Proinflammatory switch from Gαs to Gαi signaling by Glucagon-like peptide-1 receptor in murine splenic monocyte following burn injury
- 237 Downloads
Glucagon-like peptide-1 (GLP-1)-based therapy via G protein-coupled receptor (GPCR) GLP-1R, to attenuate hyperglycemia in critical care has attracted great attention. However, the exaggerated inflammation by GLP-1R agonist, Exendin-4, in a mouse model of burn injury was quite unexpected. Recent studies found that GPCR might elicit proinflammatory effects by switching from Gαs to Gαi signaling in the immune system. Thus, we aimed to investigate the possible Gαs to Gαi switch in GLP-1R signaling in monocyte following burn injury.
Materials and methods
Splenic monocytes from sham and burn mice 24 h following burn injury were treated with consecutive doses of Exendin-4 alone or in combination with an inhibitor of Gαi signaling (pertussis toxin, PTX), or a blocker of protein kinase A (H89). Cell viability was assessed by CCK-8, and the supernatant was collected for cytokine measurement by ELISA. Intracellular cAMP level, phosphorylated PKA activity, and nuclear NF-κB p65 were determined by ELISA, ERK1/2 activation was analyzed by Western blot. The expression of GLP-1R downstream molecules, Gαs, Gαi and G-protein coupled receptor kinase 2 (GRK2) were examined by immunofluorescence staining and Western blot.
Exendin-4 could inhibit the viability of monocyte from sham rather than burn mice. Unexpectedly, it could also reduce TNF-α secretion from sham monocyte while increase it from burn monocyte. The increased secretion of TNF-α by Exendin-4 from burn monocyte could be reversed by pretreatment of PTX or H89. Accordingly, Exendin-4 could stimulates cAMP production dose dependently from sham instead of burn monocyte. However, the blunt cAMP production from burn monocyte was further suppressed by pretreatment of PTX or H89 after 6-h incubation. Nevertheless, phosphorylated PKA activity was significantly increased by low dose of Exendin-4 in sham monocyte, by contrast, it was enhanced by high dose of Exendin-4 in burn monocyte after 1-h incubation. Following Exendin-4 treatment for 2 h ex vivo, total nuclear NF-κB and phosphorylated NF-κB activity, as well as cytoplasmic pERK1/2 expressions were reduced in sham monocyte, however, only pERK1/2 was increased by Exendin-4 in burn monocytes. Moreover, reduced expressions of GLP-1R, GRK-2 and Gαs in contrast with increased expression of Gαi were identified in burn monocyte relative to sham monocyte.
This study presents an unexpected proinflammatory switch from Gαs to Gαi signaling in burn monocyte, which promotes ERK1/2 and NF-κB activation and the downstream TNF-α secretion. This phenomenon is most probably responsible for proinflammatory response evoked by Gαs agonist Exendin-4 following burn injury.
KeywordsGlucagon-like peptide 1 G protein-coupled receptor Monocyte Inflammation Signal transduction Burn injury
Drugs targeting the glucagon-like peptide 1 receptor (GLP-1R), a Gs-coupled incretin receptor, to attenuate hyperglycemia with reduced hypoglycemic episodes and increased sensitivity to insulin in the field of critical care, e.g., burn injury, has attracted great attention . However, safety concerns, including pancreatitis and gastrointestinal distress, have lead to non-compliance in diabetes and the etiology has not been identified yet. Accumulating studies have suggested that GLP-1 and/or its analogues, e.g., Exendin-4 (Ex-4), be involved in immunoregulation. Increased serum GLP-1 secretion was found in patients with sepsis, metabolic syndrome or critically illness , as well as in the mice under inflammatory stimuli . In adaptive immunity, Ex-4 therapy could result in increased frequency of Tregs and decreased frequency of CD4+CD8+ T cells in non-obese diabetic (NOD) mice, as well as decreased frequency of iNKT cells in psoriasis patients. In addition, it could also reduce CD4+ T cell migration, decrease pro-inflammatory cytokine secretions in mice  and in diabetic patients. In innate immunity, Ex-4 could inhibit the recruitment and activation of macrophage , and reduce the secretions of pro-inflammatory mediators in monocytes by inactivation of NF-κB [6, 7]. It could also promote M2 polarization in macrophage, contributing to the protective effects of GLP-1 against diabetes and cardiovascular diseases .
