscle and fat after a single GH bolus in healthy human subjects is age independent, ii) we hypothesize that constitutive overexpression of CISH may contribute to the relative GH resistance in women, and iii) experimental studies on the impact of sex steroid administration and physical training on GH signaling in human subjects in vivo are required.

European Journal of Endocrinology (2014) 171, 623–631

Introduction

The activity of the growth hormone (GH)–insulin-like do not seem to involve reduced pituitary responsiveness growth factor 1 (IGF1) axis in human subjects is determined to GH secretagogues or increased sensitivity to the by several factors including age and gender, but the cause– negative feedback by IGF1 stimulation (4). Both spon- effect relationship remains uncertain (1, 2, 3). taneous and stimulated GH secretion in midlife adults are Circulating GH and IGF1 levels start to decline shortly determined by abdominal adiposity rather than chrono- after the completion of puberty and this continues during logical age (5, 6), and weight loss partly restores impaired adulthood and senescence. The underlying mechanisms GH secretion in obesity (7). It is also well known that sex

www.eje-online.org Ñ 2014 European Society of Endocrinology Published by Bioscientiﬁca Ltd. DOI: 10.1530/EJE-14-0538 Printed in Great Britain

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 624 on GH signaling

steroids regulate GH secretion at the pituitary level and test the hypothesis that age and gender have impacts on also modulate peripheral IGF1 generation (2). In adult GH signaling in human subjects. males, testosterone stimulates GH secretion (2) and some but not all studies suggest that testosterone enhances GH-induced IGF1 stimulation (8, 9). Adult females when Subjects and methods compared with males exhibit higher GH levels despite Subjects similar serum IGF1 levels (1, 2, 6, 10), which seems to be at least partly explained by a suppressive effect of estrogen A total of 20 normal-weight subjects (meanGS.E.M., BMI: on hepatic IGF1 production (11). 25G2) comprising ten ‘young’ subjects (5F/5M) with a The physiological and clinical implications of these mean (range) age of 25 (20–28) years and ten ‘old’ subjects age- and gender-dependent differences in GH secretion (5F/5M) with a mean age of 65 (60–72) years participated in and serum IGF1 levels are uncertain and controversial. this study. None were smokers or receiving prescribed It was originally suggested that the senescent decline in drugs, and none of the females received estrogen supple- GH secretion is causally linked to the concomitant mentation. The study protocol was approved by the changes in body composition and physical function and Regional Ethics committee of Denmark (m2010-0121) and that these changes could be reversed by restoration or conducted in agreement with the declaration of Helsinki II. ‘rejuvenation’ of GH levels (12). This simple paradigm has All participants received oral and written information before written informed consent was obtained. Before subsequently been refuted as numerous GH intervention inclusion, each subject underwent physical examination trials showed only marginal beneficial effects but many including plasma measurements of cholesterol, ALAT, side effects (13, 14). In this regard, it is also noteworthy HbA1c, glucose, TSH, creatinine, Hb, and electrolytes. that enhanced IGF1 activity in several species is associated with reduced longevity (15). These ambiguities emphasize the need to improve our understanding of how GH actions Body composition and physical fitness are modulated by age and gender. In that respect, Body composition was assessed by conventional anthro- experimental studies in GH-deficient adult patients pometry and dual-emission X-ray absorptiometry. MR demonstrate that elderly patients are highly responsive spectroscopy was performed using a Signa Excite 1.5 tesla to exogenous GH in terms of both serum IGF1 generation twin-speed scanner (GE Medical Systems) to quantify fat and side effects such as fluid retention and insulin in muscle (IMCL) and liver (IHL) as described previously

European Journal of Endocrinology resistance (16). Moreover, it has been confirmed that (22, 23). The spectra were quantified using the LC model female patients exhibit relative GH resistance at the levels software package (LCModel6.2; Stephen Provencher, of serum IGF1 generation, loss of body fat, stimulation of Oakville, ON, Canada) by means of dedicated muscle bone turnover, and lowering the circulating cholesterol and liver spectroscopy fitting models. The data processing levels (16, 17). However, little is known regarding the provided an estimate of the ratio of lipid to water in the molecular mechanisms underlying these differences in GH tissue within the voxel. action at peripheral target tissue levels. We have pre- Maximal oxygen uptake (VO2max) was measured viously documented that exposure to a physiological GH breath by breath during an incremental bicycle test (Jaeger bolus translates into detectable activation of signaling ER800 bicycle, Erich Jaeger, Hoechberg, Germany). The mechanisms in human skeletal muscle and adipose tissue VO2max test was performed with a pedaling rate of in vivo (18, 19, 20, 21). These studies, however, have only 70 r.p.m.; for the first 3 min the subject worked at 40 W, included young male subjects. then the workload was increased by 3 W every 10 s until In the present protocol, we therefore included females exhaustion. The test subjects were verbally encouraged as well as males, aged either below 30 years or above during the test. The maximal oxygen uptake was after- 60 years, each of whom received an i.v. GH bolus followed wards defined as a mean of a 30 s period during the highest by muscle and fat biopsies for assessment of GH signal oxygen uptake (Oxycon Pro, Erich Jaeger). Finally, transduction including phosphorylated STAT5B and the maximal isometric knee extension was measured using expression of target genes including IGF1, suppressors of the Metitur isometric muscle strength device (Metitur, cytokine signaling 1–3 (SOCS1, SOCS2, and SOCS3), and Jyva¨skyla¨, Finland). All measurements of body compo- cytokine-inducible SH2 protein (CISH), which act as sition and physical performance were performed within feedback inhibitors of GH signaling. This allowed us to 1 month before GH exposure.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 625 on GH signaling

