Tissue-specific disruption of the growth hormone receptor (GHR) in mice: An update

Abstract

The Growth hormone receptor (GHR) is expressed in many cells/tissues in the body. To investigate the specific metabolic effects of GH action in distinct tissues, several tissue-specific GHR gene disrupted or knockout (KO) mouse lines have been generated. Previously, we have described the effects of GHRKO in several known insulin sensitive tissues, namely liver, muscle and adipose tissue. In this review, we further explore and summarize the main findings of recently published GHRKO results in liver, adipocytes, intestine, bone, brain and heart.

1. Introduction

Growth hormone (GH) is a peptide hormone synthesized by somatotrophic cells of the anterior pituitary and is the principle player in early longitudinal bone and organ growth. Additionally, it also actively participates in maintenance of essential metabolic processes such as insulin regulation and fatty acid metabolism throughout life [1-3]. GH drives linear growth by binding directly to its receptor (R) on bone, cartilage and hepatocytes. The GH-GHR interaction stimulates a signaling cascade that ultimately leads to the synthesis and secretion of insulin-like growth factor I (IGF-I). Thus, the GH/IGF-I axes initiates and mediates an endocrine/autocrine/paracrine effect on growth [4]. Because IGF-I expression is mainly driven by GH action, IGF-I participates in an extensive and complex negative feedback loop that regulates synthesis and secretion of GH. Besides IGF-I, there are other molecules that also regulate GH secretion including leptin, ghrelin, glucocorticoids, corticosteroids, and free fatty acids [2, 4].

In addition to promoting longitudinal growth, other research has pointed out that GH and IGF-I have other critical physiological roles [2, 3, 5]. In fact , many of the metabolic effects of GH and IGF-I are opposite - GH has a diabetogenic effect blocking insulin action in muscle, adipose, and hepatic tissue, increasing blood-glucose, promoting lipolysis and inhibiting lipogenesis; while IGF-I promotes glucose uptake in muscle cells [1]. Both peptides decline with age in human and animal life, and because of their distinct effects on metabolism, profound changes in longevity and ageing would be expected following a variation in these two hormones. In this regard, a decrease in levels of both GH and IGF-I leads to an increase in lifespan in mice [5]. One of the most effective ways of teasing out differential effects from these molecules is by studying their physiological effects in animal models.

Researchers have studied the effects of GH in whole body homeostasis. Due to complexities in GH regulation, deficiencies in GH have been handled in a wide variety of ways: knockout for GH receptor (GHRKO mice; similar to patients with Laron Syndrome), or GH-deficient mice due to mutations in Ghrh (GHRHKO mice) or in Ghrhr expression (lit/lit mice), as well as transgenic expression of a GH antagonist (GHA mice) [6]. GHRKO mice are ~50% the size of littermates, have increased adipose tissue (AT) mass with decreased lean mass, and are resistant to cancer and diet-induced diabetes [6]. Furthermore, due to absence of GH action, serum IGF-I levels are significantly lower, thus, impairing the IGF-I negative feedback on GH release. As a consequence, the GHRKO mice not only have low IGF-I levels, but also present GH serum levels higher than WT mice.[7]. Importantly, GHRKO mice live significantly longer than their littermate controls. In fact, one of the GHRKO mice lived ~5 years in a laboratory setting, creating a record for longevity [6]. Similar to the GHRKO mice the lit/lit mice have also smaller size, increased adiposity and increased longevity [6]. Mice with GHA genotype have also been found to be smaller than littermates and also show obesity, with mild improvements in glucose handling and longevity, which could be due to higher levels of circulating IGF-I [6]. These animal models constitute only a fraction of the huge body of work on GH and IGF-I research.

