SOCS2 regulation of growth hormone signaling requires a canonical interaction with phosphotyrosine

Growth hormone (GH) is secreted by the anterior pituitary gland into the circulation, binding to a homodimeric GH receptor (GHR) on the surface of cells throughout the body to regulate growth, metabolism, and the immune response. In 1987, the GHR became the first cytokine/hematopoietin receptor superfamily member to be cloned, initiating a molecular investigation into GH signaling [1]. Despite this, it took many years to fully delineate mechanisms of ligand recruitment and signal propagation [2,3]. GHR engagement promotes a conformational change in the homodimeric receptor complex, activating the receptor-associated Janus kinase 2 (JAK2) tyrosine kinases [2,3]. Activated JAK2 in turn phosphorylates tyrosine residues within the GHR cytoplasmic domain, recruiting Src homology 2 (SH2)-containing proteins such as the signal transducers and activators of transcription (STAT) proteins to the receptor complex [4]. Once STAT5b homodimers are recruited to the receptor complex, they are phosphorylated by JAK2, triggering a conformational dimer change, translocation to the nucleus, and the transcription of genes involved in growth and development, such as insulin-like growth factor 1, Igf1 [5–7]. Excessive GH production results in acromegaly in humans and is associated with severe pathological consequences [5–7], underscoring the need for cellular mechanisms that limit the GH response.

The suppressor of cytokine signaling (SOCS) protein family, consisting of CIS and SOCS1-7, functions to limit cytokine signaling, often in a negative feedback loop [8]. The important role of SOCS2 as a negative regulator of GH signaling was revealed by SOCS2-deficient mice, which displayed a 30–40% increase in growth compared with wild-type (WT) mice [9]. The gigantism was rescued by crossing SOCS2-deficient mice with Ghrhr^lit/lit (little) mice that lacked pituitary-derived GH secretion, with both Ghrhr^lit/lit and Ghrhr^lit/litSocs2^−/− mice exhibiting comparable growth retardation [10]. Similarly, SOCS2-null mice on a STAT5b-deficient background displayed a similar growth rate to WT mice, confirming SOCS2-deficient gigantism results from aberrant activation of GH signaling [11]. Notably, the increased weight of Socs2^−/− mice was not due to excess fat gain, but resulted from a proportional increase in organ size, including muscle and bone length. Consistent with this, SOCS2-deficient mice have longer femur, tibia, radius, and humerus bones [9], associated with increased bone mass, but no change in bone mineral density [12]. However, others have reported reduced bone mineral density [13], with Dobie et al., reporting a decrease in cortical bone mineral density [14].

SOCS2 has a short 32-residue N-terminal region and two major functional domains: a central SH2 domain and a SOCS box. The SOCS2-SH2 domain is a substrate recognition module that binds to phosphorylated tyrosine residues 487 and 595 within the GHR cytoplastic domain, with SOCS2-binding proposed to inhibit cascade propagation via blocking access to other signaling intermediates [10]. The SOCS2-SH2 domain has also been reported to directly interact with JAK2 [15]. The SOCS box provides binding sites for the adaptors Elongins B and C, and the Cullin 5 scaffold, which in turn, recruit RING-box protein (Rbx)2 to form an E3 ubiquitin ligase complex that ubiquitinates substrates bound to the SH2 domain [16]. There is an underlying assumption that SOCS2 ubiquitinates the GHR complex directing it to proteasomal degradation.

The SOCS2-SH2 domain structure consists of three β-strands flanked by two α-helices, and an additional ‘SOCS-specific’ α-helix termed the extended SH2 subdomain (ESS) [17]. Interaction with phosphorylated targets occurs via a positively charged pocket (P0) that binds phosphorylated tyrosine (pTyr) and a hydrophobic patch that accommodates the third residue (+3) distal from the pTyr. Recently, we identified a noncanonical exosite on the SOCS2-SH2 domain that when occupied, enhances SH2 affinity for tyrosine phosphorylated targets [18]. Multiple SOCS2 residues, including Arg96, co-ordinate binding to the pTyr residue [19]. A naturally occurring mutation of Arg96 to Cys in ovine SOCS2 is associated with inflammatory mastitis, and increased body size and milk production [20], with mutation of Arg96 disrupting SH2 binding to phosphopeptides [18,20]. The characterization of the R96C mutation in sheep was the first study to investigate the contribution of SH2:pTyr binding to SOCS2 function in vivo [20]. However, there are no comparative studies of SOCS2-deficient sheep to address whether SOCS2 function is completely reliant on its SH2 interaction with phosphotyrosine.