Typically, GLP-1 exerts its anti-inflammatory effect through the conventional G protein-coupled receptor (GPCR)/cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) pathway. The GLP-1R is coupled to multiple signaling effectors, leading to cAMP signaling, Ca2+ mobilization, and pERK1/2 activation. In most situations, GLP-1R mainly activates the Gαs subunit to stimulate cAMP production; however, in some situations, such as in hypoxia and sepsis, the responses including PTX-sensitive, inhibitory Gαi couplings have been observed. Unlike Gsα binding, GPCR signalling via Giα could decrease the adenylate cyclase activity and increase the receptor-stimulated mitogen-activated-protein-kinase (MAPK) activation [9, 10]. Recently, the GLP-1 mimetic exenatide was found to display significant bias for Gαi pathway in the yeast chimeric Gα system  and generating low cAMP signals. Since GLP-1R therapy has been widely applied in diabetes, adverse effects on organ inflammation following prolonged use have emerged. Recently, the acknowledged anti-inflammatory effect of Ex-4 has been challenged. In preclinical studies, acute Ex-4 administration robustly induced the expression of genes encoding cytokines and chemokines in normal and injured intestine , whereas GLP-1R activation in the brain or in the mouse neuronal cell line resulted in increased IL-6 expression , indicating pro-inflammatory effect of GLP-1-based therapy. Considering that Type 2 diabetes and critical illness shared similar metabolic features, such as hyperglycemia, insulin resistance, and systemic inflammation [13, 14], the great potential for GLP-1-based therapy in critical illness was suggested, including the burn injury [15, 16] and major surgery . Therefore, the thorough understanding of GLP-1 signaling in immune cells, especially in the setting of critical care is of increasing interest.
Burn injury could elicit dramatic alterations in both adaptive and innate immune responses. Whist the T cells underwent depletion and apoptosis, along with a Th2 biased response following burn injury , the monocyte was also impaired with profound decrease in human leukocyte antigen-DR and an increase in C–C chemokine receptor type 2 (CCR2) expression , and also an increase in TNF-α production . We previously found that Ex-4 therapy could worsen burn-induced mortality in mice with immunosuppressive Th2 response and augmented inflammation (unpublished data). The unexpected findings suggested that Ex-4 might exert distinct immunoregulatory effect on burn-injured mice from sham-injured mice. Several studies had reported the imbalanced Gsα/Giα ratio in rat liver , cardiocytes  and human monocytes  in sepsis. Whether it also applies to monocytes from burn-injured mice and mediates the proinflammatory response to Ex-4 therapy has remained intriguing. Therefore, it is urgent to investigate the GLP-1R signaling in monocytes from the burn-injured mice ex vivo to assess the long-term safety of GLP-1-based therapy in critical illness, e.g., burn injury.
Materials and methods
Animal model of thermal injury
Male Balb/c mice (20–25 g, Huafukang Bioscience Co. Inc, Beijing, China) were acclimatized for 1 week before being exposed to 94 °C thermal injury for 8 s of 15% total body surface area after anesthetized with diethyl ether as previously reported . Sham-injured mice were subjected to all the procedures except the bath was 37 °C. Mice received 1.0 ml of Ringer’s solution subcutaneously immediately after injury for fluid resuscitation. A topical antibacterial agent (iodine tincture) was applied on the wound. The mice were caged individually with warm bedding in a temperature and humidity-controlled room with 12 h light and 12 h darkness. Food and water were provided ad libitum. Animals were sacrificed, between 9:00 a.m. and 10:00 a.m. by cervical dislocation. Analgesics were not used owning to the profound immunosuppressive effects of opiates . All experimental manipulations were undertaken in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, with the approval of the Scientific Investigation Board of the Chinese PLA General Hospital, Beijing, China.
Mice were killed by dislocation 24 h following injury, and splenocyte suspensions were obtained by flushing the spleen through the 70-µm mesh. The red blood cells were removed by hypertonic lysis. Cell counting was performed by trypan blue exclusion on a hemocytometer count to determine total cell numbers per spleen. Cell viability was consistently greater than 98%. Macrophage was isolated by two successive series of plastic adherence to a fetal calf serum-treated plastic dish for 2 h at 37 °C in a humidified atmosphere with 5% CO2. Non-adherent cells were removed from dishes with warm PBS, and the resulting adherent cells were scraped with a rubber policeman and represented the macrophage-purified population. Greater than 90% of the adherent cells were positive for CD11b+ staining.