Due to complex molecular interactions, studying individual hormonal effects can be extremely challenging when looking at whole-body physiology. Therefore, creating specific animal models that can survive and thrive in laboratory conditions, accounts for a necessary field of research which allows for teasing out individual physiological effects as well as delivering insight on molecular interactions. Because of the particular tissue-specific effects exerted by GH and due to the wide range of cells where GHR is expressed [8], the extent at which each cell/tissue contributes to the healthy phenotypic characteristics seen in the GHRKO mice is unknown. Furthermore, new technologies have allowed the generation and study of conditional or tissue-specific animal models with GHR ablation, thus allowing researchers to further investigate the effects of an absence of GH action in particular tissues with respect to the whole body homeostasis. Previously, we published a review covering the tissue specific mouse models that were generated to investigate the effects of lack of GH action on conventionally known insulin sensitive tissues, namely, liver, skeletal muscle and AT. In this review, we give an update and further explore the findings of new GHR tissue-specific mouse lines that were recently developed in our and other laboratories [9]. These new tissue-specific mice have the GHR ablated specifically in either of both the conventional-insulin sensitive tissues, such as liver and AT; and non-conventional insulin sensitive tissues, like macrophages, bone, heart, intestine and brain. To note, even though skeletal muscle is one of the most insulin-sensitive targets of GH action, no new GHRKO mouse lines have been generated for this tissue since our last review [9]. Therefore, the summary on muscle-specific GHR disrupted animals will not be covered in this review.

2. Update on the tissue-specific GHR ablation of the ‘insulin sensitive tissues’

2. 1. Liver GHR KO mice.

After the generation of the GHRKO mice, many laboratories have tried to identify in which manner and to what extent each of the tissues contributes to the health and extended longevity of GHRKO mice [9]. Thus, our laboratory and other laboratories have conditionally and specifically knocked down the GHR in different tissues including; liver, AT, muscle, bone, heart and intestine, between others. In terms of liver, five liver specific mouse lines have been described, four of them have a germline-GHR gene disruption and one of them has a post-natal ablation of the GHR [10-14]. Although these mouse lines have several phenotypic differences, all the germline liver-specific GHRKO mice show high GH and low IGF-I levels. This is due to the fact that the liver produces 75- 90% of circulating IGF-I [11, 13, 15]. Therefore, liver-specific GHRKO mice allow us to explore the direct effects of GH in comparison to the indirect effects produced by endocrine IGF-I. Three of the germline liver-specific GHR ablated mouse lines were generated using an exon-4 GHR floxed mouse and the albumin transcriptional regulatory sequence. These three mouse lines are the GHR liver deficient (GHRLD) mouse, described in 2009 by Fan et al. [10]; the liver-specific GHRKO (LiGHRKO) mouse, generated by List et al. in 2013 [11]; and the Li-GHRKO mouse created by LeRoith in 1999 [13], and further characterized by Yakar in 2016 [12]. The other germline liver-specific GHRKO mouse line is named L-Ghr−/− mouse and was produced by Fang et al. in 2019 [14]. The L-Ghr−/− mice also used the albumin regulatory sequence, but unlike the other three germline liver-specific GHRKO mice, the L-Ghr−/− mice were produced by using CRSPR/Cas9 and Cre recombinase to ablate exon-4 to exon-4b of the GHR gene [14].

In addition to the decreased circulating IGF-I and increased GH levels, the germline liver specific GHR ablated mouse lines, have other common characteristics. For instance, other surrogate circulatory markers of GH action in the liver were also changed; IGF binding protein 3 (IGFBP3), and ALS were decreased in the LiGHRKO, Li-GHRKO and GHRLD mouse lines; and increased circulatory IGFBP2 was found in LiGHRKO and GHRLD mice when compared to controls [10-12]. In terms of glucose homeostasis, the LiGHRKO, Li-GHRKO and the GHRLD revealed impaired glucose metabolism, with LiGHRKO and Li-GHRKO showing increased blood glucose and elevated insulin [10-12]. Also, the Li-GHRKO and the GHRLD mice were glucose intolerant and insulin resistant compared to WT mice [10-12]. It is important to note that liver-specific GHRKO mice still have GH action in the entire body except for the liver. Therefore, unlike the global GHRKO mice which lack GH action in all tissues, the diabetogenic effect of GH is still viable in the liver-specific GHRKO mouse lines and is reflected in their decreased insulin sensitivity and impaired glucose homeostasis.

