European Journal of Endocrinology (2008) 158 479–482 ISSN 0804-4643

CLINICAL STUDY Desmopressin increases IGF-binding protein-1 in humans S B Catrina1,2, R Rotarus2, I R Botusan1,2, M Coculescu2 and K Brismar1 1Department of Molecular Medicine and Surgery, Karolinska Institute, L1:O1, Stockholm S-17176, Sweden and 2Department of Endocrinology, University of Medicine and Pharmacy ‘Carol Davila’, Bucharest 011863, Romania (Correspondence should be addressed to S-B Catrina; Email: [email protected])

Abstract Context: IGF binding protein-1 (IGFBP-1) is essential for IGF-I bioavailability. High levels of IGFBP-1 are encountered in critically ill patients and are a good predictor marker in acute myocardial infarction. The mechanisms responsible for the elevated IGFBP-1 levels in these conditions are still unclear. Interestingly, high levels of vasopressin have been reported in the above-mentioned conditions. Objective: To study the effect of vasopressin on IGFBP-1 in humans. Design: Placebo-controlled cross-over study in patients with central diabetes insipidus (CDI) in whom potential interference from endogenous vasopressin secretion is minimized. After a 3-day desmopressin washout period, each patient received i.v.saline on day 1 and desmopressin (3 mg) on day 2. Blood samples were taken after administration, every 2 h during the whole night, starting at 2000 h. Patients and setting: Fourteen inpatients with CDI in an endocrinology department of a university hospital. Results: Serum IGFBP-1 increased within 4 h after 1-desamino-8-D-arginine vasopressin (DDAVP) by 375G73%, compared with a spontaneous fasting increase by 252G46% following placebo administration (P!0.05). No changes were registered in the levels of either classically regulators of IGFBP-1 (insulin, glucagon, and cortisol) or of IGF-I and glucose. The decrease in plasma osmolarity induced by DDAVP did not precede the increase in IGFBP-1. Conclusions: DDAVP increases serum levels of IGFBP-1. Further investigation is essential to unravel the clinical potential of this interaction in conditions associated with high IGFBP-1 levels.

European Journal of Endocrinology 158 479–482

Introduction the exact players are still unknown. Bearing in mind that vasopressin is a stress hormone and inappropriate high Insulin-like growth factor-binding protein 1 (IGFBP-1) is serum values are encountered in critically ill patients (9) one of six members of the IGFBP family,which specifically we tested a direct interaction between vasopressin and binds and regulates the bioavailability of IGF-I and IGF-II IGFBP-1. Two recent reports have raised even more interest (1). In humans, IGFBP-1 is mainly produced in the liver for the evaluation of a direct relation between AVP and and contributes in short-term metabolic regulation (1). IGFBP-1 as far as each of them was found to be predictor Previous investigations suggested that insulin is the main markers in acute myocardial infarction (AMI) (10, 11). determinant of IGFBP-1 expression through downregula- For this end, we conducted a placebo-controlled cross- tion of the IGFBP-1 hepatic production (2). Several other over study with i.v. administration of desmopressin hormones influence IGFBP-1 production either by (1-desamino-8-D-arginine vasopressin, DDAVP) in 14 stimulating it (e.g., glucagon (3), catecholamines (4), patients with central diabetes insipidus (CDI), who have and cortisol (5)) or by suppressing it (e.g., growth inappropriate low endogenous secretion of vaso- hormone (6)). However, these hormones are able to pressin (AVP). modulate IGFBP-1 serum levels only if the insulin levels We suggest a direct interaction between IGFBP-1 and are low or absent. Nevertheless, several cytokines vasopressin which warrants further investigation in (interleukin (IL)-1b, IL-6, and tumour necrosis factor-a, larger clinical material and in other clinical scenarios TNF-a) are able to stimulate the IGFBP-1 production and where a combination of high vasopressin levels with high overcome the effect of insulin (recently reviewed in (7)). IGFBP-1 levels is encountered, for example, diabetes. The ratio of IGFBP-1 to insulin is increased in highly stressful conditions (critically ill patients) and represents a good predictor marker for mortality as valuable as APACHE Materials and methods II score (8). The loss of the negative regulatory function of Patients insulin at the hepatic level might explain the increased IGFBP-1 in serum. Hypoxia and cytokines (8) were Fourteen inpatients with CDI were included in the study proposed to induce IGFBP-1 in this clinical condition but after giving informed consent. The major outcome of the

q 2008 Society of the European Journal of Endocrinology DOI: 10.1530/EJE-07-0662 Online version via www.eje-online.org