Treatment of monocytes
Monocytes from both sham and burn mice were grown in complete RMPI 1640 media supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were seeded on 96-well plates (2 × 105 per well) and treated with vehicle (PBS), or consecutive doses of Ex-4 (E7144, Sigma, St. Louis, MD, USA) 1 h prior to lipopolysaccharide (LPS) stimulation (100 ng/ml, Sigma) at 37 °C and 5% CO2 for another 23 h. Cell viability was assessed using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan), and the supernatant was collected for cytokine detection. For G protein signaling study, 1 h prior to Ex-4 treatment (0.3 nmol/L) , monocytes from both sham and burn mice were seeded on 24-well plates (4 × 105 per well) and treated with a specific inhibitor of Gαi-mediated signaling, pertussis toxin (PTX, 100 ng/ml, p2980, Sigma) , or an inhibitor of protein kinase (PKA) H89 (10 µM, Beyotime Biotechnology, Jiangsu, China) . Finally, LPS (100 ng/ml) was added and cells were incubated for total 24 h . The supernatants were collected for TNF-α and IL-10 analysis, each sample was performed in triplicate (ExCell Biology Inc., Shanghai, China).
cAMP and phosphorylated PKA assay
For cAMP time-dependent study, adherent purified monocytes/macrophages (1.2 × 106 per well) in 12-well plate were incubated with vehicle, differing doses of Ex-4 for 30 min, 2 h and 6 h at 37 °C/5% CO2 in the presence of 200 µM phosphodiesterase inhibitor IBMX (sc-201188, Santa Cruz Biotechnology, Inc, USA) . For G protein signaling study, monocytes from both sham and burn mice were treated with PTX (100 ng/ml, Sigma) or H89 (10 µM, Beyotime Biotechnology) , and then incubated with Ex-4 (0.3 nmol/L, Sigma) in the presence of 200 µM IBMX for another 6 h. All the cells were lysed in 0.1N HCl at room temperature for approximately 10 min and intracellular cAMP levels were assessed using the cAMP direct immunoassay kit (ADI-900-066, Enzo life sciences Inc, Farmingdale, NY, USA). Data were expressed as picomoles cAMP per 1 million cells.
For phosphorylated PKA activity study, monocytes (1.2 × 106 per well) in 12-well plate were treated with vehicle, differing doses of Ex-4 alone for 1 or 3 h, or with PTX and H89 pretreatment for 1 h. PKA activity was determined in whole-cell lysates using the ELISA PKA assay kit (ADI-EKS-390A, Enzo life sciences). Data were expressed as ng/ml of trilplicates wells represented of two experiments.
Immunofluorescence analysis of GLP-1R signaling molecules
Purified monocytes from sham and burn mice were blocked with 3% BSA in PBS for 2 h before being incubated overnight at 4 °C with one of the following antibodies: Gsα (1:100, ab83735,Abcam), Giα (1:200, sc-13534, Santa Cruz), G protein-coupled receptor kinase 2 (GRK2) (1:200, sc-13143, Santa Cruz) and GLP-1R (1:50; sc-66911, Santa Cruz). Cells incubated only in the dilution buffer were set as negative control. Then the tissue was incubated for 2 h at room temperature in dark in DyLightTM 549-conjugated AffiniPure goat anti-rabbit IgG (H + L) Abs (1:200; Jackson ImmunoResearch Laboratories, West Grove, USA). Slides were rinsed and coverslips were affixed with VECTASHIELD mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) for nucleus staining. Control experiments were performed with unspecific isotype antibodies of rabbit or mouse instead of the primary antibody. Fluorescent cell images were obtained on a confocal laser scanning microscope (LSM 700; Carl Zeiss MicroImaging, Jena, Germany).