Liver steatosis was also a consistent observation in the LiGHRKO, Li-GHRKO and the GHRLD mice when compared to control mice, with the Li-GHRKO and GHRLD mouse lines showing increased circulating FFA and triglycerides (TG) [10-12]. Increased adiposity was a feature showed only by the Li-GHRKO mice but increased leptin levels was shown in the LiGHRKO and Li-GHRKO compared to WT mice [11, 12]. Moreover, GHRLD and Li-GHRKO mice exhibited increased expression of inflammatory markers both in the circulation and in the liver [10, 12]. Increased fibrosis and oxidative stress in the liver when compared to controls were seen in the GHRLD and the Li-GHRKO mice [10, 12]. Contrary to these results, the L-GHR−/− displayed no change in blood glucose levels, fat mass, or hepatic TG [14]. These opposing results could be a reflection of the different mechanism used to generate the mouse lines or the age at which the measurements were taken, that is, while most of the analysis made in the LiGHRKO, Li-GHRKO and GHRLD mouse lines were performed after 12 weeks of age, L-GHR−/− mice were evaluated at 6-8 weeks of age [10-12, 14].

Finally, the fifth liver specific GHRKO mouse line was not generated from germline but a postnatal GHR disruption; these mice are called adult-onset, hepatocyte specific, GH receptor (GHR) knockdown (aLivGHRkd) mice[16]. These animals were produced by injecting 10-12 week old mice with an adeno-associated virus (AAV) bearing a liver specific thyroxine-binding globulin (TBG) promoter, which drives a Cre recombinase transgene that ultimately allows the recombination and ablation of the exon 4 of the GHR, specifically in hepatocytes. Surprisingly, these mice showed decreased circulating IGF-I, but no change in GH serum levels when compared to controls [16]. Contrary to the germline liver-specific GHR ablated mice, the aLivGHRkd had no change in adiposity, fasting insulin levels, glucose tolerance or insulin sensitivity with respect to WT mice [10-12, 16]. But, similar to the germline liver-specific GHRKO mice, the aLivGHRkd exhibited increased hepatic glycogen and TG content in males compared to control males [10-12, 16]. Therefore, liver steatosis is a feature that is shared by all the liver-specific GHRKO mouse lines. Aging studies performed in the LiGHRKO mice showed that longevity is not prolonged when GHR is ablated only in the liver [11]. Thus, decreasing endocrine IGF-I effect is not sufficient to extend lifespan.