Downloaded from Bioscientifica.com at 09/28/2021 03:39:08PM via free access 480 S B Catrina and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2008) 158 study has been already published (12). Briefly, nine Results patients had complete CDI, while five had CDI partially. CDI was diagnosed by water restriction test followed by i.v. IGFBP-1 injection of 1 mg desmopressin (Ferring, Malmo¨, Sweden) Mean IGFBP-1 levels (2000 h) were 6.4G1.3 and after 3 h of stable urine osmolality,despite a rising plasma 8.8G1.9 mg/l before injection of saline and DDAVP osmolality following fluid restriction (12). respectively.Intravenous injection of 3 mg DDAVPinduced a significant increase in IGFBP-1 during the first 4 h, G Study design which reached a level of 27.1 8.3 mg/l as compared with the spontaneous night increase after saline The study protocol was approved by the Ethical 14.3G3.2 mg/l (P!0.05; Fig. 1). The IGFBP-1 levels Committee of the University of Medicine ‘Carol Davila’ remain higher after DDAVP treatment (43.3G10.1 mg/l) Bucharest. The patients were instructed to discontinue compared with saline (30.5G6.2 mg/l) until the their treatment with desmopressin 3 days before morning (0600 h), however, not statistically significant admission to the clinic. At 1900 h, after a standard (PZ0.076). meal, an indwelling catheter was inserted and at 2000 h each patient received i.v. 1 ml saline in the IGF-I and blood glucose and osmolarity first evening (day 1) and desmopressin (3 mg; Ferring) in G G the second evening (day 2). Blood samples were taken at Mean IGF-I levels were 160.6 40.2 and 177.6 2-h intervals during the whole night. 52.9 mg/l before administration of saline and DDAVP respectively (PZNS). The serum levels did not change over time 177.6G52.9 and 187.4G57.8 mg/l (2200 h), Assays 202.5G55.5 and 209.3G83.3 mg/l (2400 h) after treatment with saline and DDAVP respectively (Fig. 2A). The IGFBP-1 levels were measured in the serum using a No significant changes were found in blood glucose polyclonal antibody that detects both the phosphory- levels during the study: 3.7G0.42 vs 4.6G0.7 mmol/l lated and non-phosphorylated forms. The detection (2000 h), 3.6G0.4 vs 4.1G0.5 mmol/l (2200 h), 4G limit was 3 ng/ml with inter- and intra-assay coeffi- 0.4 vs 4.1G0.3 mmol/l (2400 h) after saline and cients of variation (CVs) of 3 and 10% respectively (2). DDAVP infusion respectively. DDAVP at a concentration up to 4 mg/ml did not cross- We did not observe any difference in basal osmolarity react in the IGFBP-1 assay. levels between the 2 days of study (297.17G1.92 The IGF-I serum levels were measured by an RIA after vs 292.29G1.97 mOsm/kg). On the other hand, acid–alcohol extraction and cryoprecipitation as DDAVP decreased the serum osmolarity (287.72G described (2). The detection limit was 6 ng/ml with 1.99 mOsm/kg at 2400 h). intra- and intra-CVs of 4 and 11% respectively. The insulin levels were measured by an RIA commercial kit Glucagon, insulin, and cortisol (Pharmacia & Upjohn Diagnostics AB) with a detection limit of 2 mU/ml and intra- and inter-CV of 5.8 and Desmopressin compared with saline had no effect on the 6.4% respectively. serum levels of either cortisol or insulin. A spontaneous Glucagon concentrations were measured by an RIA commercial kit (Linco Research, St Charles, MI, USA) with a detection limit of 20 pg/ml and intra- and inter- CV of 4.6 and 13.4% respectively. Serum cortisol levels were determined by fluorescence immunoassay (Auto- DELFIA, Wallac Oy, Turku, Finland). Intra- and inter- assay coefficients of variation were 4.9 and 6.4% respectively. Blood glucose was determined by glucose oxidase method using YSI 2300 Stat Plus Analyzer.

