Growth Hormone, Insulin-Like Growth Factor-1, and the Kidney

Preparing to load PDF file. please wait...

0 of 0
100%
Growth Hormone, Insulin-Like Growth Factor-1, and the Kidney

Transcript Of Growth Hormone, Insulin-Like Growth Factor-1, and the Kidney

REVIEW

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

Growth Hormone, Insulin-Like Growth Factor-1, and the Kidney: Pathophysiological and Clinical Implications
Peter Kamenický, Gherardo Mazziotti, Marc Lombès, Andrea Giustina, and Philippe Chanson
Assistance Publique-Hôpitaux de Paris (P.K., M.L., P.C.), Hôpital de Bicêtre, Service d’Endocrinologie et des Maladies de la Reproduction, Centre de Référence des Maladies Endocriniennes Rares de la Croissance, Le Kremlin Bicêtre F-94275, France; Univ Paris-Sud (P.K., M.L., P.C.), Faculté de Médecine Paris-Sud, Le Kremlin Bicêtre F-94276, France; Inserm Unité 693 (P.K., M.L., P.C.), Le Kremlin Bicêtre F-94276, France; and Department of Clinical and Experimental Sciences (A.G., G.M.), Chair of Endocrinology, University of Brescia, 25125 Brescia, Italy
Besides their growth-promoting properties, GH and IGF-1 regulate a broad spectrum of biological functions in several organs, including the kidney. This review focuses on the renal actions of GH and IGF-1, taking into account major advances in renal physiology and hormone biology made over the last 20 years, allowing us to move our understanding of GH/IGF-1 regulation of renal functions from a cellular to a molecular level. The main purpose of this review was to analyze how GH and IGF-1 regulate renal development, glomerular functions, and tubular handling of sodium, calcium, phosphate, and glucose. Whenever possible, the relative contributions, the nephronic topology, and the underlying molecular mechanisms of GH and IGF-1 actions were addressed. Beyond the physiological aspects of GH/IGF-1 action on the kidney, the review describes the impact of GH excess and deficiency on renal architecture and functions. It reports in particular new insights into the pathophysiological mechanism of body fluid retention and of changes in phospho-calcium metabolism in acromegaly as well as of the reciprocal changes in sodium, calcium, and phosphate homeostasis observed in GH deficiency. The second aim of this review was to analyze how the GH/IGF-1 axis contributes to major renal diseases such as diabetic nephropathy, renal failure, renal carcinoma, and polycystic renal disease. It summarizes the consequences of chronic renal failure and glucocorticoid therapy after renal transplantation on GH secretion and action and questions the interest of GH therapy in these conditions. (Endocrine Reviews 35: 234–281, 2014)

I. Introduction A. GH and IGF-1: their receptors and intracellular signaling B. Anatomic and functional segmentation of the nephron
II. Molecular Bases of GH and IGF-1 Action in the Kidney A. GHR expression in the kidney B. Local IGF-1 and IGF-2 synthesis in the kidney C. IGF-1R expression in the kidney
III. The GH/IGF-1 Axis and Renal Physiology A. GH/IGF-1 in kidney growth and development B. GH/IGF-1 and glomerular function C. GH/IGF-1 and tubular function
IV. Renal Consequences of GH Hypersecretion A. Renal hypertrophy B. Changes in glomerular function C. Pathophysiology of body fluid retention D. Changes in phospho-calcium metabolism
V. Renal Consequences of GH Deficiency A. Consequences for kidney size
ISSN Print 0163-769X ISSN Online 1945-7189 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received July 18, 2013. Accepted November 15, 2013. First Published Online December 20, 2013

B. Changes in glomerular function C. Changes in body fluid homeostasis D. Changes in phospho-calcium metabolism VI. The GH/IGF-1 Axis in Renal Diseases A. GH/IGF-1 and diabetic nephropathy B. GH/IGF-1 in renal impairment C. GH/IGF-1 in kidney transplantation D. GH/IGF-1 and renal cancer E. GH/IGF-1 and polycystic kidney disease VII. Conclusion
I. Introduction
The kidney plays a central homeostatic role by adapting renal excretion of fluids and electrolytes to bodily needs. The kidney maintains a stable composition of the
Abbreviations: ANP, atrial natriuretic peptide; BNP, brain natriuretic peptide; ENaC, epithelial sodium channel; GFR, glomerular filtration rate; GHR, GH receptor; GOAT, ghrelin O-acyltransferase; GW, gestational week; IGFBP, IGF binding protein; IGF-1R, IGF-1 receptor; JAK2, Janus kinase 2; MRI, magnetic resonance imaging; NADPH, nicotinamide adenine dinucleotide phosphate; Na-Pi, sodium-phosphate; NKCC2, renal-specific Na-K2Cl cotransporter; RAAS, renin-angiotensin-aldosterone system; rh, recombinant human; SDS, SD score; Sgk1, serum- and glucocorticoid-induced kinase 1; STAT5, signal transducer and activator of transcription 5; STZ, streptozotocin; TmPO4, maximum tubular phosphate reabsorption rate; WT, wild-type; ZEB2, zinc finger E-box-binding homeobox 2.

234 edrv.endojournals.org Endocrine Reviews, April 2014, 35(2):234 –281

doi: 10.1210/er.2013-1071

doi: 10.1210/er.2013-1071

edrv.endojournals.org 235

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

Figure 1.

dependent actions (6, 7). IGF-1 may

also be synthesized independently of

A

B

GH, under the control of other regulatory factors (8). In a given tissue

or cell, GH and IGF-1 may act either

synergistically, as illustrated by their

bone-growth-promoting properties,

or antagonistically, on hepatic glu-

cose metabolism for instance (8).

Considerable advances have been

made in our understanding of the rel-

ative contributions of GH and IGF-1

to the regulation of physiological

processes such as somatic growth,

through both genetic manipulations

in mice and studies of human patho-

physiological situations.