Here, we generated and characterized mice bearing the SOCS2-R96C mutation (Socs2^R96C). Homozygous Socs2^R96C/R96C mice displayed a 25–35% increase in weight compared with WT mice during puberty, and a similar increase in body weight to SOCS2-deficient mice. The gigantism resulted from a collective increase in weight of most visceral organs, associated with increased body and bone length. GH signaling was enhanced in Socs2^R96C/R96C fibroblasts, indicating that the augmented growth during development was a consequence of a dysregulated GH response. The Socs2^R96C/R96C mice displayed a similar phenotype to Socs2^−/− mice, highlighting a critical role for SOCS2-SH2 recognition of phosphotyrosine in canonical SOCS2 function.

SOCS2-SH2 binding to pTyr occurs through a cluster of residues consisting of Arg73, Ser75, Ser76, Thr83, and Arg96 in the SOCS2-SH2 domain [19] (Figure 1A), with all five residues conserved across different species (Figure 1B). Although Arg73 corresponds to the invariant arginine residue found in all SH2 domains [21–23], mutation of Arg73 to Lys in SOCS2 resulted in reduced affinity for a phosphopeptide derived from GHR Y595 (pY595), but not loss of binding (Supplementary Figure 1).

We had previously shown that mutation of Arg96 to Cys in the SOCS2-SH2 domain abrogated binding to a phosphopeptide derived from GHR Tyr595, without impacting on peptide binding to the SH2 exosite, confirming domain integrity [18]. To verify that the R96C mutation did not impact SOCS box function, GST-SOCS2 and GST-SOCS2-R96C, were purified in a trimeric complex with Elongins B and C, and used to affinity precipitate Cullin 5 from cell lysates. Elongins B and C were present at comparable levels in both GST-SOCS2 and GST-SOCS2-R96C complexes. GST-SOCS2-R96C efficiently enriched Cullin 5 to the same extent as GST-SOCS2, evidence that mutation of R96C within the SOCS2-SH2 domain did not disrupt the SOCS box E3 complex (Figure 1C). GST-SOCS2-R96C failed to enrich tyrosine-phosphorylated proteins, confirming that pTyr binding was disrupted by the R96C mutation (Figure 1C).

To further investigate the impact of mutating Arg96 to Cys in vivo, we used CRISPR/Cas9 gene editing to generate a C57BL/6 mouse strain bearing the R96C mutation. Correct gene targeting was validated by next-generation sequencing (NGS), with routine genotyping performed by PCR (Supplementary Figure 2). Homozygous Socs2^R96C/R96C mice were viable and born at Mendelian frequencies. Socs2^R96C/R96C mice were indistinguishable to Socs2^R96C/+ and WT mice before weaning at 3 weeks of age, but subsequently developed at a faster rate (Figure 2). Analysis of immune cell populations from Socs2^−/− and Socs2^R96C/R96C mice showed increased numbers of splenic natural killer (NK) cells (Supplementary Figure 3), consistent with previous observations [15]. By 6 weeks of age, both male and female Socs2^R96C/R96C mice weighed significantly more than Socs2^R96C/+ or WT mice. The enhanced growth of Socs2^R96C/R96C mice was evident throughout puberty, with growth plateauing at week 8, and maintained in adult mice (Figure 2C–F). At 8 weeks of age, body weights of Socs2^R96C/R96C mice were comparable to Socs2^−/− mice, with both significantly heavier than WT mice (30.7% and 29.1% in male, 25.4% and 27.5% in female, respectively), suggesting that the weight difference in Socs2^R96C/R96C mice phenocopied Socs2^−/− mice (Figure 2B). At 12 weeks of age, Socs2^R96C/R96C male and female mice were visually different from WT mice (Figure 2A and not shown), with Socs2^R96C/R96C male mice attaining a 31.5% weight increase over WT males (Socs2^R96C/R96C: 37.1 ± 2.42 g, WT: 28.2 ± 1.39 g) (Figure 2C,D). Socs2^R96C/R96C females typically attained the weight of WT male mice and displayed a 26.5% weight increase over WT females at 12 weeks of age (Socs2^R96C/R96C: 27.2 ± 1.53 g, WT: 21.5 ± 0.98 g) (Figure 2E,F). Male and female Socs2^R96C/+ mice exhibited similar development and adult body weights to WT mice (Figure 2C–F). The increased size of Socs2^R96C/R96C mice was reflected in a proportional increase in the weight of individual organs, with the majority of organs (excluding brain) showing increased weight compared with organs from WT mice. Pancreas and testes in males did not show a statistical difference. Carcass weight and body length were also increased in Socs2^R96C/R96C mice. Socs2^R96C/R96C mice displayed a similar increase to Socs2^−/− mice, in organ weight, carcass weight, and body length (Figure 2G).