2.2. Adipose Tissue GHRKO mice

AT is an endocrine organ that is also greatly impacted by GH action. Increased adiposity or obesity is usually associated with diabetes, cardiovascular diseases, cancer and premature death. It is counterintuitive that the GHRKO mice are obese yet healthy with an extended lifespan. Surprisingly, even though AT mass is increased throughout the body of the GHRKO mice, different fat pads are impacted differently by GH action [17]. AT has then been postulated as one of the tissues that may help to explain the healthy aging seen in the GHRKO mice. Because of this, three mouse lines with disrupted GHR specifically in the AT have been created. Two of these mouse lines have been generated in our laboratory, one in 2013 and one in 2019 by List et al. [18, 19]., the third mouse line was also created in 2019 by Fang et al. at the University of Texas [19]. The AT specific GHRKO (FaGHRKO) mouse line, created in 2013 used the Cre-Lox system driven by the Fabp4 promoter/enhancer [19]. Importantly, it has been recently shown that the Fabp4 promoter/enhancer used to produce these mice can ‘leak’ and be expressed in non- adipose tissue cells. Therefore, another mouse line named the adipocyte-specific GHR null (AdGHRKO) mice was also generated at Ohio University [18]. To create these mice an adipocyte-specific promoter, the adiponectin promoter, was used to target the exon 4 of the GHR [18]. One of the main effects of GH action in AT is the enhancement of lipolysis. Because of the decreased GH action in the AT, both mouse lines, the FaGHRKO and the AdGHRKO mice had increased fat mass compared to control mice [18, 19]. To note, the GHRKO mice have increased adiposity, especially in the subcutaneous AT depot, but FaGHRKO and AdGHRKO mice showed increased adiposity in all fat pads in females (subcutaneous, perigonadal, retroperitoneal, and mesenteric); males also showed enlargement of all the fat pads except for the perigonadal depot [17-19]. Adipocyte size and collagen deposition are also increased in both of these mouse lines when compared to controls [18, 19]. Besides the similarities found in these two mouse lines, many differences were also found. For instance, while FaGHRKO have increased body weight and length, the AdGHRKO exhibited no differences in body weight or length compared to control mice [18, 19]. One explanation for this difference is that FaGHRKO have augmented IGF-I in males and a slightly increase in GH levels in both males and females while the AdGHRKO have no change in circulating IGF-I or GH levels compared to WT counterparts [18, 19]. There are also changes in other circulatory markers; while FaGHRKO mice showed no alterations in resistin and decreased adiponectin levels in males, the AdGHRKO mice had reduced levels of both, resistin and adiponectin [18, 19]. Liver TG results were also distinct in these two mouse lines with no difference and decreased liver TG in the FaGHRKO and the AdGHRKO mice, respectively [18, 19].

Glucose metabolism was also differentially changed in these two mouse lines, whereas glucose homeostasis was not changed in the FaGHRKO mice, with no significant differences in fasting glucose and insulin levels, glucose tolerance and insulin sensitivity[19]. The AdGHRKO mice showed a similar trend to the GHRKO mice, with decreased insulin levels and enhanced insulin sensitivity. The last adipose tissue-specific GHR ablated mouse line was created by Fang et al. and reported this year in PNAS [14]. The Fat-GHR−/− mice were created using the same enhancer/promoter than the AdGHRKO mice, that is, the adiponectin enhancer/promoter, but in this case they ablated exon 4-4b of the GHR [14]. The Fat-GHR−/− mouse, unlike the AdGHRKO mouse line, showed no change in fat mass, and glucose and GH levels compared to WT controls [14]. Overall the results found in these three mouse lines suggest that the methodology used to generate these mouse lines result in several phenotypic differences. For example, it appears that a more specific disruption of the GHR in the adipocytes as seen in the AdGHRKO have a healthier phenotype than the FaGHRKO profile. In support of this, aging studies performed in FaGHRKO showed decreased longevity in these mice compared to controls. Thus, having improved glucose homeostasis may play a significant role in extending lifespan. Aging studies on the AdGHRKO mice are currently being performed. Since this mouse line showed improved glucose homeostasis, the upcoming longevity results may help to clarify the relationship between improved glucose homeostasis and lifespan.

3. GHR ablation in ‘non-conventional-insulin sensitive tissues’

GHR is expressed in many tissue in the body [8]. To explore the effects that lack of GH action has in different tissues and their possible involvement in GH-related comorbidities, GHR has been specifically ablated in many other tissues and cell types, including macrophages, bone, heart, intestines and brain. In this section, we will aim to describe the main findings of these tissue specific mouse models. It is important to note that most of these mice were created by ablating the exon 4 of the GHR, unlike the Li-GHR−/− and the Fat-GHR−/− mice described previously that were generated by disrupting the exon4-4b of the GHR [14].