At the cellular level, GH actions

are mediated by its receptor, GH re-

ceptor (GHR), expressed on the

plasma membrane of GH target cells

Figure 1. A, Schematic representation of a nephron. PCT, proximal convoluted tubule; PST, proximal straight tubule; thin DL, thin descending limb of Henle’s loop; thin AL, thin ascending limb of Henle’s loop; MTAL, medullary thick ascending limb of Henle’s loop; CTAL, cortical thick ascending limb of Henle’s loop; DCT, distal convoluted tubule; CD, connecting duct; CCD,

(9). GHR is a member of the cytokine receptor superfamily (10). GH binding to its receptor activates intracel-

cortical collecting duct; OMCD, outer medullary collecting duct; IMCD, inner medullary collecting duct. B, Schematic representation of the glomerulus and macula densa.

lular signaling pathways, of which the Janus kinase-2 (JAK2)/signal

transducer and activator of tran-

extracellular compartment despite large quantitative and qualitative variations in dietary intake of water and solutes (1). This vital function is controlled by complex intrarenal and neurohumoral regulatory mechanisms (2). In parallel to their growth-promoting properties, GH and IGF-1 regulate a broad spectrum of biological functions in several organs, including the kidney. This review focuses

scription 5 (STAT5) and ERK1/2 pathways are the most prominent (11–13). IGF-1, like IGF-2, acts through the IGF-1 receptor (IGF-1R) belonging to the family of tyrosine-kinase receptors. IGF-1 binding activates canonical intracellular signaling cascades such as the ERK1/2 and phosphatidylinositol 3kinase/AKT pathways, reviewed in Refs. 14 and 15).

on the renal actions of GH and IGF-1, taking into account advances made since this issue was last reviewed in this journal nearly 20 years ago (3). The first three sections are devoted to molecular and physiological aspects of GH and IGF-1 actions on the nephron. The following two sections describe the impact of GH excess and deficiency on renal architecture and functions. The final part summarizes the involvement of the GH/IGF-1 axis in major renal diseases.

B. Anatomic and functional segmentation of the nephron
A brief description of renal functional segmentation is necessary before addressing the main focus of this review. Interested readers will find detailed information on renal architecture and physiology in nephrology textbooks.
Composed of approximately 1 million nephrons, the human kidney ensures body-fluid homeostasis by sequen-

A. GH and IGF-1: their receptors and

tial blood filtration in the glomeruli and through highly

intracellular signaling

regulated transport processes that take place in the suc-

GH plays an essential role in normal postnatal growth cessive tubular segments. The renal glomerulus (Figure 1),

and development and regulates a wide variety of other a small ball of capillaries (0.1 mm diameter) through

biological functions, including intermediate metabolism which the blood is filtered, is composed of a capillary

and homeostasis. Most GH actions are mediated by network formed by a thin layer of fenestrated endothelial

growth factors induced by GH in target tissues, of which cells, a central region of mesangial cells and matrix, a layer

IGF-1 is physiologically the most relevant and the most of visceral epithelial cells (podocytes), and a layer of pa-

extensively studied (4, 5). GH also has direct, IGF-1-in- rietal epithelial cells, the latter two layers forming Bow-

236 Kamenický et al GH, IGF-1, and Kidney

Endocrine Reviews, April 2014, 35(2):234 –281

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

man’s capsule. Podocytes are dynamic cells characterized by extensive interdigitating foot processes containing actin filaments. They are involved in several important cellular and physiological processes governing glomerular function, such as glomerular filtration, maintenance of the glomerular basement membrane, regulation of capillary shape and integrity, and signal transduction. In humans, approximately 170 L of plasma is ultrafiltrated daily in the “urinary space” between these two layers. The filtration barrier between blood and the urinary space is composed of a fenestrated endothelium, basement membrane, and pores between podocyte foot processes (16).
Nearly all the filtered fluid and NaCl are reabsorbed along the nephron, only 1% being excreted in urine. This necessitates a high degree of functional coordination to prevent excessive losses of fluid and electrolytes. In contrast, catabolic end products and xenobiotics are almost entirely excreted. The role of the renal tubule is thus to adjust the quantity and composition of excreted urine to maintain an appropriate composition of the interior milieu. Schematically, the renal tubule can be subdivided in three functional parts: the proximal tubule, Henle’s loop, and the distal tubule.
The proximal tubule begins at the urinary pole of the glomerulus and consists of an initial convoluted portion (proximal convoluted tubule) and a straight portion (proximal straight tubule), located in the medullary ray. The proximal tubule reabsorbs nearly all the glucose, amino acids, and phosphate; more than 70% of the water, sodium, potassium, and chloride; and a large fraction of other solutes filtered in the glomeruli (17–20). The thin descending limb of Henle’s loop begins at the boundary between the outer and inner stripes of the outer medulla and continues after the hairpin turn into the thin ascending limb. The transition between the thin and thick ascending limbs of Henle’s loop forms the border between the inner and outer medulla. The thick ascending limb expands through the medulla (medullary segment) and the cortex (cortical segment) to the glomerulus of the nephron of origin, where it ends in a specialized region called the macula densa. Henle’s loop reabsorbs 20 –30% of the sodium, 30% of the calcium, and up to 70% of the magnesium filtered in the glomeruli (21). The corticomedullar osmotic gradient initiated by the Henle’s loop and maintained by countercurrent circulation in vasa recta is essential for urine concentration. By exerting tubuloglomerular feedback, the macula densa adjusts glomerular filtration and thereby maintains a constant urinary flow to the distal nephron (22). The distal tubule begins beyond the macula densa as a distal convoluted tubule and is connected by the connecting tubule to the collecting duct. Expanding from the outermost cortex to the tip of the renal papilla, the

