Growth Hormone as a Model for Reversible Activation of Adipose Tissue Fibrosis

Background and preliminary data: Adipose tissue is a multicellular tissue surrounded by an extracellular matrix, which undergoes continuous remodeling. Pertubations in the remodeling processes may cause accumulation of excess extracellular matrix protein and hence fibrosis. Adipose tissue fibrosis is recognized as a component of the metabolic syndrome together with insulin resistance, dyslipidemia and obesity, and fibrosis is likely to play a causative role (1,2). In this context, it is fascinating that prolonged GH exposure in vivo induces insulin resistance despite a concomitant mobilization and reduction of fat mass (3). This effect of GH is expressed in patients with a GH-producing pituitary tumor (acromegaly) (4). Moreover, GH is a potent activator of collagen turnover and it also promotes fibrosis in human tendons and skeletal muscles (5-7). Increased AT fibrosis has been reported in a GH transgenic mice model (8) and we have preliminary data showing AT fibrosis in patients with active acromegaly, which reverses after disease control.

Little is known about the mechanisms underlying GH-induced fibrosis, but recent evidence points to a potential involvement of FAPα, an enzyme that is highly expressed in mouse AT FAP cells (9). Moreover, we have recently reported that human skeletal muscle FAP cells upregulate FAPα (DPP4) during fibrogenic differentiation (19). FAPα is a subunit of a heterodimeric proteinase complex attached to the cell membrane in addition to a soluble form also present in the circulation (10). Several proteins are recognized as FAPα substrates, including collagen type I (11) and III (12), and FAPα appears to play a significant role in hepatic tissue remodeling (13) and in lung fibrosis (14). We have recently recorded elevated circulating levels of FAPα in active acromegaly, which correlates with collagen turnover reverses after disease control (15). Fibro-adipogenic progenitor cells are mesenchymal progenitors with the intrinsic potential to differentiate into either collagen-producing fibroblasts or adipocytes. They have been studied in murine cardiac and skeletal muscle, where they contribute to either fibrosis or fat deposition during muscle-impaired regeneration or degeneration (16-18). We have recently demonstrated that a subset of FAP cells drives the accumulation of ECM protein and adipocytes in the muscle from patients with type 2 diabetes and likely contributes to the poor metabolic and mechanical muscle function (19). Whether GH affects adipose tissue FAP cell proliferation and differentiation is unknown, but FAP cell proliferation is regulated by IGF-I (16), which is a strongly GH-dependent peptide. Increased FAP proliferation has also been reported to contribute to intramuscular adipose tissue (IMAT) in several conditions, and we have observed IMAT after treatment in acromegaly (unpublished data). Furthermore, we have preliminary data from FACS-isolated adipose tissue FAP cells incubated with serum from acromegaly patients, which suggest GH-dependent increased FAP cell proliferation and fibrogenic appearance. Collectively, these findings suggest that GH promotes a pro-proliferative and fibrogenic FAP phenotype at the expense of adipogenic differentiation.

Hypotheses: Growth hormone: 1) Activates FAPα protein expression, 2) Increases proliferation and fibrogenic differentiation of FAP cells, and 3) Induces reversible fibrosis in adipose tissue in humans

Subjects and methods: In a single blinded, randomized, double-dummy crossover design, 10 adult, moderately overweight individuals will be subjected to one week of GH and GH receptor blockade (Pegvisomant). Pegvisomant is a modfied GH molecule that selectively blocks the GH receptor and is a licensed drug for the treatment of acromegaly. We include Pegvisomant as an 'active control' in order to suppress endogenous GH actions. The participants will receive daily subcutaneous injections of growth hormone, 0.6-2.0 mg depending on age, for 7 days in the GH intervention. In the control intervention, the participants will receive daily subcutaneous injections of either Pegvisomant or saline. Pegvisomant in a dose of 30 mg is given two times, in the beginning and in the end of the control intervention, whereas saline is given on the other 5 days of the control intervention period. The two intervention periods are separated by a wash out period of 1-4 months. The participants will be randomized to either start with the GH intervention and next be subjected to the control intervention, or start with the control intervention and next be subjected to the GH intervention. The participants will meet at the hospital daily for the injections and a small blood sample. Each intervention period is initiated by an initiation day where there will be taken blood samples, adipose tissue and muscle samples, be performed temperature measurements and bioimpedance, and be administered heavy water and the first injection of intervention either GH or control intervention. On the first initiation day a DXA scan will be performed for assessment of body composition. Each intervention period will be terminated with a study day where there will be taken blood samples, adipose tissue and muscle samples, be performed temperature measurements, indirect calorimetry, palmitate tracer kinetics and bioimpedance, and be administered the last injection of intervention either GH or control intervention. The participants will be fasting for the initiation and study days, and during the intervention periods, they will log their intake of food and beverages.

Study outcomes:

Primary: FAP cell function, FAPα protein expression and markers of fibrosis in AT biopsies obtained before and after intervention. In particular, we will perform:

• FACS to quantify and isolate cell populations, including quantification of FAP cells in adipose tissue, and in vitro determination of proliferation and fibro-/adipogenic differentiation potential
• FAPα expression in blood and adipose tissue
• Markers of fibrosis in adipose tissue assessed by light microscopy and immunohistochemically, RNA sequencing and heavy water labeled connective tissue turnover

Secondary: to study the impact of GH exposure on:

• Circulating biomarkers of collagen turnover (PINP, PIIINP)
• Whole body energy metabolism and fatty acid turnover (indirect calorimetry and palmitate tracer kinetics)
• Connective tissue turnover in muscle tissue (heavy water (D2O) labeling)
• Temperature measurements

Statistical analysis plan: Comparison between groups will be performed using standard statistical methods (t-test or equivalent nonparametric test). Within group comparison will be performed using paired t-test of equivalent nonparametric test. Moreover, ANOVA (repeated measures) will be used. A p-value less than 0.05 will be regarded as statistically significant.

Perspective and relevance: This is the first study to investigate GH effects on adipose tissue fibrosis in humans, which has implications beyond GH pathophysiology. A deeper understanding of the pathways controlling fibroblast and adipocyte balance in fat tissue is essential groundwork and may unravel new targets for combating adipose tissue dysfunction and related disorders. As recently demonstrated in type 2 diabetic patients these progenitor cells are key mediators of tissue plasticity and function in humans (19).