Micro-computed tomography (CT) X-ray imaging of skeletons from 13-week-old male WT, Socs2^−/− and Socs2^R96C/R96C mice revealed an enlarged skeleton and increased bone length of skull, hip, sternum, humerus, femur, fibula, tibia, tarsus, and metacarpus (hindfeet) in Socs2^−/− and Socs2^R96C/R96C mice. Socs2^R96C/R96C mice showed comparable bone lengths to Socs2^−/− mice (Figure 3). Histology of skin from 10-week-old Socs2^R96/R96C male mice revealed dermal thickening and increased collagen production, comparable to that observed in Socs2^−/− mice (Supplementary Figure 4), and consistent with previously reported skin pathology in Socs2^−/− mice [9].

To further investigate the impact of the SOCS2-R96C mutation on GH signaling, mouse embryonic fibroblasts (MEFs) were generated from Socs2^+/+, Socs2^R96C/+, Socs2^R96C/R96C, and Socs2^−/− day-13 embryos. GH stimulation resulted in robust tyrosine phosphorylation of STAT5 (pSTAT5; Figure 4 and Supplementary Figure 5), which peaked at 0.5 h GH treatment in MEFs of all genotypes. The intensity of the pSTAT5 response was comparable in Socs2^R96C/R96C and Socs2^−/− MEFs, and consistently increased at 1–8 h GH treatment, relative to Socs2^R96C/+ and Socs2^+/+ MEFs. Socs2^R96C/+ MEFs displayed an intermediate level of pSTAT5 compared with WT and Socs2^R96C/R96C MEFs. Total STAT5 levels were comparable in all genotypes. SOCS2 was present at baseline, reduced in Socs2^R96C/+ MEFs and expressed at comparable levels in WT and Socs2^R96C/R96C MEFs. There was a modest up-regulation of SOCS2 at 8 h. These data indicate that loss of phosphotyrosine binding is sufficient to abrogate SOCS2 regulation of GH signaling.

Five residues (Arg73, Ser75, Ser76, Thr83, and Arg96) within the SOCS2-SH2 domain form hydrogen bonds that lock the phosphorylated tyrosine residue in place [19]. As shown here, and in other studies [18,20], mutation of Arg96 to Cys results in complete loss of SOCS2-SH2 binding to phosphotyrosine and a corresponding loss of SOCS2 function. Mutation of Arg96 to Leu retained some weak binding, while R96Q completely abrogated binding to phosphotyrosine [19]. In cells, mutation of Arg73 to Lys has only a modest impact on SOCS2 regulation of GH signaling, unless combined with D74E and S75C.

We had previously shown that the SOCS2-R96C mutation did not impact on F3 peptide binding to the SOCS2 exosite [18]. In the present study, we confirmed the R96C mutation did not alter the ability of SOCS2 to recruit the SOCS box-associated-E3 ubiquitin ligase complex. Taken together, this confirmed the R96C mutation did not compromise domain integrity, enabling us to interrogate the role of SOCS2-SH2:pTyr binding in mice. Socs2^R96C/R96C mice displayed increased growth during puberty, attaining a ∼25% and ∼30% increase in weight (females and males) compared with heterozygous littermates. The enhancement in body size and weight was comparable to SOCS2 null mice. In addition, primary embryonic fibroblasts derived from Socs2^R96C/R96C mice displayed prolonged GH-induced signaling compared with WT cells, consistent with loss of a negative regulator (SOCS2), and further evidence that loss of phosphotyrosine binding was sufficient to fully disrupt SOCS2 function.