collecting duct is usually subdivided into three parts: the cortical collecting duct, the outer medullary collecting duct, and the inner medullary collecting duct. The distal nephron reabsorbs less than 10% of filtered electrolytes but plays an essential role in fine-tuning hydroelectrolytic transport, itself regulated by several hormonal systems, including aldosterone and vasopressin (22). Figure 1 represents the topological organization of the nephron. The segmental organization and functional complexity of the nephron should be kept in mind when envisaging GH/ IGF-1 actions in the kidney.
II. Molecular Bases of GH and IGF-1 Action in the Kidney
Functional GHRs are a prerequisite for direct GH actions in the kidney. Circulating and locally produced IGF-1 also requires IGF-1R to exert its biological actions. The choice of the experimental model is important because the nephron is composed of several morphologically distinct and functionally specialized cell types (2). Several rat, mouse, and human renal tissues, including the whole kidney, separated kidney zones, microdissected nephron segments, and renal cell lines, have been used to examine the anatomical pattern of GHR and IGF-1R receptor expression at both mRNA and protein levels.
A. GHR expression in the kidney Successful cloning of the human GHR in 1987 (9) en-
abled studies of its anatomical distribution. GHR was found to be expressed in most tissues (10). However, the complexity of the renal architecture hindered precise analysis of the GHR expression profile along the nephron.
Mathews et al (23), using an RNA probe in a solution hybridization assay, were the first to report GHR mRNA expression in a variety of rat tissues. Although GHR transcripts were most abundant in the liver, substantial amounts of GHR transcripts were also detected in the whole kidney, reaching adult levels 5 weeks after birth (23). Northern blotting identified two isoforms of GHR transcripts in rat kidney (24): a long transcript encoding membrane-bound GHR, and a shorter transcript generated by alternative splicing of the primary transcript encoding GH binding protein (GHBP) in rodents (25). Chin et al (26) analyzed the topology of GHR mRNA expression in the rat kidney by in situ hybridization during fetal development and adulthood. GHR mRNA was detected from embryonic day 20 in the outer stripe of the outer medulla and in the medullary rays, suggesting that GHR is mainly expressed in the proximal straight tubules. The abundance of transcripts increased until postnatal day 40,

doi: 10.1210/er.2013-1071

edrv.endojournals.org 237

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

when steady-state adult levels were reached. In adult animals, GHR mRNA was also most abundant in the proximal straight tubule, with lower levels in the medullary thick ascending limb of Henle’s loop (26). A subsequent in situ hybridization study of rat kidney, using a probe specifically recognizing GHR but not GHBP, confirmed this distribution (27).
In contrast to this restricted transcript expression, immunohistochemical mapping of GHR and GHBP proteins in rat kidney revealed substantial immunoreactivity in all nephronic segments, with the strongest signals in the distal convoluted tubule and the collecting duct, and a very weak signal in the glomeruli (28). Using the same anti-GHR antibody, the same group examined GHR expression in early second-trimester (14- to 16-wk) human fetal tissues and reported positive staining of epithelial cells in the proximal and distal tubules, mainly at the apical membrane (29). Simard et al (30) further described GHRspecific immunostaining in the human fetal kidney as early as 8.5 to 9 weeks and found that most renal tubule epithelial cells became positive by week 13. The staining was stronger in the outer medulla than in the cortex and remained similar at midgestation and after birth. Interestingly, weak staining was also found in immature glomeruli in early gestation but disappeared at later developmental stages (30), suggesting specific GH involvement in glomerular morphogenesis. Because transgenic mice overexpressing human or bovine GH (contrary to mice overexpressing IGF-1) develop progressive glomerulosclerosis (31), GH was proposed to have direct, IGF-1-independent effects on glomerular functions, even in adult animals. More recent observations based on the use of highly sensitive and specific quantitative real-time RT-PCR and more efficient and selective antibodies have extended GHR expression to mesangial cells (31, 32), as well as to podocytes in both mice and humans (33). The abundance of GHR transcripts in isolated murine glomeruli was similar to that found in the liver (33). While GHR expression was well established in the proximal straight tubule and in the medullary thick ascending limb of Henle’s loop, it remained controversial in the distal parts of the nephron. The first evidence of GHR expression in the collecting duct came from functional studies showing a rise in IGF-1 mRNA levels after GH exposure (34). More recently, we reported the expression profiles of GHR in isolated murine nephronic segments, based on quantitative real-time PCR. Besides the well-described expression of GHR in the proximal tubule and Henle’s loop, we detected substantial amounts of GHR mRNA in the distal nephron, with a descending expression gradient from the cortex to the medulla. Transcript levels were approximately 10 –20 times higher in the proximal tubule (104 molecules/mm of tu-

bule) than in downstream segments. In a cortical collecting duct cell line obtained by targeted oncogenesis in transgenic mice, we confirmed the presence of GHR at both the transcript and protein levels and showed that GHR mRNA levels increased during renal cell differentiation (35).
During the last decade, modern genomic techniques have been used to obtain exhaustive characterization of the renal transcriptome, thereby expanding our knowledge of gene networks involved in kidney functions (36). However, to the best of our knowledge, no genomic or proteomic studies have examined GHR expression along the nephron. Serial analysis of gene expression in isolated human nephronic segments established a high-resolution map of the human kidney (37). No significant amounts of GHR mRNA were detected in any of the nephron segments, most likely owing to extremely weak GHR expression, representing a major limitation of these global molecular approaches (personal communication from Jean-Marc Elalouf, CEA Saclay).
There have been few studies of the regulation of GHR expression in nonhepatic tissues. GHR expression is generally controlled by multiple factors, including sex steroids, glucocorticoids, nutrients, and GH itself (38 – 41). Results concerning homologous regulation of GHR levels by GH are conflicting and largely dependent on the time of GH exposure and the concentration used (32, 39). For example, hypophysectomy reduced GHR mRNA levels in rat tissues, including the kidney, whereas GH supplementation restored them in some studies (24, 26, 42), but not others (23). Treatment of dwarf dw/dw rats with bovine GH even resulted in a reduction in GHR and GHBP mRNA levels in the kidney (43). In human mesangial cells, GHR transcripts were either up- or down-regulated depending on the GH concentrations used (32). The regulation of GHR expression by GH involves, at least in the liver, binding of the transcription factor STAT5a to a canonical STAT response element located in the GHR promoter (44).
GHR activation initiates a cascade of intracellular signaling leading to various cell-specific biological responses. A detailed description of GHR-activated intracellular signaling pathways is available in Refs. 12 and 13. Functional integrity of GHR signaling, including the canonical JAK2/ STAT5 and ERK1/2 pathways, has been demonstrated in some renal cell lines (33, 35).
B. Local IGF-1 and IGF-2 synthesis in the kidney Renal IGF-1 originates from two different sources: 1)
circulating IGF-1, mainly synthesized in the liver, accounts for 75% of circulating IGF-1 in experimental animals (45, 46) and acts in an endocrine manner on target tissues when extracted from blood and bound to the cell surface by IGF