Immunoblotting of GLP-1R downstream signaling molecules
Purified monocytes from sham and burn mice were homogenized in lysis buffer (150 mmol/L NaCl, 50 mmol/L Tris–HCl, pH 7.8, 1% Triton X-100, 1 mmol/L EDTA, 0.5 mmol/L phenylmethanesulfonyl fluoride) with a protease inhibitor and phosphatase inhibitor (Applygen Technologies Inc., Beijing, China). Samples were centrifuged, and protein concentrations were measured by BCA assay (Applygen Technologies Inc.). A total of 25 µg of denatured protein from cells was separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) gel and transfered to polyvinylidene difluoride membrane (Millipore). The membrane was blocked with 5% nonfat milk solution and washed with buffer (Tris buffered saline, 0.05% Tween) prior to incubation of primary antibodies for GLP-1R (1:500, Santa Cruz), GRK2 (1:500, Santa Cruz), Gsα (1:1000, Abcam), or Giα (1:200, Santa Cruz). After extensive washing, protein bands were visualized by ECL reagent (Applygen Technologies Inc.) and exposed on ImageQuant LAS 4000 system (GE Healthcare, Germany). Membranes were then stripped and re-probed with β-actin (1:1000; Santa Cruz) to confirm equal protein loading. The band intensities relative to β-actin were quantified by ImageJ software.
Immunoblotting of extracellular signal-regulated kinase (ERK)1/2 and nuclear NF-κB p65
Monocytes from both sham and burn mice were seeded on 6-well plates (2 × 106 per well) and were treated with consecutive doses of Ex-4 or PBS for 2 h. Then, the cells were lysed with RIPA buffer to generate total protein for p-ERK1/2 determination, or nuclear extracts for NF-κB p65 determination (Applygen Technologies Inc.). Cellular or nuclei lysates (10 µg/lane) were electrophoresed, transferred to polyvinylidene difluoride membranes, and probed with antibody to p-ERK1/2 (1:1000, Cell signaling) or NF-κB p65 (1:1000, sc-372, Santa Cruz). After incubation with HRP-conjugated secondary antibody, specific proteins were revealed using ECL (Applygen Technologies Inc.). After retrieving, the membrabes were incubated with antibodies to ERK1/2 (1:1000, Cell signaling), or nuclear Lamin B protein (1:1000, Santa Cruz). Band densities were determined using the ImageJ software, and the ratios of the specific/control bands were shown.
ELISA for total or phosphorylated nuclear NF-κB p65 activity
The adherent monocytes from both sham and burn mice were seeded on 6-well plates (2 × 106 per well) and were treated with PBS or consecutive doses of Ex-4 for 2 h. The total and phosphorylated nuclear NF-κB p65 activity were assessed with total NF-κB p65 kit (KHO0371, Invitrogen, CA, USA) and InstantOne phosphoSer536 NF-κB p65 ELISA™ kit (eBioscience, San Diego, CA, USA), respectively. Data were expressed as ng/ml for total NF-κB p65 and OD value for phosphoSer536 NF-κB p65.
All data are expressed as mean ± SEM. Data were analyzed by Student’s t test or one-way ANOVA on SPSS 19.0 software. Differences between groups were considered statistically significant when the probability was less than 5%. Each experiment was performed at least in duplicate for statistical purposes.
Burn monocytes were resistant to immunosuppressive effect of Ex-4 with blunt cAMP production
To determine whether the immunosuppressive effect of Ex-4 would be sustained in a burn-injured model, we examined the immune response of monocyte purified from thermal-injured mice toward Ex-4 treatment ex vivo. Ex-4-inhibited LPS-increased monocyte viability from sham-injured mice in a dose-dependent manner, but not that from burn-injured mice. Although Ex-4 could generally induce IL-10 release from both groups, it could only increase IL-10 release from burn monocytes at high concentration. Unexpectedly, Ex-4 could increase TNF-α secretion in burn monocyte in a dose-dependent manner (Fig. 1a). This finding was conflict with the expected TNF-α-inhibiting effects of these Gαs-coupled agonists because cAMP is widely known to have anti-inflammatory effects on cytokine and chemokine production by activated macrophages. Therefore, we went on to examine the cAMP production in both groups following Ex-4 treatment.