The structure of SOCS2 in complex with the GHR pY595 peptide indicated that SOCS2 has the capacity to not only engage phosphopeptide in the conventional manner but can also interact with a second GHR pY595 peptide orientated in an antiparallel direction. This model provides a potential mechanism for SOCS2 interaction with both subunits in the dimerized GHR [19]. In addition to pTyr binding, the SOCS2-SH2 domain can also bind to a nonphosphorylated peptide via the exosite, providing an alternative site for SH2 interaction with protein targets [18]. These two noncanonical interaction sites within the SOCS2-SH2 domain suggest that there may be additional protein interactions and regulatory mechanisms that impact on SOCS2 function. While our results confirm the critical role of SOCS2 in regulating GH signaling, the genetic relationship to mastitis in sheep (a bacterial infection of the udder) [20], indicates SOCS2 also has an important role in regulating the immune response to bacterial infection, most likely related to its up-regulation by LPS and IFNγ, and inhibition of NFkB signaling [24]. This suggests that SOCS2 regulates proteins independently of GHR-signaling complexes and this may occur either through classic SH2:pTyr interactions or via noncanonical SH2 interactions.

SOCS2 has been associated with reduced growth in children and diverse disease pathologies. Given that mouse and human SOCS2 share 94% sequence similarity, they are likely to have a conserved role as a critical negative regulator of GH signaling [18]. This is illustrated by correlation of a SOCS2 polymorphism with a positive response to GH therapy during puberty, and increased adult height of children with GH deficiency and Turner syndrome after long-term rhGH treatment [25]. Thus, inhibition or reduction in SOCS2 has the potential to enhance the efficiency of long-term rhGH therapy. SOCS2 mRNA levels were decreased in osteoarthritis patients [26], and a polymorphism in the SOCS2 gene correlated with susceptibility to type 2 diabetes in a Japanese population [27–31]. Reduced SOCS2 protein expression is associated with a poorer outcome in diverse human cancers, such as colorectal cancer [31], breast carcinoma [32], and hepatocellular carcinoma [33], suggesting SOCS2 may have some utility as a prognostic biomarker.

SOCS2 is also an important regulator in the immune system [34]. DC-mediated T cell priming and adaptive antitumoral immunity is enhanced in SOCS2-deficient mice, resulting in reduced tumor burden [35]. Loss of SOCS2 in CD4⁺ T cells results in increased T helper (Th) 2 cell differentiation and allergic inflammation [27], while Socs2^−/− CD4⁺ cells also show reduced Treg differentiation in vivo and in vitro under TGFβ stimulation [31]. Socs2^−/− mice display increased numbers of NK cells (consistent with our observations) and up-regulated JAK2 and STAT5 activation in response to IL-15 [15].

The canonical role of SOCS2 in regulating GH signaling cannot account for the diverse role of SOCS2 in immunity and inflammatory diseases. The Socs2^R96C/R96C mice provide an accessible model to investigate candidate pTyr targets for the SOCS2-SH2 domain in vivo, which will enable us to dissect the different pathways regulated by SOCS2, as well as the identification of novel SOCS2 targets.

pGEX4T constructs for expression of GST-human SOCS2 (residues 32-198), SOCS2-R73K (residues 32-198), and SOCS2-R96C (residues 32-198), and pACYCDuet constructs for expression of human Elongin B (residues 1-118) and Elongin C (residues 17-112), have been described previously [18]. The GST-SOCS2/BC trimeric complex was produced by co-expression in E. coli and purified by affinity chromatography as described previously [18]. The R73K mutation was introduced using the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Technologies). The primer for mutagenesis (AATGCGAGCTATCTTTAATCAAGAAAGTTCCTTCTGGTGCC) was from Integrated DNA Technologies (Singapore).

Human embryonic kidney cells (HEK293T) were pretreated with 10 μM MG132 for 4 h and sodium pervanadate for the final 30 min [36]. 2 × 10⁷ cells were lysed in 1 mL NP-40 lysis buffer (1% v/v NP-40, 50 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM EDTA, 10% glycerol, 1 mM PMSF, 1 mM Na₃VO₄, 1 mM NaF) and protease inhibitors (Complete Cocktail tablets, Roche). Cell lysates were clarified by centrifugation at 13000 g for 15 min at 4°C. Supernatant was precleared with glutathione-sepharose resin (GE Healthcare). A total of 5 μg GST or 10 μg GST-SOCS2 protein was then added to cell lysates and incubated for 2.5 h at 4°C. A total of 30 μL of 50% glutathione sepharose resin (GE Healthcare) was then added and incubated for 60 min at 4°C. Glutathione sepharose resins were then washed with 3 × 1 mL of NP-40 lysis buffer and boiled with 20 μL SDS sample buffer.