238 Kamenický et al GH, IGF-1, and Kidney

Endocrine Reviews, April 2014, 35(2):234 –281

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

binding proteins (IGFBPs); and 2) IGF-1 synthesized locally in kidney acts as an autocrine or paracrine regulatory factor for renal cell metabolism (47, 48). IGF-1 levels are higher in renal venous blood than in renal arterial blood, suggesting renal IGF-1 biosynthesis (49). In their seminal contributions, D’Ercole et al (48) demonstrated that ovine GH administration to hypophysectomized rats increased the amount of extractable IGF-1 in several tissues, including the whole kidney, thus pioneering the concept of paracrine/autocrine IGF-1 effects. Further studies showed that GH treatment increased IGF-1 mRNA levels in the kidney of hypophysectomized rats (50, 51), confirming local renal IGF-1 production.
The spatiotemporal pattern of IGF-1 expression in the kidney is unclear. Local IGF-1 synthesis usually takes place in connective tissue cell types such as stromal cells and may act on adjacent (epithelial) cells that do not actually synthesize IGF-1 (8). In human fetal kidney, IGF-1 mRNA was not detected in the nephrogenic zone, including the glomeruli and tubules, whereas IGF-1 immunostaining was positive in the epithelium of proximal and distal tubules, probably reflecting sequestration of this peptide from the circulation (52). Bortz et al (34), by applying immunohistochemistry and a soluble hybridization assay to rat isolated nephronic segments, identified IGF-1 peptide and IGF-1 mRNA in the collecting ducts and inferred that local synthesis of IGF-1 took place in this nephronic segment. The same group showed elevated IGF-1 immunolabeling in collecting ducts in two rat models of acromegaly (53), as well as increased IGF-1 mRNA levels in rat isolated renal collecting ducts after in vitro incubation with GH (54). In situ hybridization mapping in the rat kidney showed an alternative pattern of IGF-1 mRNA distribution, exclusively in the medullary thick ascending limb of Henle’s loop (26, 55), long considered the only site of IGF-1 synthesis along the nephron (3). Studies of IGF-1 expression during mouse kidney development revealed IGF-1 mRNA expression in all renal cells at embryonic day 15, with a drastic decrease after birth. Interestingly, 2 weeks after birth, IGF-1 mRNA was only found in the peritubular capillaries of the outer medulla and inner cortex (56). In addition, no IGF-1 mRNA was detected either at baseline or after GH treatment in cultures of differentiated podocytes (33) and principal cells of the cortical collecting duct (35), despite the functional integrity of GHR and its signaling pathways. Thus, IGF-1 synthesis has been clearly demonstrated in renal connective tissue but is questionable in renal epithelial cells (8). The local tissue availability and distribution of IGF-1 is regulated by six high-affinity IGFBPs. The renal expression profile of IGFBPs is summarized in Ref. 57.

IGF-2 plays an important role during embryonic and fetal development, but its function after birth has not been fully elucidated (5, 58). IGF-2 mRNA is strongly expressed in the rat and mouse fetal kidney and falls markedly after birth, except in blood-vessel endothelial cells (56, 59). GH increases IGF-2 levels only minimally compared to IGF-1 levels (8, 58).
C. IGF-1R expression in the kidney Initial evidence for a direct IGF-1 action in the kidney
came from studies showing that prolonged treatment with recombinant human (rh) IGF-1 increased kidney size in hypophysectomized rats (60) and enhanced the glomerular filtration rate (GFR) in healthy men (61). Molecular cloning of genes encoding the IGF-1R a and b subunits enabled studies of their tissue expression and ontogenesis (62, 63). The first studies were performed in rats by in situ hybridization. During early embryogenesis (E14-E15), the receptor mRNA is expressed in the rat mesonephros (59, 63). In adult rat kidney, IGF-1R mRNA was detected throughout the nephron, including the glomerulus, the thick ascending limb of Henle’s loop, and along the distal nephron and collecting duct, with the lowest levels in the proximal tubules (26, 55, 64). In the human kidney, the pattern of IGF-1R gene expression is very similar to that found in rats, with strong expression in glomeruli and the tubular epithelium of the medulla, whereas IGF-1R transcripts are barely detectable in proximal tubules (65, 66). Lindenberg-Kortleve et al (56) used in situ hybridization and quantitative RT-PCR to investigate IGF-1R expression during mouse kidney development. IGF-1R mRNA levels were highest during the initial period of metanephric development, with transcripts being detected throughout the kidney, whereas their levels declined during further development, being lowest in the postnatal period. In proximal tubules, the receptor was expressed until birth. In the mature mouse kidney, the distribution of IGF-1R mRNA largely matched that previously reported in rats and humans (56). We used RT-PCR to quantify the expression profile of IGF-1R in freshly isolated murine nephronic segments obtained by microdissection and detected substantial amounts of IGF-1R transcripts throughout the tubular system (Ͼ104 molecules/mm of tubule). In contrast to previous studies, we found the strongest expression (Ͼ105 molecules/mm of tubule) in the proximal tubule (35). Global analysis of gene expression in human isolated nephronic segments did not reveal significant amounts of IGF-1 mRNA in any of the portions (37). As already stated, this was probably due to the limited sensitivity of the transcriptomic approach.
IGF-1R mRNA expression is modulated by a number of physiological and pathophysiological stimuli (15). Fast-