The study of macrophages and other immune cells with respect to GH-related comorbidities such as obesity, diabetes and cancer has become of much interest to many researchers, due to the macrophage’s ability to infiltrate and subsequently modulate the metabolism of distinctive organs. The macrophage-specific GHR ablated mice or the GHRMacD mouse, was reported in 2010 by Menon’s laboratory at the University of Michigan, and was created using the lysozyme 2 (Lyz2) locus to drive the Cre recombinase to ablate the GHR [20]. Even though these mice are called macrophage-specific, the Lyzs locus is known to be expressed not only in mature macrophages, but also in monocytes and granulocytes [21]. Therefore, outcomes obtained with this mouse line are the result of ablated GHR in these three immune cell populations. Results from the GHRMacD mice showed no alterations in terms of body weight or glucose metabolism compared to controls [22]. However, when subject to high fat diet, the GHRMacD mice present impaired glucose metabolism and increased fat and adipocyte size in the epididymal AT depot [22]. Also, the epididymal AT depot showed an increased stromal vascular fraction, with augmented activated M1 macrophages, inflammatory markers and crown-like structures; all these indicators of an inflammatory status [22]. Furthermore, when AT derived macrophages isolated from GHRMacD mice are co-cultured with 3T3-L1 preadipocytes, a decreased in preadipocyte differentiation and adipogenesis is found [20]. Together, these results suggest that lack of GH action in macrophages present in the AT can contribute to a deterioration in glucose homeostasis and increased AT inflammation at least in high fat diet. Also, in vitro studies suggest that GH action in macrophages is needed for proper preadipocyte differentiation.

The GH/IGF-I axes are well known for its effects in bone development and maintenance, but the specific contributions of GH vs IGF-I in bone structure and biology are still debatable. Therefore, in 2016 Yakar’s laboratory at New York University generated a bone-specific GHR disrupted mouse called the dentin matrix protein (DMP)-1 GHRKO (DMP-GHRKO) mouse [23]. These mice were generated using the DMP-1 promoter/enhancer to drive the Cre recombinase to ablate the GHR specifically in the bone. DMP-GHRKO did not have any significant changes in IGF-I levels or body weight, composition and growth [23]. As expected, they did possess differences in terms of bone acquisition. MicroCT analysis performed in trabecular and cortical bone showed that although cortical and trabecular osteocyte gross morphology was not changed, slender bones and reduced cortical and trabecular areas was seen in DMP-GHRKO compared to WT mice. Also, increased number of osteoclasts, as well as decreased bone formation, mineral deposition rate and number of osteoblasts were found in DMP-GHRKO mice compared to controls [23]. Furthermore, these mice had decreased serum parathyroid hormone and activity [23]. Therefore, GH action (and presumably parathyroid hormone) in bone is needed for proper bone acquisition and maintenance.

Excess or diminished GH action can affect the heart [24]. In humans, patients with acromegaly have GH hypersecretion and present with cardiac abnormalities including cardiomyopathy, hypertension and vascular dysfunction; In fact, untreated patients with acromegaly can die from cardiac failure. On the other side of the spectrum, GH deficient (GHD) and Laron Syndrome (LS) patients, which present lower and no GH action, respectively, have decreased cardiac size and function [24]. To elucidate the specific effects of GH in the heart, Jara et al. in 2016 created a cardiac-specific GHR disrupted (iC-GHRKO) mouse, using the myosin heavy chain (Myh6) promoter/ enhancer [25]. Surprisingly, even though no significant differences in heart size were apparent, echocardiography or systolic blood pressure were seen in iC-GHRKO compared to WT mice. Also, body composition and glucose homeostasis were changed in an age-specific manner. That is, decreased fat mass and increased lean mass was observed in iC-GHRKO compared to controls between 4.5 and 8.5 months of age [25]. In terms of glucose homeostasis, while insulin sensitivity was enhanced at 6 months of age, decreased glucose tolerance, insulin sensitivity and circulatory IGF-I levels were found in iC-GHRKO compared to WT mice at 1 year of age [25]. Therefore, results suggest that disrupted GH action in the heart have repercussions in whole body composition and glucose homeostasis at an adult age, but not in specific heart performance suggesting that GH, or lack thereof, can induce ‘cardiokines’ that effect whole body metabolism.