The affinity of SOCS2 and SOCS2-R73K protein for a phosphopeptide derived from the GHR receptor was measured by competitive surface plasmon resonance (SPR), as described previously [18]. Briefly, biotin-GSGS-GHR pY595 peptide was immobilized to a streptavidin-coated SA chip and 100 nM of SOCS2 proteins preincubated with titrations of GHR pY595 competitor peptide (10, 3.3, 1.1, 0.3, 0.1 μM) was flowed over the chip. Peptides were purchased from Genscript.

Mice carrying a germline mutation of Arg96 to Cys in SOCS2 (Socs2^R96C) were generated on a C57BL/6J background by the MAGEC laboratory at the Walter & Eliza Hall Institute, as previously described [37]. A total of 20 ng/μL Cas9 mRNA, 10 ng/μL single-guide RNA (CAGCTGGACCGACTAACCTG), and 40 ng/μL oligo donor template (TCGCATTCAGACTACCTACTAACTATATCCGTTAAGACGTCAGCTGGACCGACTAAtCTatGtATTGAGTACCAAGATGGGAAATTCAGATTGGATTCTATCATATGTGTCAAGTCCAAGCTT) were injected into fertilized one-cell stage embryos. Additional silent changes were incorporated into the donor template to enable genotyping by PCR. Founder mice were analyzed by NGS to confirm the correct sequence change. Mice carrying the mutation were backcrossed to C57BL/6J mice for three generations to eliminate potential off-target events, and NGS repeated. Heterozygous Socs2^R96C/+ mice were intercrossed to generate a homozygous Socs2^R96C/R96C line. Mice were routinely genotyped using genomic DNA extracted from ear biopsies with the Direct PCR Lysis tail reagent (Viagen) and 5 mg/mL proteinase K (Worthington), according to the manufacturer’s instructions. Sequencing primers are shown in Supplementary Table 1. Socs2^−/− mice have been described previously [9], and both Socs2^−/− and WT C57BL/6 mice were bred and housed in the same room as Socs2^R96C mice. Experiments were performed at the Walter and Eliza Hall Institute (WEHI) in accordance with the NHMRC Australian code for the care and use of animals for scientific purposes. All experiments were approved by the WEHI Animal Ethics Committee (AEC 2021.002). Animals were euthanised using CO₂ inhalation.

14-week-old Socs2^R96C/R96C and WT mice (n=3) were euthanized, and skeletal bones analyzed in intact carcasses by skeleton X-ray imaging in 3D using a Bruker Skyscan 1276 micro-CT. X-ray images were reconstructed into 3D volumes using Bruker’s NRecon software, skeletons were visualized, and length of individual bones measured in Imaris 8.7.

Dorsal skin sections were taken from 10-week-old male Socs2^R96C/R96C, Socs2^−/−, and WT mice (n=3), fixed with 10% buffered formalin, embedded in paraffin, and stained with Van Giessen stain using standard techniques. Images were captured using the Panoramic Scan II (3D HISTECH Ltd.).

Spleens, axial, and inguinal lymph nodes, and cardiac bleeds were collected from 8- to 12-week-old mice. Single-cell suspensions from spleens and lymph nodes were acquired by mashing organs through a 70 µM cell strainer with FACS buffer (phosphate-buffered saline (PBS) + 2% FCS). Cardiac bleeds were collected in K3-EDTA tubes, and whole blood separated by density gradient, underlaying Histopaque®-1077 (Sigma-Aldrich) and centrifuging at room temperature for 15 min at 400× g with low acceleration and brake. The opaque interface was collected and washed twice with FACS buffer. Spleen and lymph node samples were resuspended in 5 mL of red-cell removal buffer (156 mM NH₄Cl, 11.9 mM NaHCO₃, and 0.097 mM EDTA) for 5 min at room temperature and washed twice with PBS.

Single-cell suspensions were resuspended in FACS buffer and 20 µL removed to determine absolute cell numbers, using 123count eBeads (Invitrogen™; 10000 beads/tube) and propidium iodine (1 µg/mL). The remaining cells were incubated with Zombie UV (1:1000 in PBS) for 15 min at room temperature, washed twice with FACS buffer, FC blocked (FcR Blocking Reagent, mouse, Miltenyi; 1:100 in FACS buffer) for 10 min at 4°C in the dark, and stained for 30 min at 4°C with the relevant antibody cocktail (Table 1). Cells were washed twice with PBS and fixed with 100 µL BD Cytofix™ Fixation Buffer for 30 min at 4°C. Fluorescent labeling and fixation was performed in the dark. Finally, cells were washed twice with PBS and analyzed using a BD FACSymphony™ cell analyser. Data analysis was performed with FlowJo software v10.