doi: 10.1210/er.2013-1071

edrv.endojournals.org 239

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

ing, for instance, increases IGF-1 binding and IGF-1R mRNA abundance in rat tissues, including the kidney (67). Experimental diabetes in rats also induces an increase in renal IGF-1R mRNA levels (68), likely contributing to the renal hypertrophy observed in diabetic nephropathy (15). Unilateral nephrectomy in immature rats leads to a significant increase in IGF-1R mRNA in the remaining kidney, mediating its compensatory growth (69). Although an increase in IGF-1 levels usually reduces IGF-1R levels in cell lines (15), hypophysectomy did not significantly affect IGF-1 mRNA levels in rat kidney (55). In murine cortical collecting duct cells, IGF-1R expression increased with cell differentiation (35). Several studies of renal cell lines have shown the functional integrity of IGF-1R-associated signaling (35, 70, 71).
III. The GH/IGF-1 Axis and Renal Physiology
A. GH/IGF-1 in kidney growth and development Given that GH and IGF-1 play critical roles as regula-
tors of renal development, growth, and function, it was anticipated that animal models of gene inactivation, as well as pathophysiological models, would provide important new insights into the mechanisms and role of the GH/ IGF-1 axis in renal organogenesis. We will first briefly summarize renal ontogenesis in mammals, focusing on humans, to facilitate understanding of the potential impact of these growth factors on renal development. The pronephros emerges during the first 3 gestational weeks (GWs) and disappears after 4 GWs, leading to the development of the mesonephros, which itself vanishes during the fourth month of gestation. Metanephronic development occurs from the fifth GW when the first nephrons start to appear, becoming functional around the eighth GW. Nephronic development takes place in a centrifugal manner from the medulla to the renal cortex. Twenty percent of nephrons are formed at 3 months gestation, whereas nephronic development is achieved by the 34th GW, with approximately 106 nephrons per kidney (72).
Consistent with the pivotal role of GH in IGF-1 secretion, GHR knockout leads to dwarfism in mice, but the body weight deviation only became apparent at 3 weeks of age. Interestingly, at variance with most other organs that were proportionally smaller and allosterically scaled, even after normalizing to body weight, the kidneys of GHRϪ/Ϫ mice were smaller than those of controls (73), indicating that GH and its renal receptor exert important and unique actions on renal development. However, to our knowledge the renal histology of these animals has not been directly assessed, and there are no detailed reports of their development or renal function. Compelling evidence for a

direct role of GH in renal growth came from the unambiguous demonstration, in uninephrectomized model mice, that compensatory renal hypertrophy was directly dependent on GH-induced, locally secreted IGF-1, as revealed by blunted responses in the presence of a GHR antagonist (74).
IGF-1Ϫ/Ϫ mice do not exhibit major renal abnormalities (75). Except for a massive reduction in body weight at birth, associated with 95% perinatal lethality, surviving homozygous mutant newborns have a proportional reduction in kidney weight (76). Subtle abnormalities of nephrogenesis are observed, with smaller glomeruli and a 20% reduction in the number of glomeruli in IGF-1Ϫ/Ϫ mice (77). Although IGF-1 appears to be required for normal murine embryonic growth (78), its absence was not crucial for survival, thus pointing to the existence of another, partially redundant, overlapping growth factor required for metanephric development.
As mentioned in other sections of this review, the relative growth contributions of circulating IGF-1 and of locally produced, GH-driven IGF-1 are poorly understood. Using a very elegant, specific knockout mouse model in which the major GH signaling mediator JAK2 was specifically invalidated in the liver, Nordstrom et al (79) demonstrated that hepatic IGF-1 production was crucial for GH-mediated kidney mass stimulation, suggesting that locally produced renal IGF-1 had little or no effect on kidney growth, as opposed to skeletal muscle for instance. It has also been shown that both IGF-1 and IGF-2 are produced locally by the metanephros, thus participating in renal development (80). Results obtained in several rodent models show the importance of the GH/IGF-1 axis in renal growth during ontogenesis and development. In sharp contrast, little attention has been paid to kidney growth in human disorders associated with defective GH/IGF-1 signaling.
B. GH/IGF-1 and glomerular function Our understanding of glomerular structure and func-
tion in physiological and pathophysiological conditions has improved markedly in recent decades, especially the functional properties of the glomerular barrier (16). In humans, close to 170 L of primary urine is produced each day. This clearly requires fine regulation of glomerular functions and the participation of various hormonal systems.
More than 50 years ago, Corvilain et al (81) demonstrated in humans that short-term treatment with GH increased the GFR. This GH action is due to an IGF-1-mediated decrease in renal vascular resistance, leading to increased glomerular perfusion (61, 82– 86). The decreased renal vascular resistance is due to an IGF-1-in-