Intestine is a key organ for the regulation of food absorption and, therefore, for nutrient and glucose metabolism [26]. GH action in the intestine is not well characterized, yet recombinant GH is prescribed to treat several short bowel syndrome and inflammatory bowel disease due to its reported ability to improve the intestinal barrier [26, 27]. On the other hand, excess of GH has also been shown to increase intestinal dysfunction [28]. Therefore, to clarify the effects of GH in the intestines in 2019, Young et al. described the generation and characterization of an intestinal epithelial cell-specific GHRKO (IntGHRKO) mouse line, produced using the villin promoter enhancer-driven Cre recombinase [29]. Decreased intestinal length and fat absorption was shown in IntGHRKO male mice, when compared to WT mice [29]. Interestingly, as stated above, although recombinant GH has been reported to improve intestinal barrier in an intestinal-disease-state, it seems that germline ablation of GH action in the intestinal endothelial cells may improve intestinal barrier function. That is , compared to controls, male IntGHRKO mice showed small improvement in intestinal barrier function since expression of the gut barrier gene, occludin, was increased [29]. Similarly, IntGHRKO females showed a decreased fecal albumin compared to controls, indicating an improved barrier function given that blood was not crossing the intestinal barrier [29]. Also, In terms of whole body homeostasis, although no changes in body composition or glucose levels were found, glucose intolerance and insulin resistance was found in IntGHRKO female mice compared to WT mice [29]. Overall these results suggest that GH action has a minor impact on the intestines. The effect of ablated GH action on other cells of the intestine, different from endothelial cells, is yet to be explored. Furthermore, it has been shown that the villin promoter/enhancer used for the generation of these mice is also active in the proximal convoluted tubular cells of the kidney also which also express villin [29].

Studies show that GHR is active in the central nervous system, including hypothalamic areas [30]. Furthermore, regulation of growth, metabolism and energy balance is tightly regulated by the brain [30]. To investigate the effects of decreased GH action in the brain and in energy balance four mouse lines that disrupted the GHR in distinct cells in the brain were generated. In 2017, Cady et al. generated mice with an ablated GHR in the leptin receptor expressing neurons, these mice called the Lepr^EYFPΔGHR, were created by using the the Lepr promoter/enhancer [30, 31]. The other three mouse lines were reported this year by Furigo et al. In this paper GH-responsive neurons that are also involved in energy metabolism were used as targets to specifically ablate the GHR [30]. The two cell-specific mouse lines generated carry the name on the protein that are expressed by this neurons, the third mouse line was generated to ablate the GHR in the whole brain. The three mouse lines created were the agouti-related protein (AgRP) GHR KO, the leptin receptor (LepR) GHR KO, and the Brain GHR KO mice; the AgRP-Ires, LepR-Ires and the Nestin promoter/ enhancer were used to express CRE in these mouse lines [30]. Results showed that in terms of body composition and measurements, there no change in the body weight and length of Lepr^EYFPΔGHR and AgRP GHR KO compared to controls [30, 31], but during caloric restriction a higher rate of weight loss is seen in AgRP GHR KO, LepR GHR KO and Brain GHR KO mice with respect to WT mice. As for glucose homeostasis, Lepr^EYFPΔGHR present no change in insulin sensitivity and decreased glucose intolerance with respect to controls [30, 31]. On the other hand, no changes in, food intake, energy expenditure, respiratory quotient, ambulatory activity, adiposity, lean body mass, glucose tolerance and insulin sensitivity were seen in AgRP GHR KO mice compared to WT mice [30]. Also, caloric restricted (CR) AgRP GHR KO mice showed impaired responses that are expected under CR, such as reduced serum leptin levels and increased serum GH concentration [30]. Furthermore, inhibition of thermogenesis is one of the main responses of CR, with WT mice showing reduced levels of uncoupling protein-1 (UCP-1) mRNA [30]. In AgRP GHR KO this reduction was decreased when compared to control mice [30]. In terms of the other two mouse lines, both LepR GHR KO and brain GHR KO mice showed increased body weight and length [30]. However, LepR GHR KO mice showed reduced adiposity, glucose and serum leptin levels compared to control animals [30]. Also, similar to AgRP GHR KO mice, both LepR GHR KO and brain GHR KO mice showed inefficiency to save energy during CR, leading to an increased weight loss when compared to control mice [30]. These studies suggest that disruption of the GHR in distinctive neurons lead to different metabolic effects under normal vs. CR diet. That is, while under normal diet conditions, decreased GH action in AgRP neurons do not result in alterations in energy homeostasis, ablation of the GHR in LepR neurons induces changes in the whole body glucose metabolism. Furthermore, it appears that both, the AgRP and the LepR neurons need GH action for normal metabolic responses under CR.

