A novel leptin receptor antagonist uncouples leptin’s metabolic and immune functions
Leptin links body energy stores to high energy demanding processes like reproduction and immunity. Based on leptin’s role in autoimmune diseases and cancer, several leptin and leptin receptor (LR) antagonists have been developed, but these intrinsically lead to unwanted weight gain. Here, we report on the uncoupling of leptin’s metabolic and immune functions based on the cross talk with the epidermal growth factor receptor (EGFR). We show that both receptors spontaneously interact and, remarkably, that this complex can partially overrule the lack of LR activation by a leptin antagonistic mutein. Moreover, this leptin mutant induces EGFR phosphorylation comparable to wild-type leptin. Exploiting this non-canonical leptin signalling pathway, we identified a camelid single-domain antibody that selectively inhibits this LR-EGFR cross talk without interfering with homotypic LR signalling. Administration in vivo showed that this single-domain antibody did not interfere with leptin’s metabolic functions, but could reverse the leptin-driven protection against starvation-induced thymic and splenic atrophy. These findings offer new opportunities for the design and clinical application of selective leptin and LR antagonists that avoid unwanted metabolic side effects.
KeywordsLeptin Leptin receptor Receptor cross talk Uncoupling Metabolism Immunity
Bimolecular fluorescence complementation
Bioluminescence resonance energy transfer
Cytokine receptor homology
Epidermal growth factor receptor
- FN III
Fibronectin type III
Fluorescence resonance energy transfer
Granulocyte-colony stimulating factor
Insulin-like growth factor-I receptor
Kinase substrate sensor
Lipoprotein receptor-related protein
Rat pancreatitis associated protein
Secreted alkaline phosphatase
Signal transducer and activator of transcription
Transforming growth factor-alpha
Tumour necrosis factor
Time-resolved fluorescence energy transfer
Vascular endothelial growth factor
Variable domain of heavy chain antibody
The crucial role of leptin in long-term body weight control is well established. This cytokine with hormone-like characteristics is mainly, but not exclusively, produced and secreted by adipocytes, and plasma leptin levels positively correlate with body fat energy stores [1, 2]. Loss-of-function mutations in the leptin-encoding ob gene give rise to a complex syndrome that includes not only morbid obesity, but also abnormalities in lipid and glucose metabolism , haematopoiesis , innate and adaptive immunity (leading to increased risk of infection in mice and men) [5, 6, 7, 8, 9], reproduction , development , angiogenesis , vascular remodelling , blood pressure , and bone formation . Indeed, leptin is more than just a satiety signal and rather acts as a ‘metabolic switch’ that connects the body’s energy stores to high energy demanding processes like immunity and reproduction [16, 17]. In innate immunity, leptin promotes secretion of inflammatory cytokines and the activation of macrophages, neutrophils and natural killer cells . Functions in adaptive immunity include thymic homeostasis, naïve CD4+ cell proliferation, promotion of T helper 1 (TH1) responses and suppression of CD4+CD25high regulatory T cells (Tregs) . Consequently, leptin can contribute to the onset and progression of several T cell-controlled autoimmune diseases, including Crohn’s disease [19, 20], rheumatoid arthritis [21, 22], multiple sclerosis [23, 24] and autoimmune hepatitis [25, 26, 27]. Furthermore, clinical reports unequivocally link elevated serum leptin levels (caused by obesity) to an increased risk of certain cancers including prostate , breast [29, 30], colorectal , renal cancers  and multiple myeloma [33, 34].