Experimental design 5mMNa4P2O7, 250 mM sucrose, 1% (v/v) Triton X-100, 2 mM dithiothreitol, 0.1 mM benzamidine, 0.5 mM In a single blinded and randomized crossover design, each phenylmethylsulphonyl fluoride, 50 mg/ml soybean subject was studied twice receiving either a single i.v. bolus trypsin inhibitor, and 4 mg/ml leupeptin, pH 7.4, in of 0.5 mg GH (genotropin miniquick, Pfizer) or saline. a Precellys24 homogenizer (Bertin Technologies, At least 14 days elapsed between each study; all five Montigny-le-Bretonneux, France) at 6500 r.p.m. for ‘young’ females were examined 14 days after the first 2!30 s. Samples were rotated for 15 min at 4 8C and menstrual day and again 1 month later. centrifuged at 13 000 r.p.m. for 20 min at 4 8C. The Each subject attended the clinical research unit at supernatant was collected and the protein concentration 0800 h after an overnight fast. The subjects abstained from was determined using the Bradford assay (Protein Assay, strenuous physical exercise 2 days before each experi- #500-0006, Bio-Rad Laboratories, Inc.). mental day. At 0830 h (tZ0), the first blood sample was Adipose tissue was homogenized in a solubilization drawn followed by an i.v. bolus of GH or saline. Muscle buffer containing 20 mM HEPES, 10 mM NaF, 1 mM and fat biopsies were taken at 30 and 120 min after the Na VO , 1 mM EDTA, 5% SDS, 50 mg/ml soybean trypsin Z 3 4 bolus. Blood samples were drawn at t 0, 5, 10, 20, 30, 40, inhibitor, 4 mg/ml leupeptin, 0.1 mM benzamidine, 50, 60, 90, and 120 min. 2 mg/ml antipain, and 1 mg/ml pepstatin in a Precellys24 homogenizer (Bertin Technologies) for 2!30 s at Serum hormones and metabolites 5000 r.p.m. After homogenization, the samples were thermomixed at 1000 r.p.m. for 1 h at 37 8C. The K Serum samples were frozen and stored at 20 8C. Serum infranatant was collected, frozen in liquid nitrogen, and insulin and GH levels were measured using time-resolved stored at K80 8C until analyses. fluoroimmunoassay (TR-IFMA, AutoDELFIA; PerkinElmer, Western blot analyses were performed by SDS–PAGE Turku, Finland), and non-esterified free fatty acids (NEFAs) on Stain-Free 4–15% gels using the CriterionXT-system were analyzed using a commercial kit (Wako Chemicals, (Bio-Rad). Proteins were transferred onto PVDF membranes, Neuss, Germany). Total serum IGF1 was determined by blocked for 2 h in 2.5% skim milk. Antibodies against time-resolved immunofluorometric assay with modifi- phosphorylated STAT5B and total STAT5B from Cell Z cations as published recently (24),att 0, 30, 60, and Signaling Technology were used, Cat. nos: #9359 and 120 min after GH and saline bolus. #9358 respectively (Cell Signaling Technology, Beverly, MA, USA). STAT5B was visualized by ECL chemi- European Journal of Endocrinology Muscle and adipose tissue biopsies luminescence using a ChemiDoc XRS system (Bio-Rad) and quantified using Image Lab (ver. 4.0.1, Bio-Rad). Quantifi- Under sterile conditions, using local anesthesia (Lidokain: cation of STAT5B phosphorylation is expressed as a ratio of Amgros, Copenhagen, Denmark), skeletal muscle and total STAT5B expression measured on the same membranes. adipose tissue biopsies were taken. The muscle specimen Total RNA was isolated from muscle and adipose tissue was obtained from the vastus lateralis of the quadriceps samples using Trizol (Gibco BRL, Life Technologies); RNA was femoris muscle 12–15 cm proximal to the superior border quantified by measuring absorbance at a wavelength of 260 of the patella, using a Bergstro¨m biopsy needle. The and 280 nm with a ratio R1.8 using a NanoDrop 8000 muscle tissue was immediately dissected free from fat spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, and connective tissue and snap-frozen in liquid nitrogen. MA, USA). Integrity of the RNA was checked by visual The adipose tissue was obtained by liposuction from inspection of the two rRNAs, 18S and 28S, on an agarose gel. the periumbilical subcutaneous tissue. The adipose tissue cDNA was synthesized using the Verso cDNA Kit AB 1453 was washed and subsequently snap-frozen in liquid (Thermo Fisher Scientific, Inc.) using random hexamers. Real- nitrogen. time PCR for target genes were carried out with b2-micro- globulin levels as an internal control, and this expression did not change during intervention. The sequences of the used Western blotting and quantitative RT-PCR primers are as follows: IGF1: GACAGGGGCTTTTATTTCAA Muscle biopsies were freeze dried and the remaining fat and CTCCAGCCTCCTTAGATCAC; SOCS1: ACGCACTT- was removed. Freeze-dried muscle tissue (5–6 mg) was CCGCACATTC and CGAGGCCATCTTCACGCTAAG; homogenizedinanice-coldsolubilizationbuffer SOCS2: GGTCGAGGCGATCAGTG and TCCTTGAAGTCA- containing 20 mM Tris, 50 mM NaCl, 50 mM NaF, GTGCGAATC; SOCS3: GGCCACCTGGACTCCTATGA and