4. Summary

GHR disruption in mice as seen in the GHRKO mouse line, leads to healthy and long-lived dwarf mice. To understand the contribution of individual cells and tissues to the healthy phenotype of GHRKO mice, many tissues-specific mouse lines have been created. In this paper, we reviewed some of the new tissue-specific GHR KO mouse lines and their main characteristics (Table 1). Even though none of these tissue-specific GHR KO mice had the exact phenotype as the global GHRKO mice, some characteristics of the GHR KO tissue-specific mouse lines are similar to the global GHRKO mice. More importantly, it is interesting how GHR deletion in specific cells and tissues such as liver, AT, macrophages, heart, intestine and some neurons can have an impact in the whole body glucose homeostasis. In general, it seems that while KO of the GHR in liver, macrophages and intestines have a negative effect in the glucose homeostasis, specific ablation of the GHR in adipocytes, heart and certain neurons (LepR GHR KO mice) have a positive effect in glucose metabolism measured by insulin sensitivity, and insulin and/or glucose levels. Other parameters such as oxidative stress and immune response markers seem to also be associated with the lack of GHR expression in specific tissues. Therefore, it appears that even though a specific tissue is not responsible for the favorable effects seen in the GHRKO mice, disruption of the GHR in particular tissues is not always beneficial and can have either positive or negative effects on health and longevity.

Table 1.

Phenotypic characteristics of tissue specific mice with disrupted GH action. Symbols indicate increases (↑), decreases (↓), no change (↔), sex dependent changes (*), and diet dependent variations (**) relative to controls. Insulin growth factor −1 (IGF-1), Insulin tolerance test (ITT), glucose tolerance test (GTT), triglycerides (TG).

Mouse line	Tissue disrupted	Promoter	Serum IGF-1	Serum insulin	Serum glucose	ITT	GTT	Liver TG	Serum adiponectin	Fat mass	Life- span
LI-GHRKO	Liver	Albumin	↓↓	↑	↑	↓	↓	↑		↑
LIGHRKO	Liver	Albumin	↓↓	*↑	*↑			*↑	↑	↓	↔
GHRLD	Liver	Albumin	↓	↑	↔	↓	↓	↑		↔
L-GHR−/−	Liver	Albumin			↔			↔		↔
aLivGHRkd	Liver	Thyroxine-binding globulin	↓	↔	↔	↔	↔	↑		↔
AdGHRKO	Adipose tissue	Adipocyte specific	↔	↓	↔	↑	↔	↓	↓	↑
FaGHRKO	Adipose tissue	Fabp4	*↔↑	↔	↔	↔	↔	↔	*↔↓	↑	↓
Fat-GHR−/−	Adipose tissue	Adipocyte specific			↔					↔
GHRMacD	Macrophages monocytes granulocytes	lysozyme 2				**↔↓	**↔↓			↑
DMP-GHRKO	Bone	Dentin matrix protein	↔							↔
iC-GHRKO	Heart	Myosin heavy chain	↓	↔	↔	↓	↓			↓
IntGHRKO	Intestine	Villin			↔	↓	↓			↔
LeprEYFPΔGHR	Lepr neurons	leptin receptor		↔	↔	↔	↓
AgRP GHR KO	AgRP neurons	AgRP Ires,		**↔↓		↔	↔			**↔↓
LepR GHR KO	Lepr neurons	LepR-Ires		**↔↓						↓
Brain GHR KO	Brain	Nestin		**↔↓