240 Kamenický et al GH, IGF-1, and Kidney

Endocrine Reviews, April 2014, 35(2):234 –281

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

duced decrease in both afferent and efferent arteriolar resistance (83). The effect of IGF-1 on peripheral resistance requires nitric oxide release and cGMP signaling (87), as well as cyclooxygenase metabolites such as vasodilating prostaglandins (3). In addition to increasing glomerular perfusion, GH and IGF-1 augment extracellular volume and plasma volume (88), thereby also contributing to increased glomerular filtration.
Recently, a direct IGF-1-independent action of GH on the glomerulus was demonstrated by Reddy et al (33), who detected GHR mRNA and protein by real-time PCR, immunohistochemistry, and Western blot analysis in murine podocyte cells (MPC-5) and murine kidney glomeruli. GH treatment of both murine and human podocytes was associated with increases in STAT5, JAK2, and ERK1/2 proteins. Exposure of podocytes to GH modified the intracellular distribution of JAK2 SH2-B and Janus kinase-2 adapter protein Src homology 2-B and also stimulated focal adhesion kinase expression, increased reactive oxygen species production, and induced reorganization of the actin cytoskeleton (33). By using microarray and quantitative RT-PCR analyses of immortalized human podocytes, the same group showed that zinc finger E-boxbinding homeobox 2 (ZEB2) was up-regulated in a GH dose- and time-dependent manner, and that the increased ZEB2 expression was partly related to an increase in the expression of a ZEB2 natural antisense transcript. The same authors showed that the GH-dependent increase in ZEB2 expression resulted in a loss of P- and E-cadherins in podocytes, thus explaining the increased podocyte permeability to albumin (89).
The GH/IGF-1 axis exerts effects on all the component cells of the glomerulus. GH and especially IGF-1 stimulate mesangial cell proliferation and migration. The physiological role of GH and IGF-1 in podocyte function is complex because dual effects of these hormones have been observed. In a model of hyperhomocysteinemia-induced podocyte dysfunction, GH inhibited the epithelial-to-mesenchymal transition (a crucial event leading to glomerulosclerosis) by blocking nicotinamide adenine dinucleotide phophate (NADPH) oxidase activation (90). IGF-1 inhibited podocyte apoptosis by concomitant stimulation of cell migration and differentiation (91). Negative effects of GH on podocyte function were also described, with increased production of reactive oxygen species that affected reorganization of the podocyte actin cytoskeleton and increased the permeability of the filtration barrier (33). These latter effects were subsequently implicated in the pathogenesis of baseline and exercise-induced proteinuria occurring in clinical disorders such as diabetes mel-

litus (92) and acromegaly, although hemodynamic factors could be more important than structural abnormalities in this latter condition (93).
C. GH/IGF-1 and tubular function Over the last two decades, major efforts have been
made to describe the transport properties of the kidney tubule and to identify regulatory factors at the cellular and molecular levels (20). Cloning of membrane transporters and hormone receptors, involved in solute transport and its hormonal control, and deciphering of intracellular signaling pathways were essential prerequisites for fine analysis of these highly regulated transport processes. GH and IGF-1 are involved in hormonal fine-tuning of tubular handling of sodium, water, and phosphate and, to a lesser extent, other electrolytes, and are also known to regulate tubular gluconeogenesis (3). This section will describe the principal physiological actions of GH and IGF-1 in the renal tubule, examine the relative contributions of GH and IGF-1, and clarify the tubular topology of transport processes and their underlying molecular mechanisms.
1. Regulation of renal sodium reabsorption The kidney tubule plays a pivotal role in regulating
body fluid homeostasis and blood pressure by adjusting renal excretion of sodium and water to dietary intake (1). The GH-IGF-1 system has long been recognized as a hormonal modulator of renal tubular sodium and water reabsorption (88). Soon after the isolation and purification of GH in 1944 (94), the sodium-retaining properties of extracted GH were clearly demonstrated in rats (95) and healthy volunteers (96, 97). A large number of metabolic studies in rodents and uncontrolled studies in humans have analyzed the effects of chronic and acute GH administration on sodium and water homeostasis, confirming initial observations (88). The impact of GH and IGF-1 on extracellular volume and sodium balance is revealed in situations of GH hypersecretion and deficiency. Acromegaly and GH deficiency thus represent important pathophysiological models for studying, in parallel to physiological approaches, the mechanisms underlying the sodium-retaining action of GH (reviewed in Sections IV and V). The first controlled study in healthy men showed an increase in extracellular volume, as shown by 82Br dilution, after short-term treatment with GH, with no change in plasma volume or blood pressure (98). Another study showed that this effect was stronger in men than in women (99). Recently, similar conclusions were drawn in a large study of 96 recreational athletes, with 10.2 and 7.9% increases in extracellular volume (measured by 82Br dilution) in GH-treated subjects of both genders and in women, respectively (100).

doi: 10.1210/er.2013-1071

edrv.endojournals.org 241

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

Although the antinatriuretic action of the GH/IGF-1 axis in the kidney is well established, the underlying mechanisms were controversial. The first important question was whether the antinatriuretic effect resulted from a direct GH/IGF-1 action on the kidney tubule or from indirect mechanisms involving the renin-angiotensin-aldosterone system (RAAS) or antinatriuretic peptides. Increased aldosterone excretion after GH administration to hypophysectomized rats (101) and to humans (96, 97) indeed suggested RAAS stimulation, but the pituitary-extracted GH used in these studies might have contained contaminating pituitary peptides. Ho and Weissberger (102) reported for the first time that biosynthetic human GH induced a rapid increase in plasma renin activity and aldosterone levels in healthy men.
Treatment with captopril, an angiotensin-convertingenzyme inhibitor, and spironolactone, a mineralocorticoid receptor antagonist, abolished the GH-induced increase in extracellular volume, further suggesting the involvement of the RAAS system (103). However, this concept of an indirect, RAAS-dependent antinatriuretic action of GH was subsequently challenged by several studies. The first conflicting evidence came from the demonstration that sodium retention after GH administration could also occur in the absence of the adrenal glands (104, 105). In the study by Møller et al, plasma angiotensin II and aldosterone concentrations did not increase during GH treatment, but atrial natriuretic peptide (ANP) concentrations fell significantly (98). A study by Christiansen’s group (106) in healthy volunteers and two important studies of GH-deficient subjects (107, 108) (see Section V.C) further demonstrated a RAAS-independent sodiumretaining action of GH. Despite some divergences, most recent data favor a direct stimulatory action of GH on sodium and water reabsorption in the kidney tubule.
Few studies have attempted to distinguish the intrinsic effects of GH and those mediated by IGF-1. Guler et al (61) were the first to investigate the metabolic effects of infused recombinant IGF-1 in two healthy subjects. Because body weight and sodium excretion did not change, the authors inferred that fluid accumulation might be a direct, IGF1-independent effect of GH (61, 84). Another study showed no change in sodium homeostasis in subjects treated with IGF-1 sc (109, 110). Yet the antinatriuretic properties of IGF-1 were clearly documented in children with GH insensitivity due to inactivating GH receptor mutations (111). The respective effects of GH, IGF-1, and their combination on extracellular volume were first compared in obese postmenopausal women. Extracellular volume estimated by 82Br dilution increased similarly in all the treatment groups, pointing to an absence of synergy between GH and IGF-1 (112). The fluid- and sodium-