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 626 on GH signaling

GCCCTTTGCGCCCTTT; and CISH: GCCCTGAGCCCTGG- transformation was applied and data were retested. TAGTCC and GACACATCCACAGACGGGTGG. Normally distributed data are expressed as meanGS.E.M. The PCRs were performed in duplicate using the Two-way repeated measurements (2 ANOVA) were applied KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Inc., to analyze the effect of GH. If there was a significant Woburn, MA, USA) in a LightCycler 480 (Roche Applied effect of GH, placebo data were removed. Thereafter, Science) using the following protocol: one step at 95 8C for 2 ANOVA was applied to test for different outcomes 3 min, then 95 8C for 10 s, 60 8C for 20 s, and 72 8C for 10 s. in group, age, and sex respectively. Non-parametric The increase in fluorescence was measured in real time statistics was used to test for differences in age, and basal during the extension step. The threshold cycle (CT) was levels of SOCS1, SOCS2, SOCS3, CISH, and IGF1 mRNA calculated, and the relative gene expression was estimated in muscle and adipose tissue (Mann–Whitney rank sum essentially using the default ‘Advanced Relative Quantifi- test), and data are expressed as median (range). Corre- cation’ mode of the software version LCS 480 1.5.0.39 lation was assessed with the Pearson correlation analysis. (Roche Applied Science). P!0.05 was considered significant. Statistics and figures were produced using SigmaPlot version 11.0 (Systat GHR polymorphism Software, Germany).

Genomic DNA was extracted from blood lymphocytes. The frequency of GHR transcript variants with retention (fl-GHR) or exclusion (d3-GHR) of exon 3 was tested by the Results multiplex PCR assay described by Pantel et al. (25). This Body composition and physical performance was performed with primers G1, G2, and G3 (GenBank accession no. AF155912) as follows: initial step of As expected, males weighed more than females, whereas denaturation for 3 min at 95 8C, followed by 25 cycles fat mass was increased in females when compared with consisting of 30 s at 95 8C, 1 min at 64 8C, and 1 min at males (Table 1). In addition, muscle strength reached 72 8C, followed by an extension period at 72 8C for 5 min. higher values in males when compared with females and Amplification of DNA fragments was analyzed by electro- in the young subjects when compared with the older phoresis on a 1% agarose gel stained with ethidium subjects. The amount of IMCL and IHCL fat did not differ bromide. A 935-bp band represented the full-length (fl) significantly between groups. allele (fl-GHR), and a 532-bp fragment represented the d3

European Journal of Endocrinology allele (d3-GHR). We have previously published the Serum levels of GH, IGF1, and NEFAs methodology (26). Following the GH bolus, Cmax (meanGS.E.M.: 93.2G 3.6 mg/l) was recorded after 5 min (Fig. 1a). The serum Statistical analysis GH proﬁle following the GH bolus, as deﬁned by Cmax,

The Shapiro–Wilk test was used to test for normal Tmax, T1/2, and AUC, did not differ according to either age distribution. If data were not normally distributed, log or gender (data not shown). In the GH-unstimulated state

Table 1 Data on body composition and physical ﬁtness in the 20 participants.

Young (nZ10) Old (nZ10) P!0.05

Male (nZ5) Female (nZ5) Male (nZ5) Female (nZ5) Gender Age

Age (years) 25 (20; 27) 25 (23; 28) 66 (63; 69) 65 (60; 72) * Weight (kg) 86G13 72G567G862G8* BMI (kg/cm2)25G224G125G324G3 Fat (kg) 14.8G5 23.0G5 18.1G4 21.3G11 Serum IGF1 (mg/l) 197G19.7 219G24.7 136.6G19.4 122.6G19.8 * Knee force (N) 851G149 440G25 562G40 396G128 * VO2max (mlO2/min per kg) 5712G574 3123G460 2839G292 2855G663 IMCL (AU) 0.006G0.003 0.004G0.001 0.002G0.002 0.005G0.005 IHCL (AU) 1.8G1.6 1.3G0.6 3.2G1.2 1.6G0.3

IMCL, intra-myo-cellular lipid; IHCL, intra-hepatic cellular lipid. For further details, see text. Signiﬁcant differences (*P!0.05) between the ‘young’ group (nZ10) and the ‘old’ group (nZ10) and between males (nZ10) and females (nZ10).