Leptin mediates its effects upon binding and activation of the single-membrane spanning leptin receptor (LR), encoded by the db gene , and belonging to the class I cytokine receptor family. Its ectodomain is composed of two cytokine receptor homology (CRH1 and CRH2) domains, which are separated by an immunoglobulin-like domain (IGD) and is followed by two membrane-proximal fibronectin type III (FN III) domains. The CRH2 domain is necessary and sufficient for leptin binding [36, 37], but functional receptor clustering requires interaction with the IGD since leptin site III mutants that fail to contact this domain behave as leptin antagonists [38, 39]. Like all class I cytokine receptors, the LR lacks intrinsic kinase activity and relies for signalling on constitutively associated Janus tyrosine kinases (JAKs) . LR clustering results in JAK trans-phosphorylation and activation of several intracellular signalling cascades including the STAT (signal transducer and activator of transcription), MAPK (mitogen-activated protein kinase), PI3K (phosphatidylinositide 3-kinases) and mTOR pathways [41, 42].
In a recent study, we demonstrated the possible uncoupling of leptin’s metabolic and immune functions at the LR level . Indeed, the interaction between leptin’s site III (consisting of residues S120 and T121) and the LR IGD domain is strictly necessary for the control of body weight and metabolism, but is dispensable for immune signalling. We hypothesised that this non-canonical LR signalling may depend on receptor cross talk mechanisms, as documented for the insulin-like growth factor-I receptor (IGF-IR) [44, 45], epidermal growth factor receptor (EGFR) [46, 47] or the estrogen receptor-alpha (ERα) [48, 49].
Here, we identified a cross talk between LR and EGFR, defined the structural requirements for this cross talk and used this information to design a LR antagonist that selectively interferes with the immunomodulatory effects of leptin without interfering with the canonical LR signalling that controls the metabolic function.
Materials and methods
Cell culture, transfection and reagents
Hek293T (human embryonic kidney) cells (ATCC) were grown in DMEM (Dulbecco’s modified Eagle’s medium; Gibco, life technologies) with 4500 mg/l glucose and 10% FCS (foetal calf serum; Invitrogen) in a 10% CO2 humidified atmosphere at 37 °C. Cells were transfected with standard calcium phosphate precipitation. Recombinant mouse leptin and leptin S120A-T121A were purified and refolded from E. coli inclusion bodies . To prolong the half-life in circulation, leptin was pegylated with a 40 kDa PEG (polyethylene glycol) variant (Sunbright) according to the manufacturer’s instructions. Generation and screening of a mouse leptin receptor-specific VHH library has been described previously . Mono- and bispecific (genetically fused to the sequence of anti-mouse serum albumin (mAlb) via a flexible GGS-linker) VHHs were cloned in the pHEN6C vector for periplasmic E. coli expression. Proteins were purified after osmotic shock using the C-terminal 6*His-tag with the TALON metal affinity resin (Clontech). Imidazole (Merck) was used for elution and removed using PD-10 gel filtration columns (GE Healthcare). Protein concentrations were determined using the absorbance at 280 nm (Nanodrop; ThermoFisher Scientific) and purity via SDS-PAGE. Endotoxins were washed away during purification with 0.5% EMPIGEN (Calbiochem, Millipore) and 0.5% CHAPS (Sigma-Aldrich) and afterwards quantified using limulus amoebocyte lysate (LAL) QCL-1000 (Lonza). If still present, endotoxins were removed using Pierce Endotoxin Removal Resin (Thermo Scientific).
Hek293T cells were transfected with tagged EGFR, LR, LR-FATT variants or combinations thereof. 24 h later, cells were lysed in modified RIPA buffer and cleared lysates were subjected to immunoprecipitation with anti-FLAG M2 affinity resin (Sigma). Proteins in cell lysates (input) and after co-immunoprecipitation were analysed using SDS-PAGE and Western blotting using mouse anti-FLAG (Sigma) and mouse anti-E-tag antibodies (Bethyl), diluted according to the manufacturer’s guidelines in StartingBlock blocking buffer (Pierce) supplemented with 0.1% Tween-20, and revealed using the SuperSignal West Pico Chemiluminescent Substrate (Pierce) after incubation with HRP-conjugated anti-mouse secondary antibody (Jackson ImmunoResearch).