retaining properties of IGF-1 have also been convincingly documented in healthy men (113).
The precise nephron segment where GH/IGF-1 regulates sodium reabsorption has been a subject of debate. As described in detail in Section II, GH and IGF-1Rs are expressed all along the nephron. Microperfusion of rabbit proximal tubules exposed to GH and IGF-1 (114) and measurements of lithium clearance (an important index of proximal tubular sodium reabsorption) in rats (115) and GH-treated patients (106, 107) ruled out a prominent role of the proximal tubule in GH-induced sodium transport. A recent study showed that acute GH injection in rats resulted in increased phosphorylation of the renal-specific Na-K-2Cl cotransporter (NKCC2) cotransporter in the thick ascending limb of Henle’s loop (115). NKCC2 phosphorylation is associated with short-term actions of vasopressin in this nephron segment (116). Nevertheless, the lack of a concomitant GH-induced change in sodium transport in the microperfused murine thick ascending limb of Henle’s loop (115) challenged the physiological relevance of this observation and suggested that the putative site of GHmediated sodium reabsorption would lie beyond Henle’s loop. Indeed, the above-mentioned human metabolic studies suggested that GH might exert its effects in the distal tubule (106, 107), a nephron segment with a pivotal role in fine-tuning sodium homeostasis, via its regulation by several hormonal systems, including the mineralocorticoid hormone aldosterone (117, 118).
The molecular mechanisms by which GH and IGF-1 regulate sodium transport in the distal nephron have been studied in several cell-based systems, leading to the identification of their main molecular targets. The role of IGF-1 in the regulation of transepithelial sodium transport in the distal nephron has been convincingly documented in amphibian (119, 120) and mammalian (70, 71) cell systems. Such an action could have been anticipated from the well-known effects of insulin on sodium handling in the distal tubule (121, 122). The first study was performed on the toad bladder epithelium, an experimental system that was critical in early demonstrations of the actions of aldosterone on sodium transport (123). IGF-1, like insulin, rapidly stimulated transcellular sodium transport. The effect of both hormones was blocked by amiloride (119), suggesting the involvement of the epithelial sodium channel (ENaC), a key regulator of sodium entry through the apical membrane of polarized epithelial cells (124). These findings were later confirmed in the murine cell lines mCCDcl1 and mpkCCDc14 derived from the principal cells of the collecting duct (70, 71). Rossier’s group (70) compared the effects of insulin and IGF-1 in mCCDcl1 cells and showed that concentrations of insulin 50 times higher than

242 Kamenický et al GH, IGF-1, and Kidney

Endocrine Reviews, April 2014, 35(2):234 –281

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

Figure 2.

␣-subunit of the ENaC. This effect

appeared to be related to JAK2/

STAT5 activation because GH treat-

ment led to phospho-STAT5 binding

to a response element located in the

promoter of the SCNN1A gene (en-

coding the ␣ENaC subunit), highly

conserved among mice, rats, and hu-

mans (35). However, these in vitro

observations of ␣ENaC transcrip-

tional regulation do not directly

demonstrate that GH activates

ENaC- dependent sodium transport

in the distal tubule. Convincing evi-

dence from amiloride-sensitive,

short-circuit current measurements

or patch clamp experiments, espe-

cially on isolated kidney tubules, is

lacking. Because cross talk between

cytokine receptor signaling path-

ways and steroid receptors had been

Figure 2. Proposed model of cooperative GH and IGF-1 action on cortical collecting duct cells. GH documented (131), a functional link

binding to GHR triggers activation of the JAK2/STAT5 and MAPK pathways, leading to transcriptional activation of kidney-specific GH target genes, including ␣ENaC. IGF-1, either synthesized locally or trapped from the circulation, binds to IGF-1R and regulates the apical membrane abundance of ENaC via phosphatidylinositol 3-kinase (PI-3K)-dependent Sgk1 activation, as well as its open probability. The two hormones act synergistically to stimulate

between the mineralocorticoid receptor (the key transcriptional regulator in collecting duct cells) and STAT5 was very likely, given the

transepithelial sodium transport (35). P, phosphorylated; Ub, ubiquitinylated; STAT-RE, STAT response element; MR, mineralocorticoid receptor; MRE, MR response element.

proximity of their respective response elements on the SCNN1A

promoter. Based on these data, we

IGF-1 were needed to obtain the same effect on sodium proposed a model of cooperative GH and IGF-1 action in

transport. Both hormones are thus likely to regulate so- the distal tubule, providing intricate control of transepi-

dium transport through the IGF-1R rather than the insulin thelial sodium reabsorption, as schematized in Figure 2.

receptor.

Another point of convergence of GH and IGF-1 signaling

Aldosterone regulates ENaC-dependent sodium reab- in the control of sodium transport may be ERK1/2, which

sorption in the kidney via various transcriptional and has been shown to stimulate the activity of Na/K-ATPase

post-transcriptional mechanisms, including induction of in some (132) but not all studies (133).

the ␣ENaC gene (125, 126) and of the serum- and gluco-

It is interesting to note that prolactin and GH signaling

corticoid-induced kinase 1 (Sgk1) (127). Sgk1, once phos- plays an important role in the osmoregulation and salinity

phorylated, inhibits the ubiquitin-ligase Nedd4 –2 (128) adaptation of teleost fish (134, 135). Prolactin also in-

and subsequently increases the membrane residency of ac- duces sodium transport via stimulation of ENaC and Na/

tive ENaCs. Insulin (129) and IGF-1 (70) regulate ENaCs K-pump activities in amphibian skin (136). Lactogen/GH-

mainly through Sgk1 activation. Patch-clamp experiments dependent ENaC activation could thus represent an

on renal tubular cells and isolated cortical collecting ducts important phylogenetic axis in the control of sodium and

of rats showed that IGF-1 acutely regulates the open prob- water homeostasis, conserved through hundreds of mil-

ability of individual ENaCs through inositol-triphosphate lions of years of evolution.