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 627 on GH signaling

(a) (b) expression in muscle (PZ0.002) and a 130% higher 1.4 l/l) 120

µ expression in fat (PZ0.05) (Fig. 4), whereas IGF1 mRNA 100 1.2 P<0.001

g/l) expression in adipose tissue was more pronounced in 80 µ 1.0 60 0.8 males compared with females (PZ0.013), corresponding 40 0.6 to a 83% higher level in males (Fig. 4). The GH-unstimu- 20 0.4

Serum NEFA ( lated expression of SOCS1, SOCS2, and SOCS3 did not 0 0.2 0.0 differ between the groups (data not shown). In muscle Serum GH concentration ( 0 20 40 60 80 100 120 140 0 20406080100 120 140 tissue, a significant GH-stimulated response in IGF1 mRNA Time (min) Time (min) was observed after 120 min in all subjects (PZ0.003) (Fig. 5). No significant gender effect was recorded Figure 1 (PZ0.437). Furthermore, muscle SOCS1, SOCS2, SOCS3, G Mean S.E.M. of serum GH levels (a) and serum non-esterified and CISH mRNA expression increased significantly after fatty acids (NEFAs) (b) after exposure to GH (open circles) and 120 min (P!0.001) (Fig. 5), with no impact of age saline (filled circles). The injection of GH significantly increased (PZ0.950) or gender (PZ0.136). Adipose tissue CISH ! the serum concentration of GH after 5 min (P 0.001) and serum mRNA was significantly increased 120 min after injection NEFAs after 120 min. of GH (P!0.001) without any impact of age (PZ0.157) or gender (PZ0.114) (Fig. 5). We did not record a significant GH-induced increase in mRNA expression of IGF1 (saline injection), spontaneous GH peaks of z10 mg/l were (PZ0.962), SOCS1 (PZ0.386), SOCS2 (PZ0.091), nor observed in three subjects after 20–30 min. Serum IGF1 SOCS3 (PZ0.637) in adipose tissue. levels (mg/l) at baseline were significantly higher in the ‘young’ group when compared with the ‘old’ group ((208G15.3) vs 129.6G13.3 (P!0.001)) without any Correlations impact of gender (PZ0.878), and no significant changes The GH-induced increase in muscle pSTAT5b activity in in serum IGF1 were recorded in response to either GH or all participants at tZ30 min was inversely correlated saline. Serum NEFA levels increased significantly with with LBM (rZK0.558, PZ0.010) and positively correlated time only after the GH bolus, and the peak level occurring with the concomitant increase in muscle CISH mRNA at at tZ120 min was significantly elevated when compared tZ120 min (rZ0.533, PZ0.028). No correlations were with the GH-unstimulated state (P!0.001). observed between the GH-induced elevations in muscle European Journal of Endocrinology pSTAT5b and serum NEFA levels (rZK0.155, PZ0.513), ZK Z ZK STAT5B phosphorylation strength (r 0.326, P 0.186), or VO2max (r 0.261, PZ0.296) respectively. A significant inverse correlation Phosphorylated STAT5B (pSTAT5B) (thy694) in muscle was between IHL and serum IGF1 was observed for the study significantly increased in all subjects at 30 min after GH population as a whole (rZK0.552, PZ0.027). injection followed by a reduction in activity after 120 min. In the GH-unstimulated state, no consistent changes in pSTAT5b activity were recorded (Fig. 2). The response to (a)STAT5 response to saline (b) STAT5 response to GH GH was more pronounced in females when compared 400 ! P<0.001 with males (P 0.001) without a significant impact of age 300 Male (PZ0.132) (Fig. 3a). A significant GH-induced pSTAT5b Female 200 response (PZ0.003) was also recorded in adipose tissue 100 with a maximal activity after 30 min without any significant impact of either age (PZ1.0) or gender 0 Muscle pSTAT5/STAT5 (AU) (PZ0.65) (Fig. 3b). 30 120 30 120 Time (min) Time (min)

IGF1, SOCS, and CISH mRNA levels Figure 2 In the GH-unstimulated state, CISH mRNA expression in Individual changes in pSTAT5b activity relative to total STAT5 in muscle and fat was more pronounced in females when muscle biopsies 30 and 120 min after exposure to saline (a) and compared with males, corresponding to 359% higher GH (b) in females (open triangles) and males (ﬁlled triangles).

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 628 on GH signaling

The pSTAT5b response to GH in adipose tissue was not correlated with body composition, serum NEFA levels, or CISH mRNA (data not shown).