Male and female Socs2^R96C/+ mice were set-up in timed matings, and e13 embryos harvested to generate Socs2^+/+, Socs2^R96C/+, and Socs2^R96C/R96C MEFs. Socs2^−/− MEFs were generated from Socs2^−/− matings. Heads and fetal livers were removed, and embryos mechanically disaggregated and digested using trypsin, before culturing in tissue culture plates coated in 0.1% gelatin/PBS. MEFs were cultured in DMEM (Thermo) supplemented with 100 U/mL penicillin, 0.1 ng/mL streptomycin, and 10% FBS (Thermo) at 37°C in a humidified incubator with 10% CO₂. MEFs at passage 5 were treated with 50 ng/mL of human GH for 0.5–8 h, lysed in NP-40 lysis buffer, and analyzed by immunoblotting. Human GH was kindly provided by Dr Andrew Brooks.

Cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretically transferred to nitrocellulose membranes (Amersham). Membranes were blocked overnight in 5% w/v BSA and incubated with primary antibody for 2 h. Anti-GST-HRP (RPN1236, 1:10000) was from Sigma-Aldrich. Anti-phosphotyrosine antibody (4G10) was obtained from Millipore. Anti-Elongin C antibody (610761, 1:3000) was from BD Biosciences. Anti-p44/42 (9102S, 1:2000), anti-STAT5 (94205; 1:3000), and anti-phospho-STAT5 (9359; 1:2000) were purchased from Cell Signaling. Anti-actin-HRP antibody (C4) was obtained from Santa Cruz (sc-47778 HRP; 1:1000). Anti-Cullin 5 (ab184177), anti-Elongin B antibody (ab154854, 1:2000), and anti-SOCS2 antibody (ab109245, 1:1000) were purchased from Abcam. Antibody binding was visualized with peroxidase-conjugated sheep antirabbit immunoglobulin (Southern Biotech; 4010-05; 1:15000), or sheep antimouse immunoglobulin (GE Healthcare; NA931-1ML; 1:10000) and the enhanced chemiluminescence (ECL) system (Amersham or Millipore).

Statistical analyses were performed with Prism 9 software. All statistical parameters including the exact value of n, the statistical test, error bars, and significance are reported in associated figure legends.

All supporting data are included within the main article and Supplementary file.

The authors declare that there are no competing interests associated with the manuscript.

This work was supported by an Australian Government National Health and Medical Research Council (NHMRC) Research Fellowship [grant number 1121755 (to J.J.B.)] and a Melbourne Research Scholarship (University of Melbourne) (to K.L.). This work was supported under a collaborative research project with Servier, and was supported in part through Victorian State Government Operational Infrastructure Support and the Australian Government NHMRC Independent Research Institutes Infrastructure Support Scheme (IRIISS).

Open access for the present article was enabled by the participation of University of Melbourne in an all-inclusive Read & Publish agreement with Portland Press and the Biochemical Society under a transformative agreement with CAUL.

Kunlun Li: Investigation, Writing—original draft. Lizeth G. Meza Guzman: Investigation, Writing—review & editing. Lachlan Whitehead: Software. Evelyn Leong: Investigation. Andrew Kueh: Resources. Warren S. Alexander: Resources, Visualization. Nadia J. Kershaw: Supervision, Visualization. Jeffrey J. Babon: Supervision. Karen Doggett: Supervision, Investigation, Writing—review & editing. Sandra E. Nicholson: Supervision, Visualization, Writing—original draft.

The authors thank Dr Andrew Brooks (The University of Queensland Diamantina Institute) for providing human growth hormone. The authors thank Natasha Blasch and Sophia Russo for mouse husbandry.

ANOVA

analysis of variance

CT

computed tomography

ECL

enhanced chemiluminescence

ESS

extended SH2 subdomain

FCS

fetal calf serum

GH

growth hormone

GHR

homodimeric GH receptor

HEK

human embryonic kidney cell

Igf1

insulin-like growth factor 1

JAK2

Janus kinase 2

LPS

lipopolysaccharide

MEF

mouse embryonic fibroblast

NGS

next-generation sequencing

NK

natural killer

P0

positively charged pocket

PBS

phosphate buffered saline

pTyr

phosphorylated tyrosine

S.D.

standard deviation

S.E.M.

standard error of the mean

SH2

Src homology 2

SOCS

suppressor of cytokine signaling

SPR

surface plasmon resonance

STAT

signal transducers and activators of transcription

Th

T helper

WT

wild-type