production (71). IGF-1, which displays stable plasma con-

centrations throughout the day, may play a physiological 2. Regulation of renal phosphate reabsorption

role in maintaining basal sodium transport in the distal

The kidney ensures body phosphate homeostasis by

tubule, irrespective of aldosterone status (70).

adjusting renal phosphate excretion to dietary intake. In

Using highly differentiated KC3AC1 cortical collecting physiological conditions, 80 –90% of the filtered phos-

duct cells (130), we showed that GH exerts direct phate load is reabsorbed. Renal phosphate reabsorption

stimulatory effects on transcriptional regulation of the occurs almost exclusively in the proximal tubule (19).

doi: 10.1210/er.2013-1071

edrv.endojournals.org 243

Downloaded from https://academic.oup.com/edrv/article/35/2/234/2354705 by guest on 17 June 2022

GH and IGF-1 play important roles in adapting phosphate metabolism to increased requirements during juvenile growth, a period of increased bone formation. The renal phosphate-retaining action of GH was first observed more than 50 years ago in studies of anterior pituitary extracts, showing decreased urinary phosphate excretion and increased plasma phosphate concentrations in men after GH treatment (137). Corvilain and Abramow demonstrated in healthy men (81) and dogs (138) that this antiphosphaturic action of GH was due to an increase in the maximum tubular phosphate reabsorption rate (TmPO4). GH enhanced renal phosphate reabsorption independently of PTH because a similar antiphosphaturic effect was observed in parathyroidectomized dogs (138). Conversely, hypophysectomy and selective inhibition of pulsatile GH release in rats causes a significant decline in TmPO4 and increased urinary phosphorus losses (139, 140). Several clinical trials of rhGH confirmed these observations in subjects without GH deficiency (141–143) and also in GH-deficient patients (144 –146).
The impact of GH and IGF-1 on the phosphate balance also becomes apparent during GH hypersecretion (Section IV). Although chronic GH treatment increases renal tubular transport of phosphate, acute GH administration does not produce similar effects. For instance, tubular phosphate reabsorption remained unchanged 2 hours after GH injection to normal and parathyroidectomized dogs (147), suggesting an indirect, IGF-1-mediated action. Elegant in vitro perfusion studies on rabbit isolated proximal convoluted tubules by Quigley and Baum (114) clearly showed that the phosphate-retaining action of GH is entirely IGF-1 mediated and, as expected, takes place in the proximal tubule. IGF-1 was shown to stimulate phosphate transport via both the basolateral and apical membranes of proximal tubular cells. However, when IGF-1 was added to the apical membrane, the maximal response was greater (up to a 46% increase in phosphate transport), and phosphate transport was stimulated by IGF-1 concentrations 100 times lower (114).
The cellular and molecular mechanisms of phosphate transport in the proximal tubule are comprehensively summarized in Ref. 19. Phosphate is taken up from tubular fluid by apical membrane sodium-phosphate (Na-Pi) cotransporters and leaves cells through basolateral transport pathways. Apical Na-Pi cotransporters are key players in transcellular phosphate flux and are molecular targets for various physiological regulatory mechanisms, including IGF-1 (19). IGF-1 was first shown to stimulate Na-Pi cotransport in brush-border membrane vesicles isolated from rat renal cortex (148). Experiments with OK cells, a model of proximal tubule cells, further showed that IGF-1 directly increased levels of a specific type IIa Na-Pi

cotransporter protein at the plasma membrane (149). The IGF-1 effect on Na-Pi-dependent transport is mediated by IGF-1Rs because it can be blocked by anti-IGF-1R monoclonal antibodies (150). GH and IGF-1 are crucial for early developmental up-regulation of brush-border membrane Na-Pi-II cotransporters in juvenile growing rats, leading to enhanced phosphate reabsorption along the proximal tubule and thereby ensuring the positive phosphate balance critical for increased bone formation during this life period (151).
3. Regulation of renal calcium reabsorption The plasma calcium concentration is kept within nar-
row limits by coordination of intestinal absorption, renal reabsorption, and bone resorption. The kidney tubule has a central role in maintaining calcium homeostasis by adjusting renal calcium losses to highly variable dietary calcium intake. Two major calciotropic hormones tightly control the calcium balance, namely the vitamin D metabolite calcitriol and PTH (152).
GH and IGF-1 play an important role in adapting calcium homeostasis to the increased demands during the period of juvenile growth with accelerated bone formation. Indeed, during childhood and early adolescence, 24hour urinary calcium excretion is low, increasing from 1 mmol/24 h in young children to 2 mmol/24 h just before puberty and reaching adult levels by the end of puberty. This rise in urinary calcium excretion at the end of adolescence reflects decreased skeletal calcium requirements (153). GH and IGF-1 affect calcium homeostasis mainly through their effect on vitamin D metabolism, whereas their relationship with PTH is controversial. Indeed, Spanos et al (154, 155) reported more than three decades ago that GH stimulated calcitriol production in experimental animals (154) and men (155), whereas further investigations in mice and isolated cells showed that this GH action was mediated by IGF-1 stimulation of 1␣-hydroxylase in the proximal tubule (156).
Interestingly, acute administration of recombinant IGF-1 to healthy volunteers had only modest effects on calcium handling, leading to a reduction in calcium excretion only during the first day of treatment in normovolemic subjects (109, 110). In situations of GH hypersecretion and GH deficiency, renal calcium excretion is modified, but the renal action of GH/IGF-1 is difficult to evaluate because of parallel changes in intestinal calcium absorption and, consequently, in filtered calcium loads. Indeed, GH/IGF-1-induced calcitriol production increases intestinal calcium absorption via the epithelial calcium channel TRPV6, resulting in a positive calcium balance (152, 157). However, the calcitriol-mediated calcium-retaining effect of IGF-1 also involves the kidney. We have recently demonstrated in acromegalic patients
IgfKidneyTubuleNephronStudies