Luciferase reporter assay
Hek293T cells were transiently transfected with (a combination of) receptors and the STAT3-responsive pXP2d2-rPAP1 (rat pancreatitis associated protein 1)-luciferase reporter . Transfected cells were stimulated overnight as indicated. Cell lysates were prepared [lysis buffer: 25 mM Tris/HCL, pH 7.8, 2 mM EDTA, 2 mM DTT (dithiothreitol), 10% glycerol and 1% Triton X-100], and 35 µl of luciferase substrate buffer [20 mM Tricine, 1.07 mM (MgCO3)4 Mg(OH)2·5H2O, 2.67 mM MgSO4·7H2O, 0.1 mM EDTA, 33.3 mM DTT, 270 µM coenzyme A, 470 µM luciferin and 530 µM ATP, final pH 7.8] was added per 50 µl of lysate. Light emission was measured for 5 s in a TopCount chemiluminescence counter (Packard).
EGFR phosphorylation was analysed in transfected Hek293T cells that were serum starved overnight and left untreated or stimulated with indicated amounts of leptin or leptin antagonist for 5 min. Lysates were prepared in Laemmli loading buffer and proteins were blotted overnight. Phosphorylated and total protein levels were visualised using rabbit anti-phospho-EGFR (Tyr1068) or EGFR antibodies (both Cell Signalling Technology) and an HRP-conjugated anti-rabbit secondary antibody (Jackson ImmunoResearch) as described earlier.
Time-resolved fluorescence energy transfer (TR-FRET) measurement
The TR-FRET saturation curve is generated in conditions where the expression of the energy donor (EGFR) is kept constant with increasing expression of the energy acceptor (LR) [52, 53]. Hek293 cells were therefore transfected with plasmid vectors encoding SNAP-tagged EGFR (10 ng), and HALO-tagged LR (0–170 ng) constructs (completed to 200 ng/well of 96-well plate with noncoding plasmid) were plated in polyornithine-coated 96-well white plates. Two days after transfection, adherent cells were washed and incubated with Taglite buffer (Cisbio Bioassay) containing 200 nM HALO-Tb and/or SNAP-d2 or SNAP-Tb substrates (Cisbio Bioassay) for 1 h at 4 °C. Cells were washed four times with Taglite buffer and the fluorescence of Lumi4-Tb (excitation at 340 nm, emission at 620 nm, 150 μs delay, and 500 μs integration time) and the TR-FRET signals (340 nm, 665 nm, 150 μs, and 500 μs) were read using an Infinite F500 spectrofluorimeter (Tecan). Background signals were measured on cells transfected with empty vectors and subtracted to obtain the specific signal.
Kinase substrate sensor (KISS) experiments
The extracellular and transmembrane domains of the EGFR, LR, LR-FATT and the granulocyte-colony stimulating factor receptor (G-CSFR) were fused as prey proteins to the cytoplasmic tail of glycoprotein 130 (gp130), or as bait proteins to a C-terminal fragment of human Tyk2 (Aa 589–1187) as described in . In a standard KISS experiment, combinations of equal amounts of bait and prey DNAs together with the STAT3 responsive pXP2d2-rPAP1-luciferase reporter were transiently transfected in Hek293T cells . Luciferase activity was measured 2 days after transfection as described earlier.
Specificity of the different VHHs was determined in a plate-binding assay using a series of mouse LR extracellular subdomains (or a combination thereof) fused to secreted alkaline phosphatase (SEAP) . VHHs were bound via a penta-His antibody (Qiagen) on Maxisorp plates (Costar). Excess unbound VHH was washed away with washing buffer (PBS + 0.05% Tween-20) and plates were blocked (PBS + 0.1% Casein) and further incubated with conditioned medium containing LR-SEAP fusions for 2 h at room temperature. After 4 successive washes, bound SEAP activity was measured using the chemiluminescent CSPD substrate (PhosphaLight, Tropix) in a TopCount chemiluminescence counter (Canberra Packard).