As expected, stimulation of sham monocyte with Ex-4 resulted in a transient increase of intracellular cAMP level in a dose-dependent manner (Fig. 1a). Then, it fell down to the comparative levels as the placebo controls after 6-h stimulation. On the contrary, burn monocyte was resistant to Ex-4 treatment in terms of cAMP production at any dose or time examined (Fig. 1b), indicating the dysfunction of GLP-1R signaling in burn monocyte. In light of the immediate activation of PKA by cAMP, which in turn phosphorylates myriad downstream targets , we went on to detect the downstream PKA activity in both monocytes following Ex-4 treatment. Unexpectedly, Ex-4 could promote PKA activity marginally after 1-h incubation in sham monocytes (P = 0.050, One-way ANOVA), but significantly in burn monoctytes (P = 0.024, One-way ANOVA). In detail, Ex-4 at low dose could promote PKA activity in sham monocytes, but could only do so in burn monocytes at high dose. However, after 3 h incubation, higher dose of Ex-4 could significantly inhibit PKA activity in burn monocyte (P = 0.032), indicating a peripheral protective effect against high dose of GLP-1 in pathophysiological situation. The retarded PKA activity in response to Ex-4 in burn monocytes after 1 h incubation indicated the desensitization of GLP-1R pathway in the context of burn injury.
Burn monocyte was biased toward Giα signaling in response to Ex-4 ex vivo
Recent studies found that β-adrenoceptors (β-AR), a typical GPCR, might elicit proinflammatory effects by switching from Gαs to Gαi signaling in synoviocytes under hypoxia in patients with rheumatoid arthritis  and in monocytes from patients with septic shock . The unexpected increase of TNF-α level by Ex-4 in burn monocyte suggested the possible switch of Gαs to Gαi signaling in burn monocyte. To further substantiate these observations, we used the specific Gαi inhibitor pertussis toxin (PTX) or the specific PKA blocker H89 to block the key effectors of the switch in the downstream of GLP-1R signaling. In sham monocyte, the decreased TNF-α secretion by Ex-4 was rescued by PTX or H89 pretreatment (Fig. 2a), whereas in burn monocytes, the increased TNF-α secretion by Ex-4 was reversed by pretreatment of PTX or H89 (Fig. 2a). However, Ex-4 induction of IL-10 release was potentiated by PTX in burn monocytes, without any effect of H89 (Fig. 2b). One pathway known to downregulate proinflammatory TNF-α production is that elicited by cAMP [31, 32]. To address the role of G-protein signaling in monocytes during the inflammatory response, we further examined the intracellular cAMP levels. Although inactivation of Giα by PTX alone did not influence cAMP level significantly in both groups, it could reverse Ex-4-induced cAMP to extremely lower level than that in placebo controls in both groups. In addition, single H89 treatment could elevate cAMP level in sham monocyte but decrease it in burn monocyte. Nevertheless, Ex-4 could not increase cAMP production in presence of H89 in neither sham monocytes nor burn monocyte after 6 h incubation (Fig. 2c). Considering that cAMP could further activate PKA activity, we wondered if PKA activity was differentially activated between the two groups. As expected, Ex-4 at middle dose could induce PKA activity significantly in sham monocytes after 1-h incubation, but slightly decrease it in burn monocytes which was reversed by PTX and H89, indicating a PKA-dependent Giα signaling in burn monocyte. It was intriguing that TNF-α secretion be inversely correlated with cAMP level in both groups (Fig. 2a, c),but not with PKA activity (Fig. 2a, d),suggesting possibly an alternative PKA-independent pathway in TNF-α secretion.
Imbalanced ratio of Gsα/Giα with reduced GRK2 expression in burn monocytes
To understand whether G-protein-mediated cell signaling is altered in monocyte during burn injury, we investigated the monocytic expression of Gsα, Giα, GLP-1R and GRK2 24 h following burn injury. It is reasonable to hypothesize that the Gsα/Giα ratio has important implications for the regulation of basal adenylate cyclase activity. As reported here, Western blots of monocyte lysates revealed decreased expression of Gsα vs. increased expression of Giα from burn-injured mice as compared with sham-injured mice (Fig. 3). Comparatively, the profound abundance of GLP-1R and GRK2 expressions were both significantly reduced in burns as compared with shams (Fig. 4). Therefore, our results suggest that alterations in GLP-1R signaling in burn injury specifically affect the ex vivo Ex-4-mediated control over monocyte immunity.