300 (a) Adipose tissue GHR polymorphism 250 A total of 18 subjects were analyzed for GHR poly- 200 morphism in exon 3, of whom 12 were homozygous (fl/fl), four were heterozygous (d3/fl), and two were 150 homozygous (d3/d3). In the analysis, the two d3-GHR 100 groups were pooled, hereafter t-test was applied to test for differences in outcome between the GHR WT and the 50 d3-GHR isoform. No differences were found between the 0 GHR WT and the d3-GHR isoform in any of serum IGF1 Muscle (PZ0.613), muscle pSTAT5b (PZ0.962), adipose tissue 300 pSTAT5b (PZ0.962), NEFA (PZ0.7077), IMCL (b) P<0.001 (PZ0.8131), or IHCL (PZ0.8131).

pSTAT5 increase (%) pSTAT5 250

200 Discussion 150 This study was undertaken to test the hypothesis that GH 100 signaling in human muscle and fat in vivo is determined by age and gender. To this end, muscle and adipose tissue 50 biopsies were obtained in 20 healthy human subjects of 0 both sexes before and after i.v. exposure to either GH or Male Female Young Old saline. In the GH-unstimulated state, the expression of CISH mRNA in muscle and fat was signiﬁcantly elevated (c) Adipose tissue Muscle in females when compared with males, whereas IGF1 Sa30 Sa120 GH30 GH120 Sa30 Sa120 GH30 GH120 mRNA expression in adipose tissue was more pronounced

European Journal of Endocrinology pSTAT5 YM STAT5 in males. GH exposure significantly induced pSTAT5 in both tissues in all subjects, but the response was more pSTAT5 YF pronounced in females without age dependency. This was STAT5 accompanied by a significant increase in the transcription pSTAT5 of canonical GH-dependent genes after 2 h without a OM STAT5 distinct impact of age or gender. Based on our data, we hypothesize that the sexually pSTAT5 OF STAT5 dimorphic pattern of GH activity in human subjects may be causally linked to constitutive overexpression of CISH in females. This is in accordance with the data obtained in Figure 3 primary hepatocytes from male and female rats (27). In the The meanGS.E.M. increases in pSTAT5b activity (percent of latter study, pSTAT5 following episodic GH exposure was maximal response in young males) 30 min after GH exposure in more pronounced in male hepatocytes. In contrast to this, adipose tissue (a) and in muscle tissue (b) relative to the activity we observed increased pSTAT5b in women following a afterwards in the GH-unstimulated state. Representative single exogenous GH bolus. It has, however, also been Western blots of subgroups for total and phosphorylated recorded in vitro that the inhibitory effect of CISH on STAT5B 30 and 120 min after saline (sa) and GH are inserted (c). STAT5b activity is more complete with lower amounts of The activity of pSTAT5b was significantly increased in all groups STAT5b and following more prolonged GH exposure (28). after GH exposure, but the response was significantly higher in It is therefore conceivable but as yet unproven that the females when compared with males (P!0.001). YM, young elevated unstimulated levels of CISH mRNA recorded in male; YF, young female; OM, old male; OF, old female. our female subjects are induced by continuous elevations

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 629 on GH signaling

Baseline any impact of age on GH signaling either at the level of 1000 P=0.002 STAT5B phosphorylation or the expression of IGF1, Male Female SOCS1, SOCS2, SOCS3, and CISH mRNA. We are not P=0.05 aware of published data on the impact of age on GH P=0.013 signaling in any species, but it is well recognized that aging rats as well as human subjects are highly responsive to GH 100 stimulation induced by administration of both exogenous GH and GH secretagogues (3, 31). This could indicate that the senescent decline in GH secretion and serum IGF1 mRNA expression (%) levels is primarily driven by concomitant changes in, e.g. 10 physical activity, dietary habits, and body composition. In CISH IGF1 CISH IGF1 our study, the GH-induced activation of STAT5b in muscle Muscle Adipose tissue was inversely correlated with lean body mass. A similar inverse correlation has been reported between LBM and Figure 4

MeanGS.E.M.ofCISH and IGF1 mRNA expression (percent of maximal response in young males) in muscle and adipose 1000 Adipose tissue tissues after saline exposure (Zbaseline). CISH mRNA expression in muscle was signiﬁcantly elevated in females compared with males (PZ0.002). IGF1 mRNA expression in § adipose tissue was signiﬁcantly elevated in males compared with females (PZ0.013). 100

in endogenous GH levels and that its inhibitory effect on STAT5b activation therefore predominates during chronic 10 GH exposure. We did not measure endogenous GH secretion in our

European Journal of Endocrinology 10 000 subjects, but a recent comprehensive study in 100 healthy Muscle adults of both sexes over a wide range of age reveals that

mRNA expression (%) mRNA expression § § female subjects exhibit higher mean and nadir levels of GH in addition to greater irregularity (29). We furthermore 1000 § § recorded reduced IGF1 mRNA expression in adipose tissue from females when compared with males in the * GH-unstimulated state, which is compatible with relative 100 GH resistance. It is well established that estrogen status is an important determinant of GH secretion although it does not fully account for the sexual dimorphism (2, 11, 29). 10 One mechanism of estrogen in this regard is direct IGF1 CISH suppression of hepatic IGF1 production, where the SOCS1 SOCS2 SOCS3 subsequent decline in serum IGF1 levels is assumed to stimulate GH secretion secondary to reduced feedback Figure 5

inhibition (11). At the molecular level, it has been MeanGS.E.M. increases (percent of maximal response in young reported that estrogen inhibits GH activation of the males) in mRNA expression of IGF1, SOCS1, SOCS2, SOCS3, and JAK/STAT pathway in vitro via stimulation of SOCS2 CISH in muscle and adipose tissue 120 min after GH exposure expression (30). In our study, we did not record differences relative to the GH-unstimulated state. The GH-induced increase in the levels of SOCS2 expression between males and in mRNA expression was signiﬁcant for CISH in adipose tissue females either before or after GH exposure. We did not see and for all genes in the muscle (*P!0.01, §P!0.001).