Affinities of the different VHHs for the LR were determined in a similar plate-binding assay as described before: VHH-coated plates were incubated with a serial dilution of the mouse LR extracellular domain fused to SEAP (mLREC-SEAP)  for 2 h. The GraphPad Prism software was used to calculate the affinities from the bound SEAP measurements.
Serum VHH quantification
Penta-His-coated plates were incubated with a serial dilution of purified VHH as a standard, or a 1:1000 dilution of mouse serum samples for 2 h, and subsequently allowed to bind mLREC-SEAP. Serum VHH concentrations were calculated from the standard curves using the GraphPad Prism software.
In vivo experiments
Mice were maintained in pathogen-free conditions in a temperature-controlled environment with 12/12 h light/dark cycles and received food and water ad libitum.
To evaluate the metabolic effects, 8-week-old female C57BL/6 mice were treated with 200 µg/mouse/day of mAlb fusions of the selected VHHs intraperitoneally (i.p.) or with PBS for 2 weeks. On a daily basis, animals were weighed and food intake (average food intake per cage/number of mice in cage) and body temperature were monitored. At the end of the experiment, blood was collected for the measurement of VHH concentrations (see above) and serum leptin levels using the mouse leptin DuoSet ELISA (R&D Systems Europe Ltd). Further haematological analysis was performed using a Hemavet 950FS (Drew Scientific) whole blood counter. From each group, one mouse was used for a glucose tolerance test: mice were fasted for 6 h, i.p. injected with 2 g glucose/kg body weight and glucose levels were measured from tail vein blood at the indicated time points using AlphaTRAK 2 Blood Glucose Test Strips (Abbott). After animals were killed via CO2 asphyxiation, subcutaneous and visceral fat tissue was dissected and weighed.
To evaluate immune-related effects, 8-week-old male C57BL/6 mice were fed ad libitum or starved for 48 h. The latter group either received a daily i.p. injection of PBS, pegylated leptin (50 µg/mouse/day) or a combination of pegylated leptin (50 µg/mouse/day) and 4.80-mAlb (200 µg/mouse/day). At the end of the experiment, animals were killed via CO2 asphyxiation and spleen and thymi were isolated. Organs were weighed and the number of white blood cell, lymphocytes, neutrophils and monocytes in homogenates were quantified using the Hemavet 950FS (Drew Scientific) according to the manufacturer’s guidelines.
Data were analysed with the GraphPad PRISM software using one-way ANOVA followed by Tukey multiple-comparison tests. Results with P value less than 0.05 were considered statistically significant.
Ligand-independent LR-EGFR clustering
Effect of EGFR on non-canonical LR signalling
Interfering with LR-EGFR cross talk
We applied a plate-binding assay using a mLREC-SEAP fusion protein to determine the affinity of the different VHHs. Figure 3d shows that VHH 4.80 binds the receptor with low nanomolar affinity (0.47 ± 0.017 nM) comparable to VHH 2.17 (0.544 ± 0.065 nM) and VHH 4.10 (a neutralising IGD-specific VHH; 0.66 ± 0.029 nM). A similar plate-binding assay using different LR subdomains or combinations thereof was then used to show that VHH 4.80 binds to the LR FN III domains (Fig. 3e). VHHs 2.17 and 4.10 only bind CRH2 or IGD containing chimeras, respectively.