Ex-4-induced ERK activation and NF-κB nuclear translocalization in burn monocyte
In addition to the alteration of Gαs/Gαi balance in burn monocyte, another indication of the Gαs to Gαi switch is ERK1/2 activation, which is a proinflammatory signal in most cells . We next sought to determine whether the GLP-1 pathway also mediates the proinflammatory response in burn monocyte. In good correlation with the changes in TNF-α production, pERK1/2 expression was decreased after Ex-4 treatment in sham monocytes, in contrast, pERK1/2 expression was increased after treatment with Ex-4 in burn monocyte (Fig. 5a). The densitometric analysis of Western blot bands (pERK/ERK relationship) confirmed these observations (Fig. 5b).
Since a proinflammatory event through MAPK signaling could increase TNF-α and other cytokines , it might be necessary to evaluate the inflammatory NF-κB translocation and expression in response to Ex-4. As shown in Fig. 6, Ex-4 could marginally reduce NF-κB nuclear translocation in sham monocytes (Fig. 6b, P = 0.056, One-way ANOVA), while showed a trend of increase in burn monocytes after 2-h incubation. Furthermore, phosphorylated nuclear NF-κB p65 was also suppressed by Ex-4 in sham monocytes (Fig. 6b, P = 0.046, One-way ANOVA), however, Ex-4 could not do so in burn monocytes (Fig. 6b, P = 0.351, One-way ANOVA).
This study demonstrates a major finding in GLP-1R signaling in murine monocyte following burn injury. An expected inhibitory effect of Ex-4 via Gαs signaling on TNF-α secretion from monocytes was reversed under burn injury. Since severe burn injury induced inflammatory monocyte subpopulation with inflammasome priming and activation, and an increased infiltration into adipose tissue [34, 35], these findings are most probably more relevant than findings under normal conditions.
Typically, intracellular cAMP has been shown to regulate various cytokines, e.g., increasing anti-inflammatory IL-10, whereas decreasing proinflammatory TNF-α secretions from distinct cell types including monocytes [6, 29]. The elevation of intracellular cAMP has been shown to suppress the proliferation of peritoneal macrophages in guinea pig and bone marrow-derived macrophages in mice . GLP-1R, like many other GPCRs, has been documented to be coupled to numerous G proteins and primarily activates the Gαs subunit to stimulate cAMP production . However, in the present study, such a stimulation of cAMP production was completely absent in burn monocytes with a GLP-1R agonist, Ex-4. Accordingly, the small inhibition on TNF-α secretion was absent in burn monocyte with Ex-4 treatment ex vivo, to make things worse, Ex-4 promoted TNF-α secretion in monocytes from burn-injured mice, and a delayed increase of IL-10 secretion as well. Moreover, the burn monocyte was resistant to the inhibitory effect of Ex-4 in cell viability at all doses examined. Taking together, the results indicated an important alteration in response to Gαs agonist Ex-4 in burn monocyte.
With this information, we hypothesized that the downstream GLP-1R signaling molecules, e.g., Gsα/Giα are responsible for the observed alterations in burn monocyte in response to Ex-4. It was reported that a Gsα/Giα switch occurred in the synoviocytes in hypoxic rheumatoid arthritis , in murine peritoneal macrophages after β-adrenergic stimulation , and in peripheral CD14+ monocytes from septic shock patients , resulting in intensifying inflammatory mediator production. Even in rat ventricular myocytes, similar switch occurred after LPS treatment, resulting in the inotropic effect of adrenomedullin  and acute cardiac failure during sepsis. Considering that Gαi2 transcription was increased by LPS in monocyte [39, 40], or by TNF-α in human airway smooth muscle cells [40, 41], rat cardiomyocytes [22, 42] and liver , the Gαi upregulation in burn monocyte might be attributed to the inflammatory stimuli, such as TNF-α. While the inhibition of Gαi protein by PTX [30, 43, 44, 45, 46] or genetic deletion of Gαi2 protein  could enhance inflammatory response, increased Gαi function or overexpression will suppress TLR-induced inflammation . These changes are recapitulated in our murine model of thermal injury. We found that Gαi blocking by PTX alone could increase TNF-α secretion, or restore the decreased TNF-α by Ex-4 in sham monocyte. On the contrary, pretreatment of PTX could reduce Ex-4-induced TNF-α secretion in burn monocyte. It was reasonable to speculate that GLP-1R signaling be biased to PTX-sensitive Giα in burn monocytes, resulting in either unresponsiveness in cAMP production or an increase in TNF-α secretion.