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 630 on GH signaling

mean 24-h GH concentrations in a group of males and Declaration of interest females aged between 27 and 59 years (6). In the previous The authors declare that there is no conﬂict of interest that could be study, however, visceral adiposity was the major and perceived as prejudicing the impartiality of the research reported. negative determinant of several attributes of GH secretion in both sexes. In this study, we only measured total fat mass by means of DEXA scan, which did not correlate with Funding This study was supported by a research grant from the Novo Nordisk GH signaling. We did, however, observe a signiﬁcant Foundation. inverse correlation between intrahepatic fat and serum IGF1. This is in line with studies on rodents and human

subjects, where intrahepatic fat is determined by reduced Acknowledgements GH secretion and action (32, 33, 34). Moreover, treatment Mrs Else Horneman, Mrs Lone Kvist, and Mrs Merete Møller from Medical of acromegalic patients with a GH antagonist increases Research Laboratory, Department of Clinical Medicine, Aarhus University Hospital, are thanked for their skilled technical assistance. Brian Vendelbo, intrahepatic fat content (22). Department of Growth and Reproduction, University Hospital of Copen- The common exon 3-deleted GH receptor poly- hagen, is thanked for his skilled technical assistance in Ghr genotyping. morphism (d3-GHR) is attributed to enhanced GHR activity both in vitro and in vivo (35, 36), and we therefore found it relevant to evaluate its impact on GH signaling References in vivo. We could not detect a significant difference in GH 1 Ho KY, Evans WS, Blizzard RM, Hochs H, Veldhuis JD, Merriam GR, responsiveness at the level of STAT5b activation when Samojlik E, Furlanetto R, Rogol AD, Kaiser DL et al. Effects of sex and comparing carriers and non-carriers of the supposedly age on the 24-hour profile of growth hormone secretion in man: more active d3-GHR variant. Whether this reflects a low importance of endogenous estradiol concentrations. Journal of Clinical Endocrinology and Metabolism 1987 64 51–58. sample size remains yet to be elucidated. (doi:10.1210/jcem-64-1-51) Certain limitations of this study merit attention. First, 2 Meinhardt UJ & Ho KKY. Modulation of growth hormone action our sample size was relatively small, which increases the by sex steroids. Clinical Endocrinology 2006 65 413–422. (doi:10.1111/j.1365-2265.2006.02676.x) risk of type 2 errors. Secondly, we focused on the impact of 3 Nass R. Growth hormone axis and aging. Endocrinology and a single exogenous GH bolus rather than different patterns Metabolism Clinics of North America 2013 42 187–199. (doi:10.1016/ of exogenous GH exposure or endogenous GH levels, j.ecl.2013.02.001) 4 Hersch EC & Merriam GR. Growth hormone (GH)-releasing hormone and which would have been informative in the context of the GH secretagogues in normal aging: fountain of Youth or Pool of Tantalus? Clinical Interventions in Aging 2008 3 121–129. (doi:10.2147/CIA.S3247)

European Journal of Endocrinology sexual dimorphism of endogenous GH secretion. Thirdly, we only studied skeletal muscle and adipose tissue; in 5 Vahl N, Jorgensen JO, Jurik AG & Christiansen JS. Abdominal adiposity and physicalfitness are major determinants of the age associated declinein particular, it is likely that the liver – also in human subjects stimulated GH secretion in healthy adults. Journal of Clinical Endocrinology – exhibits more pronounced gender differences in GH and Metabolism 1996 81 2209–2215. (doi:10.1210/jcem.81.6.8964853) responsiveness. Nevertheless, we consider our model and 6 Vahl N, Jorgensen JO, Skjaerbaek C, Veldhuis JD, Orskov H & Christiansen JS. Abdominal adiposity rather than age and sex predicts the data it generates to be important for advancing our mass and regularity of GH secretion in healthy adults. American understanding of how GH works in human subjects. Journal of Physiology 1997 272 E1108–E1116. In summary, this study detects distinct gender 7 Rasmussen MH, Hvidberg A, Juul A, Main KM, Gotfredsen A, Hilsted J & Skakkebaek NE. Massive weight loss restores 24-hour growth hormone differences in intracellular GH signaling in adult human release profiles and serum insulin-like growth factor-I levels in obese subjects in vivo. In the unstimulated state, women exhibit subjects. Journal of Clinical Endocrinology and Metabolism 1995 80 1407–1415. (doi:10.1210/jcem.80.4.7536210) elevated expression levels of CISH mRNA in concomitance 8 Gibney J, Wolthers T, Johannsson G, Umpleby AM & Ho KK. with reduced expression levels of IGF1 mRNA. Activation Growth hormone and testosterone interact positively to enhance of STAT5b is detectable in human muscle and adipose protein and energy metabolism in hypopituitary men. American Journal of Physiology. Endocrinology and Metabolism 2005 289 E266–E271. tissue in vivo 30 min after i.v. exposure to an exogenous (doi:10.1152/ajpendo.00483.2004) GH bolus and is more pronounced in women. The 9 Fisker S, Norrelund HF, Juul A, Skakkebaek NE, Christiansen JS & mechanistic link between these two observations remains Jorgensen JO. The growth hormone (GH)-insulin-like growth factor axis during testosterone replacement therapy in GH-treated yet to be characterized. In contrast to gender, age does not hypopituitary males. Growth Hormone & IGF Research 2001 11 104–109. seem to be a significant determinant of stimulated GH (doi:10.1054/ghir.2001.0195) signaling in our model system, adding support to the 10 Jorgensen JO, Christensen JJ, Vestergaard E, Fisker S, Ovesen P & Christiansen JS. Sex steroids and the growth hormone/insulin-like notion that the senescent decline in GH activity is an growth factor-I axis in adults. Hormone Research in Paediatrics 2005 extrinsic property. 64 (suppl 2) 37–40. (doi:10.1159/000087752)