Effect of VHH 4.80 on metabolism
Effect of VHH 4.80 on leptin immune signalling
Cytokine receptors often appear as inactive, pre-formed complexes on the cellular surface in the absence of ligands. Well-studied examples include the erythropoietin (Epo) receptor , the growth hormone (GH) receptor , and the interleukin-6 (IL-6) receptor . Bimolecular fluorescence complementation (BiFC) experiments illustrated not only that all members of the EGFR family form homo- and heterodimers in the absence of their respective ligands, but also that this clustering already occurs in the endoplasmic reticulum (ER) and depends on the intracellular receptor domains [62, 63]. Similarly, the high basal signal in the absence of leptin in bioluminescence resonance energy transfer (BRET)  and FRET , and co-immunoprecipitation experiments [66, 67] suggest that the LR forms pre-formed dimers (or oligomers) on the cellular surface. This clustering likely involves cysteine residues in the CRH2 domain and, as for the EGFR, already occurs intracellularly . Our co-immunoprecipitation and TR-FRET or adapted KISS experiments in Fig. 1 showed that both the EGFR and LR not only form homodimers (or oligomers), but also heterodimers (or oligomers) with each other on the cellular surface. TR-FRET experiments at the plasma membrane with fluorophores coupled to the N-terminus of the extracellular part of the receptors demonstrate close proximity between the extracellular chains of both receptors (Fig. 1b). Inversely, lack of FRET signal between C-terminal tagged receptors, observed by Eisenberg and colleagues , might suggest that the FRET probes on the cytoplasmic domains are likely too distant for the energy transfer. The clustering between LR and EGFR appears to be independent of the LR IGD, since LR-FATT, a naturally occurring and a signalling-deficient LR variant lacking the complete IGD identified in obese fatt/fatt mice , is as efficient in immunoprecipitating the EGFR as wild-type LR. Likewise, co-expression of EGFR and LR-FATT KISS bait and prey proteins resulted in clear STAT3-dependent signalling.
The EGFR can be activated by its “natural” ligands (including EGF, transforming growth factor-α (TGF-α), heparin-binding EGF-like growth factor, amphiregulin, and epiregulin), but also by cross-activation with G-protein-coupled receptors, the tumour necrosis factor receptor (TNFR), or IGF-IR (reviewed in [70, 71, 72, 73]). Activation of the EGFR is widely recognised as a potent signal in the regulation of diverse biological processes including cell proliferation, survival and motility, depending on the cellular system, the stimulus and the repertoire of signalling molecules recruited . Likewise, not only leptin binding, but also cross talk with ERα, IGF-IR, lipoprotein receptor-related protein 1 and 2 (LRP1 and LRP2), and vascular endothelial growth factor (VEGF) receptor has been reported to activate the LR .
Cross talk and cross-activation between LR and EGFR is well documented mainly in cancer cell lines, but also in muscle, salivary gland, and mucosal cells and in rat kidneys (reviewed in ). Several of these studies showed reciprocal cross-phosphorylation of both receptors and observed that EGFR inhibitors can inhibit leptin-induced signalling events. Both long and short isoforms of the LR were shown to phosphorylate the EGFR family member ErbB2 (both kinase positive and kinase negative) in a leptin-dependent way, suggesting that this cross-activation is independent of downstream LR signalling . The principal function of EGFR is to regulate epithelial tissue development and homeostasis. In pathological settings, such as lung and breast cancer, the EGFR is a driver of tumourigenesis . Aberrant activation of the EGFR and its downstream pro-oncogenic signalling pathways, such as the MAPK and PI3K pathway, promotes cancer cell proliferation, including their chronic initiation and progression through the cell cycle (reviewed by ). Leptin was shown to participate in human breast cancer progression and metastasis  and induces, via EGFR phosphorylation, clonogenicity, anchorage-independent growth, migration, Notch activation and survivin upregulation in human breast cancer cells , illustrating that this LR-EGFR cross talk also has possible biological implications in the context of EGFR signalling.