Nevertheless, the simple imbalance of Gsα/Giα expression might not account for the overall immunological alteration in burn monocyte. The downstream PKA might also be involved in the event. Previous studies have found that PKA-mediated phosphorylation of the β-AR could decrease its affinity for the Gas subunit, whereas increase Gαi binding, so as to switch the predominant coupling of β-AR from Gsα to Giα [47, 48]. In our observation, PKA blocked by H89 could reverse the increased TNF-α secretion by Ex-4 in burn monocyte, suggesting a switch from Gsα to PTX-sensitive, PKA-dependent Giα coupling of GLP-1R signaling in burn monocytes. Accordantly, the unresponsiveness of cAMP toward Ex-4 in burn monocytes was further greatly potentiated by pretreatment of H89 with even lower cAMP level, further indicating a switch from Gαs to Gαi signaling in burn monocyte. By contrast, in sham monocyte, pretreatment with H89 could further facilitate the increased level of cAMP induced by Ex-4 and restore the consequent TNF-α suppression, suggesting the Gαs-cAMP-PKA signaling in sham monocyte.
Intriguingly, we found that low dose of Ex-4 could not induce PKA activity in burn monocyte as in sham monocytes following 1 h incubation. Instead, it could greatly raise its activity at high dose of treatment following 1 h incubation, whereas slightly decrease it after 3 h incubation in burn monocytes. It was noteworthy that the elevated PKA activity in response to Ex-H in burn monocytes was not in concord with blunt cAMP level at the same time points (Fig. 1b, c). Furthermore, the equivalent PKA activity between the two groups (Figs. 1c, 2d) did not translate into similar level of TNF-α secretion. Actually, burn monocyte secreted higher levels of TNF-α than sham monocyte (Fig. 2a). The puzzling data could be explained by recently recognized novel targets directly activated by cAMP including the guanidine nucleotide exchange factor (Epac), a PKA-independent actions of cAMP associated with sustained AKT phosphorylation [49, 50]. Differential roles of Epac-1 vs. PKA in the inhibitory effects of cAMP had been defined in alveolar macrophages, the former for phagocytosis suppression, the later for leukotriene B4 and TNF-α inhibition . In our study, whether Ex-4 at highest dose might activate the alternative pathway of Epac without an increase of PKA activity in sham monocyte should be determined later. Furthermore, the unexpected pronounced PKA activity by high dose of Ex-4 in burn monocyte could be attributed to the possible alteration in upstream of GLP-1R signaling. Instead of via Gαs signaling, agonist-induced GLP-1R internalization is mediated by the Gaq pathway, inducing ERK1/2 phosphorylation . An increase of internalization and a prolonged cycling of ligand-activated GLP-1Rs was observed in HEK293 cells by higher dose of Ex-4 compared with lower dose of Ex-4, which is suggested to be correlated with a prolonged cAMP signal . Therefore, in burn monocyte, higher dose of Ex-4 might also activate PKA with sustained stimulus of low cAMP via Gaq/ ERK1/2 pathway.
Since Ex-4 exerted anti-inflammatory properties in most situations, it is conceivable that Ex-4 could inhibit pERK1/2 and nuclear NF-κB p65 activity as expected in sham monocytes. Nonetheless, it is often asserted that the switch from Gαs to Giα protein-coupled ERK1/2 signaling pathway mediate LPS-induced signaling [30, 46]. Likewise, in burn monocytes, the activation of ERK1/2 was augmented by Ex-4 after 2-h treatment, confirming the proinflammatory ERK1/2 signaling pathway for GLP-1R in burn monocytes. All the above hypothesis was further verified by increased NF-κB translocation and absence of nuclear phosphor-NF-κB p65 inhibition by Ex-4 in burn monocytes. Collectively, these data indicates that the proinflammatory effect of Ex-4 on burn monocyte may account for Gαs to Gαi switch in GLP-1R signaling in splenic monocytes.