Statistical analysis The difference between the levels of the hormones measured after treatments was analyzed by repeated measurement of variance using Statistica 7.1 (StatStoft ! Inc., Tulsa, OK, USA). Differences of P 0.05 were Figure 1 The serum levels of IGFBP-1 after injection of DDAVP considered significant. The data are represented as (3 mg; filled circles) or saline (open circles). Data are expressed as meanGS.E.M. percentage (meanGS.E.M.) from value at 2000 h (*P!0.05).

www.eje-online.org

Downloaded from Bioscientifica.com at 09/28/2021 03:39:08PM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2008) 158 DDAVP induces IGFBP-1 481

Figure 2 The serum levels of (A) IGF-I, (B) cortisol, (C) glucagon, and (D) insulin during the first 4 h after infusion of DDAVP (3 mg) or saline. Box: meanGS.E.M., whisker 95% CI (*P!0.005 different hours versus 2000 h inside the same treatment). decrease in the mean concentration of both cortisol agonist for V2 subtype receptors, has been shown to have a (Fig. 2B) and insulin (Fig. 2D) occurred during the first similar affinity for V1b receptors (14). hours of follow-up (P!0.05). The mean concentration of The main source of circulating IGFBP-1 in males and glucagon did not significantly change and there was no non-pregnant women is the liver where it is produced by difference between groups (Fig. 2C). hepatocytes (1) that express V1a and V1b receptors (15) but not V2. Thus, DDAVP might induce IGFBP-1 production through a direct liver effect mediated via a Discussion V1b-dependent pathway. In agreement with this, in vitro, DDAVP induced an increased IGFBP-1 secretion by We report here for the first time that DDAVP increases HepG-2 cells (human hepatic carcinoma cell line; data IGFBP-1 blood levels. The study was performed on not shown). We cannot exclude that DDAVP stimulates patients with CDI, in a clinical setting where a potential IGFBP-1 from other sources as well, such as the kidneys, confounding effect from endogenous secretion of AVP is even though the renal production may not contribute decreased. The increase in IGFBP-1 after DDAVPis specific substantially to the circulating IGFBP-1 level (1). and is not due to a change in water balance, which would We did not observe a modulation of classical have had an opposite effect through dilution. regulators of IGFBP-1 after administration of DDAVP. Vasopressin acts via three subtypes of receptors that are Nevertheless, due to the sampling schedule, we could classified according to the second messenger system to not completely reject the possible contribution of some which they are coupled: V1a and V1b (also known as V3) of them. For instance, cortisol has been reported to be receptors linked to the phosphoinositol signaling pathway stimulated by DDAVP (16). However, the DDAVP effect and V2 receptors linked to the adenylate cyclase signaling on cortisol levels lasts less than an hour thus making pathway (13). DDAVP, originally developed as a selective this explanation less relevant (16). We even observed a