The interaction between the LR IGD and leptin’s binding site III is essential for the formation of an activated LR complex and downstream signalling. Deletion of this domain results in a receptor completely devoid of biological activity [36, 37]. Likewise, mutation of residues in leptin involved in this interaction resulted in an antagonist both in vitro and in vivo [38, 39]. In strong contrast, we gathered several lines of evidence that LR-mediated immune signalling, but not the control of body weight, can occur in the absence of this interaction: (i) the spontaneous natural S120C mutation in the leptin gene gives rise to high incidence of early-onset obesity in the Nochurli population. Surprisingly, affected patients are still fertile and not more susceptible to infections (unpublished data). This latter is in strong contrast to earlier described human loss-of-function mutations in leptin or LR genes . (ii) Fatt/fatt mice carry a spontaneous splice mutation that causes deletion of the complete IGD in all LR isoforms . These animals are hyperphagic and morbidly obese, but, in contrast to other leptin or LR-deficient rodents, display only minimal changes in size and cellularity of the thymus and respond to concavalin A in a model for autoimmune hepatitis. (iii) Treatment of healthy mice with an IGD-specific neutralising VHH induced weight gain and hyperinsulinemia, but failed to block development of experimentally induced autoimmune multiple sclerosis, arthritis and hepatitis .
In this study we now show that the leptin S120A-T121A antagonist, which is unable to interact with the LR IGD, is capable of inducing EGFR phosphorylation comparable to wild-type leptin (Fig. 2a). Furthermore, co-expression of EGFR partially restores STAT3-dependent signalling by this mutant through the LR (Fig. 2b). The signalling capacities of the leptin S120A-T121A mutant suggest that the structural requirements for LR signalling vs LR-EGFR cross talk appear to be fundamentally different. To further explore this hypothesis, we evaluated the effect of mouse LR-selective VHHs from a previously designed library  on both types of signalling. We found that VHH 4.80, directed against the LR FN III domains, selectively interfered with LR-EGFR cross talk (measured by leptin S120A-T121A-induced STAT3 signalling in LR and EGFR co-transfected cells), but not with canonical LR signalling (leptin stimulation of LR expressing cells)(Fig. 3), while VHH 2.17, which was previously shown to interfere with leptin binding to its cognate receptor , clearly blocks both pathways. VHH 4.80 also blocks leptin or leptin S120A-T121A driven EGFR phosphorylation.
In a next step, we compared the effect of the selective VHH 4.80 with previously characterised, neutralising VHHs 2.17 and 4.10  on metabolism in healthy mice. All VHHs were coupled to a mouse serum albumin-specific VHH (mAlb) to prolong half-life in circulation . Both VHH 2.17 and 4.10 treatments resulted in a significant increase in daily food intake, body weight and subcutaneous and visceral fat mass, while a 14-day treatment with VHH 4.80 did not alter these parameters in a significant manner (Fig. 4). This difference could not be explained by a different accumulation of the VHHs in circulation, since VHH 4.80 levels were comparable to VHH 2.17 levels and higher than VHH 4.10 levels after 2 weeks of treatment. No significant differences in basal blood glucose levels and glucose tolerance were observed between the treated groups, which is in line with the study of Levi et al. where 3 days treatment with a pegylated mouse leptin antagonist (L39A/D40A/F41A) did not alter glucose disposal .
The LR-EGFR cross talk could provide a molecular explanation for the LR immune signalling in the absence of the IGD interaction. EGFR is known to be critically involved in tissue development and homeostasis as well as in the pathogenesis of cancer. EGFR antagonists were one of the first anti-cancer treatments developed targeting a receptor tyrosine kinase. Several studies illustrate that a substantial part of the clinical responses observed following EGFR targeting treatments may not only be mediated by direct effects on the tumour, but also by regulation of immune responses, an aspect of EGF biology that remains undervalued . For example, the EGFR is expressed on many immune cells including macrophages , monocytes , plasma cells  and some T cell subsets such as effector CD4+ T cells and FoxP3-expressing Treg cells , thereby guiding the cellular functionality. The observation that VHH 4.80 blocks leptin-driven EGFR phosphorylation and the partial restoration of signalling by leptin S120A-T121A upon LR-EGFR co-expression prompted us to study the role of LR-EGFR cross talk in vivo. We could show that VHH 4.80, at least in part, reverses the leptin-driven protection against starvation-induced thymic and splenic atrophy measured by cellularity and weight of these organs (Fig. 5). In line with the latter, starvation conditions also severely decrease the number of lymphocytes, neutrophils and monocytes, and leptin reverses those effects, while VHH 4.80 interferes with the leptin-induced normalisation of these immune cellular counts.