Our hypothesis was further corroborated by the following findings of key molecules downstream of GLP-1R signaling. The first intriguing finding was the reliably decreased expression of GRK2 in burn monocytes. GRK2 is a member of a kinase family originally discovered to phosphorylate and desensitize G-protein-coupled receptors. Earlier studies already identified the presence of GRK2 in myeloid cells and its levels are altered in many inflammatory disorders including sepsis  and surgery [30, 43, 46]. GRK2 could negatively regulate NFκB1p105-ERK pathway and limit endotoxemic shock in mice . On the other hand, mice bearing GRK2 deletion in myeloid cells exhibited exaggerated inflammatory cytokine/chemokine production in response to LPS . The reduced expression of GRK2 in burn monocyte indicated its possible role in augmenting inflammatory cytokine/chemokine production in burn monocyte. In fact, burn monocytes secreted higher level of TNF-α, bearing more nuclear NF-κB translocation and p-ERK1/2 expression compared with sham monocytes in our study, reinforcing the inflammatory monocyte in burn injury. The second finding was the internalization and reduced protein level of GLP-1R in burn vs. sham monocyte. Upon its activation, most GPCRs internalize from the cell surface to dampen the biological response. Nevertheless, early sensibilization of GLP-1R is still induced by reduced GRKs, thus, Gαs-stimulated adenylyl cyclase activity is increased. The early rise in cAMP concentration could induce PKA activation as seen in burn monocytes, that showing higher PKA activity in response to high dose of Ex-4 treatment at 1-h incubation. Over the time, cAMP could not be induced any more, thereby resulting in lower PKA activity in response to Ex-H at 3 h incubation in burn monocytes (Fig. 1c). The third finding was that burn injury could lead to increased ratio of Giα/Gsα 24 h after injury (Fig. 3), the expressional imbalance may be in favour of Giα-biased signaling of GLP-1R. All the above molecular changes downstream of GLP-1R could result in less cAMP production and consequent reduced PKA activation in burn monocytes. Moreover, owing to Giα biased binding and activation, the increased MAPK signaling through ERK1/2 could elicit a proinflammatory effect that increases TNF-α secretion via NF-κB. Thereby, the ligand-biased signaling by a single receptor, e.g., GPCR, could help elicit effective immunity against invasive bacteria in mice subjected to burn injury.
Taking together, this study demonstrated that Ex-4 treatment of burn monocytes result in a proinflammatory GLP-1R switching from Gαs to Gαi signaling. These findings might explain the enhanced tissue inflammation in burn mice following Ex-4 therapy. It also explains that typical Gαs-coupled receptor agonist do not exert anti-inflammatory but even proinflammatory effects, e.g., on TNF-α secretion in the context of burn injury. This study also indicates that loss of GRK2 may account for the early sensitization of GPCR including GLP-1R in monocytes to activate PKA, which in turn phosphorylate GPCR and increases Gαi binding for Ex-4.
This study was supported by National Natural Science Foundation (81272089), the National Basic Research Program of China (2012 CB518102), and Twelve-Five Plan for Military Scientific Foundation (BWS12J050).
QHZ: development of the concept, conduction of most experiments, generation of all the data and figures, drafting and final approval the paper. JWH, GLL, XJJ and XDY: conduct part of the experiments. YMY: providing part of the fund and the personnel.
Compliance with ethical standards
Conflict of interest
The authors report no conflict of interest.
- 9.Montrose-Rafizadeh C, Avdonin P, Garant MJ, Rodgers BD, Kole S, Yang H, et al. Pancreatic glucagon-like peptide-1 receptor couples to multiple G proteins and activates mitogen-activated protein kinase pathways in Chinese hamster ovary cells. Endocrinology. 1999;140:1132–40.CrossRefPubMedGoogle Scholar
- 15.Watada S, Yu YM, Fischman AJ, Kurihara T, Shen CA, Tompkins RG, et al. Evaluation of intragastric vs intraperitoneal glucose tolerance tests in the evaluation of insulin resistance in a rodent model of burn injury and glucagon-like polypeptide-1 treatment. J Burn Care Res. 2014;35:e66–72.CrossRefGoogle Scholar
- 44.Solomon KR, Kurt-Jones EA, Saladino RA, Stack AM, Dunn IF, Ferretti M, et al. Heterotrimeric G proteins physically associated with the lipopolysaccharide receptor CD14 modulate both in vivo and in vitro responses to lipopolysaccharide. J Clin Investig. 1998;102:2019–27.CrossRefPubMedPubMedCentralGoogle Scholar