www.eje-online.org

Downloaded from Bioscientifica.com at 09/28/2021 03:39:08PM via free access 482 S B Catrina and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2008) 158 tendency of reduction of the cortisol levels over time 4 Fernqvist-Forbes E, Hilding A, Ekberg K & Brismar K. Influence of following DDAVP injection, even though statistically circulating epinephrine and norepinephrine on insulin-like this was non-significant. This might reflect the dilution growth factor binding protein-1 in humans. Journal of Clinical Endocrinology and Metabolism 1997 82 2677–2680. effect of the free water retained after DDAVP injection. 5 Conover CA, Divertie GD & Lee PD. Cortisol increases plasma DDAVP induces also production of catecholamines that, insulin-like growth factor binding protein-1 in humans. Acta in turn, might result in increased IGFBP-1 serum levels Endocrinologica 1993 128 140–143. (17). However, it has previously been reported that the 6 Norrelund H, Fisker S, Vahl N, Borglum J, Richelsen B, effect of catecholamines on IGFBP-1 lasts !120 min Christiansen JS & Jorgensen JO. Evidence supporting a direct suppressive effect of growth hormone on serum IGFBP-1 levels. after inoculation (4). Furthermore, DDAVP in doses Experimental studies in normal, obese and GH-deficient adults. similar to those used in the present study did not Growth Hormone and IGF Research 1999 9 52–60. influence epinephrine but only norepinephrine level 7 Baxter RC. Changes in the IGF-IGFBP axis in critical illness. Best (17), which has only a marginal effect on IGFBP-1 (4). Practice and Research. Clinical Endocrinology and Metabolism 2001 We could not detect any difference in the blood levels of 15 421–434. glucagon or insulin after injection of DDAVP compared 8 Mesotten D, Delhanty PJ, Vanderhoydonc F, Hardman KV, Weekers F, Baxter RC & Van Den Berghe G. Regulation of with saline. Even though there is a clear evidence that insulin-like growth factor binding protein-1 during protracted vasopressin is able to stimulate the secretion of both critical illness. Journal of Clinical Endocrinology and Metabolism hormones through V1b receptors (18), it is not surprising 2002 87 5516–5523. that we did not register it, taking into account that the 9 Jochberger S, Mayr VD, Luckner G, Wenzel V, Ulmer H, Schmid S, effect is transitional and lasts for less than 60 min. Knotzer H, Pajk W, Hasibeder W, Friesenecker B, Mayr AJ & Dunser MW. Serum vasopressin concentrations in critically ill Injection of DDAVP did not change the concen- patients. Critical Care Medicine 2006 34 293–299. trations of total IGF-I or glucose. Although IGFBP-1 is 10 Khan SQ, Dhillon OS, O’Brien RJ, Struck J, Quinn PA, an acute regulator of IGF-I bioavailability (1), short- Morgenthaler NG, Squire IB, DaviesJE, BergmannA & Ng LL. C-ter- term changes in its blood levels are not always followed minal provasopressin (copeptin) as a novel and prognostic marker in by changes in glucose or in total or free IGF-I levels (19). acute myocardial infarction: Leicester Acute Myocardial Infarction In conclusion, we demonstrated that DDAVP Peptide (LAMP) study. Circulation 2007 115 2103–2110. 11 Wallander M, Norhammar A, Malmberg K, Ohrvik J, Ryden L & increases the blood levels of IGFBP-1 suggesting a Brismar K. Insulin-like growth factor binding protein 1 predicts novel mechanism to explain the high levels of IGFBP-1 cardiovascular morbidity and mortality in patients with acute observed in critically ill patients or patients with myocardial infarction and type 2 diabetes. Diabetes Care 2007 30 diabetes mellitus. Further investigations are needed to 2343–2348. confirm the clinical relevance of our finding. 12 Catrina SB, Rotarus R, Wivall IL, Coculescu M & Brismar K. The influence of vasopressin deficiency and acute desmopressin admin- istration on melatonin secretion in patients with central diabetes insipidus. Journal of Endocrinological Investigation 2004 27 47–51. Acknowledgements 13 Jard S. Vasopressin receptors. A historical survey. Advances in Experimental Medicine and Biology 1998 449 1–13. We thank nurses Constanta Navodaru, Eugenia Mircea, 14 Saito M, Tahara A & Sugimoto T. 1-desamino-8-D-arginine and Micsunica Ilie for their care of the patients, Inga-Lena vasopressin (DDAVP) as an agonist on V1b vasopressin receptor. Biochemical Pharmacology 1997 53 1711–1717. Wivall for exceptional technical assistance and Agneta 15 Saito M, Sugimoto T, Tahara A & Kawashima H. Molecular Hilding for statistical advice. We also thank the Family cloning and characterization of rat V1b vasopressin receptor: Stefan Persson Foundation, von Kantzow Foundation, evidence for its expression in extra-pituitary tissues. Biochemical Thurings Foundation, Tore Nilson Foundation, and Elin and Biophysical Research Communications 1995 212 751–757. Danielsson for their financial support. 16 Scott LV, Medbak S & Dinan TG. ACTH and cortisol release following intravenous desmopressin: a dose–response study. Clinical Endocrinology 1999 51 653–658. 17 Grant MB, Guay C & Lottenberg R. Desmopressin stimulates References parallel norepinephrine and tissue plasminogen activator release in normal subjects and patients with diabetes mellitus. Thrombosis 1 Lee PD, Giudice LC, Conover CA & Powell DR. Insulin-like growth and Haemostasis 1988 59 269–272. factor binding protein-1: recent findings and new directions. 18 Yibchok-Anun S, Cheng H, Heine PA& Hsu WH. Characterization of Proceedings of the Society for Experimental Biology and Medicine receptors mediating AVP- and OT-induced glucagon release from the 1997 216 319–357. rat pancreas. American Journal of Physiology 1999 277 E56–E62. 2 Brismar K, Fernqvist-Forbes E, Wahren J & Hall K. Effect of insulin on 19 Frystyk J, Nyholm B, Skjaerbaek C, Baxter RC, Schmitz O & the hepatic production of insulin-like growth factor-binding protein- Orskov H. The circulating IGF system and its relationship with 1 (IGFBP-1), IGFBP-3, and IGF-I in insulin-dependent diabetes. 24-h glucose regulation and insulin sensitivity in healthy subjects. Journal of Clinical Endocrinology and Metabolism 1994 79 872–878. Clinical Endocrinology 2003 58 777–784. 3 Hilding A, Brismar K, Thoren M & Hall K. Glucagon stimulates insulin-like growth factor binding protein-1 secretion in healthy subjects, patients with pituitary insufficiency, and patients with insulin-dependent diabetes mellitus. Journal of Clinical Endocrinology Received 11 January 2008 and Metabolism 1993 77 1142–1147. Accepted 17 January 2008

www.eje-online.org

Downloaded from Bioscientifica.com at 09/28/2021 03:39:08PM via free access