Together, our data show that the structural requirements for signalling via the LR or LR-EGFR cross talk are fundamentally different. Based on this observation we were able to identify a VHH that selectively inhibits the cross talk, but not the canonical LR activation mechanism. Daily administration of this VHH apparently did not alter any of the tested metabolic parameters, but had significant effects on leptin signalling in the immune compartment as observed in our starvation experiments. Leptin and LR antagonists (leptin mutants, leptin-derived peptides, neutralising antibodies and soluble LR variants) were shown to have a clear potential as therapeutics for the treatment of several autoimmune diseases and cancer . However, the major drawback of most, if not all, current leptin and LR antagonistic strategies is that they also give rise to considerable undesirable weight gain. Uncoupling of leptin’s metabolic and immune functions, for example based on the EGFR cross talk, might revitalise the potential of leptin and LR antagonists as a therapeutic strategy.
This work was supported by a European Research Council (ERC) Advanced Grant (CYRE-No. 340941 (JT)) and by grants from the « Agence Nationale de la Recherche » ANR-15-CE14-0025 (JD & RJ), the département hospitalo-universitaire (DHU) « autoimmune and hormonal diseases» , the “Who am I?” laboratory of excellence No. ANR-11-LABX-0071 (JD).
Compliance with ethical standards
Conflict of Interests
The authors declare that they have no conflict of interest.
- 20.Karmiris K, Koutroubakis IE, Kouroumalis EA (2005) The emerging role of adipocytokines as inflammatory mediators in inflammatory bowel disease. Inflamm Bowel Dis 11:847–855. https://doi.org/10.1097/01.mib.0000178915.54264.8f CrossRefGoogle Scholar
- 39.Niv-Spector L, Gonen-Berger D, Gourdou I et al (2005) Identification of the hydrophobic strand in the A-B loop of leptin as major binding site III: implications for large-scale preparation of potent recombinant human and ovine leptin antagonists. Biochem J 391:221–230. https://doi.org/10.1042/bj20050457 CrossRefGoogle Scholar
- 45.Saxena NK, Taliaferro-Smith L, Knight BB et al (2008) Bidirectional crosstalk between leptin and insulin-like growth factor-I signaling promotes invasion and migration of breast cancer cells via transactivation of epidermal growth factor receptor. Cancer Res 68:9712–9722. https://doi.org/10.1158/0008-5472.CAN-08-1952 CrossRefGoogle Scholar
- 46.Chao H-H, Hong H-J, Liu J-C et al (2007) Leptin stimulates endothelin-1 expression via extracellular signal-regulated kinase by epidermal growth factor receptor transactivation in rat aortic smooth muscle cells. Eur J Pharmacol 573:49–54. https://doi.org/10.1016/j.ejphar.2007.06.051 CrossRefGoogle Scholar
- 47.Ogunwobi O, Mutungi G, Beales ILP (2006) Leptin stimulates proliferation and inhibits apoptosis in Barrett’s esophageal adenocarcinoma cells by cyclooxygenase-2-dependent, prostaglandin-E2-mediated transactivation of the epidermal growth factor receptor and c-Jun NH2-terminal kinase activation. Endocrinology 147:4505–4516. https://doi.org/10.1210/en.2006-0224 CrossRefGoogle Scholar
- 51.Broekaert D, Eyckerman S, Lavens D et al (2002) Comparison of leptin- and interleukin-6-regulated expression of the rPAP gene family: evidence for differential co-regulatory signals. Eur Cytokine Netw 13:78–85Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.