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access Clinical Study P F Vestergaard and others Impact of age and gender 171:5 631 on GH signaling

11 Leung KC, Johannsson G, Leong GM & Ho KK. Estrogen regulation of insulin-like growth factor system and inflammatory biomarkers: a growth hormone action. Endocrinology Reviews 2004 25 693–721. randomized open-label crossover study. BMC Nephrology 2013 14 80. (doi:10.1210/er.2003-0035) (doi:10.1186/1471-2369-14-80) 12 Rudman D, Feller AG, Nagraj HS, Gergans GA, Lalitha PY, Goldberg AF, 25 Pantel J, Machinis K, Sobrier ML, Duquesnoy P, Goossens M & Schlenker RA, Cohn L, Rudman IW & Mattson DE. Effects of human Amselem S. Species-specific alternative splice mimicry at the growth growth hormone in men over 60 years old. New England Journal of hormone receptor locus revealed by the lineage of retroelements Medicine 1990 323 1–6. (doi:10.1056/NEJM199007053230101) during primate evolution. Journal of Biological Chemistry 2003 275 13 Blackman MR, Sorkin JD & Mu¨nzer T. Growth hormone and sex steroid 188664–188669. (doi:10.1074/jbc.M001615200) administration in healthy aged women and men: a randomized 26 Jensen RB, Vielwerth S, Larsen T, Greisen G, Leffers H & Juul A. controlled trial. Journal of the American Medical Association 2002 288 The presence of the d3-growth hormone receptor polymorphism is 2282–2292. (doi:10.1001/jama.288.18.2282) negatively associated with fetal growth but positively associated with 14 Liu H, Bravata DM, Olkin I, Nayak S, Roberts B, Garber AM & postnatal growth in healthy subjects. Journal of Clinical Endocrinology Hoffman AR. Systematic review: the safety and efficacy of growth and Metabolism 2007 92 2758–2763. (doi:10.1210/jc.2007-0176) hormone in the healthy elderly. Annals of Internal Medicine 2007 146 27 Thangavel C & Shapiro BH. Inherent sexually dimorphic expression 104–115. (doi:10.7326/0003-4819-146-2-200701160-00005) of hepatic CYP2C12 correlated with repressed activation of growth 15 Bartke A, Sun LY & Longo V. Somatotropic signaling: trade-offs between hormone-regulated signal transduction in male rats. Drug Metabolism growth, reproductive development, and longevity. Physiological Reviews and Disposition 2008 36 1884–1895. (doi:10.1124/dmd.108.021451) 2013 93 571–598. (doi:10.1152/physrev.00006.2012) 28 Ram PA & Waxman DJ. SOCS/CIS protein inhibition of growth 16 Gasco V, Prodam F, Grottoli S, Marzullo P, Longobardi S, Ghigo E & hormone-stimulated STAT5 signaling by multiple mechanisms. Aimaretti G. THERAPY OF ENDOCRINE DISEASE: GH therapy in adult Journal of Biological Chemistry 1999 274 35553–35561. (doi:10.1074/ GH deficiency: a review of treatment schedules and the evidence for jbc.274.50.35553) low starting doses. European Journal of Endocrinology 2013 168 R55–R66. 29 Veldhuis JD, Roelfsema F, Keenan DM & Pincus S. Gender, age, body (doi:10.1530/EJE-12-0563) mass index, and IGF-I individually and jointly determine distinct 17 Burman P, Johansson AG, Siegbahn A, Vessby B & Karlsson FA. Growth GH dynamics: analyses in one hundred healthy adults. Journal of hormone (GH)-deficient men are more responsive to GH replacement Clinical Endocrinology and Metabolism 2011 96 115–121. (doi:10.1210/ therapy than women. Journal of Clinical Endocrinology and Metabolism jc.2010-1669) 1997 82 550–555. (doi:10.1210/jcem.82.2.3776) 30 Leung KC, Doyle N, Ballesteros M, Sjogren K, Watts CK, Low TH, 18 Jorgensen JO, Jessen N, Pedersen SB, Vestergaard E, Gormsen L, Lund SA Leong GM, Ross RJ & Ho KK. Estrogen inhibits GH signaling by & Billestrup N. GH receptor signaling in skeletal muscle and adipose suppressing GH-induced JAK2 phosphorylation, an effect mediated by tissue in human subjects following exposure to an intravenous GH SOCS-2. PNAS 2003 100 1016–1021. (doi:10.1073/pnas.0337600100) bolus. American Journal of Physiology. Endocrinology and Metabolism 2006 31 Velasco B, Cacicedo L, Melian E, Fernandez-Vazquez G & 291 E899–E905. (doi:10.1152/ajpendo.00024.2006) Sanchez-Franco F. Sensitivity to exogenous GH and reversibility of the 19 Nielsen C, Gormsen LC, Jessen N, Pedersen SB, Moller N, Lund S & reduced IGF-I gene expression in aging rats. European Journal of Jorgensen JO. Growth hormone signaling in vivo in human muscle and Endocrinology 2011 145 73–85. (doi:10.1530/eje.0.1450073) adipose tissue: impact of insulin, substrate background, and growth 32 Ichikawa T, Nakao K, Hamasaki K, Furukawa R, Tsuruta S, Ueda Y, hormone receptor blockade. Journal of Clinical Endocrinology and Taura N, Shibata H, Fujimoto M, Toriyama K et al. Role of growth Metabolism 2008 93 2842–2850. (doi:10.1210/jc.2007-2414) hormone, insulin-like growth factor 1 and insulin-like growth 20 Krusenstjerna-Hafstrom T, Madsen M, Vendelbo MH, Pedersen SB, factor-binding protein 3 in development of non-alcoholic fatty Christiansen JS, Moller N, Jessen N & Jorgensen JO. Insulin and GH liver disease. Hepatology International 2007 1 287–294. (doi:10.1007/ European Journal of Endocrinology signaling in human skeletal muscle in vivo following exogenous GH s12072-007-9007-4) exposure: impact of an oral glucose load. PLoS ONE 2011 6 e19392. 33 Ichikawa T, Hamasaki K, Ishikawa H, Ejima E, Eguchi K & Nakao K. (doi:10.1371/journal.pone.0019392) Non-alcoholic steatohepatitis and hepatic steatosis in patients 21 Moller L, Dalman L, Norrelund H, Billestrup N, Frystyk J, Moller N & with adult onset growth hormone deficiency. Gut 2003 52 914. Jorgensen JO. Impact of fasting on growth hormone signaling and (doi:10.1136/gut.52.6.914) action in muscle and fat. Journal of Clinical Endocrinology and Metabolism 34 Barclay JL, Nelson CN, Ishikawa M, Murray LA, Kerr LM, McPhee TR, 2009 94 965–972. (doi:10.1210/jc.2008-1385) Powell EEM & Waters MJ. GH-dependent STAT5 signaling plays an 22 Madsen M, Krusenstjerna-Hafstrom T, Moller L, Christensen B, important role in hepatic lipid metabolism. Endocrinology 2011 152 Vendelbo MH, Pedersen SB, Frystyk J, Jessen N, Hansen TK, Stodkilde- 181–192. (doi:10.1210/en.2010-0537) Jorgensen H et al. Fat content in liver and skeletal muscle changes in a 35 Dos Santos C, Essioux L, Teinturier C, Tauber M, Goffin V & reciprocal manner in patients with acromegaly during combination Bougneres P. A common polymorphism of the growth hormone therapy with a somatostatin analog and a GH receptor antagonist: receptor is associated with increased responsiveness to growth a randomized clinical trial. Journal of Clinical Endocrinology and hormone. Nature Genetics 2004 36 720–724. (doi:10.1038/ng1379) Metabolism 2012 97 1227–1235. (doi:10.1210/jc.2011-2681) 36 Wassenaar MJ, Dekkers OM, Pereira AM, Wit JM, Smit JW, Biermasz NR 23 Moller L, Stodkilde-Jorgensen H, Jensen FT & Jorgensen JO. Fasting in & Romijn JA. Impact of the exon 3-deleted growth hormone (GH) healthy subjects is associated with intrahepatic accumulation of lipids receptor polymorphism on baseline height and the growth response to as assessed by 1H-magnetic resonance spectroscopy. Clinical Science recombinant human GH therapy in GH-deficient (GHD) and non-GHD 2008 114 547–552. (doi:10.1042/CS20070217) children with short stature: a systematic review and meta-analysis. 24 Reinhard M, Frystyk J, Jespersen B, Bjerre M, Christiansen J, Flyvbjerg A Journal of Clinical Endocrinology and Metabolism 2009 94 3721–3730. & Ivarsen P. Effect of hyperinsulinemia during hemodialysis on the (doi:10.1210/jc.2009-0425)

Received 29 June 2014 Revised version received 7 August 2014 Accepted 26 August 2014

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 08:00:42PM via free access