Fetuin-A regulates adipose tissue macrophage content and activation in insulin resistant mice through MCP-1 and iNOS: Involvement of IFNγ-JAK2-STAT1 pathway

In the context of obesity-induced adipose tissue inflammation, migration of macrophages and proinflammatory subtype is considered a pivotal event in the loss of adipose insulin sensitivity. Two major chemoattractants, monocyte chemoattractant protein-1 (MCP-1) and Fetuin A (FetA), have been reported to stimulate macrophage migration into inflamed adipose tissue instigating inflammation. Moreover, FetA could notably modulate macrophage polarization, yet the mechanism(s) is unknown. The present study was undertaken to elucidate the mechanistic pathway involved in the actions of FetA and MCP-1 in obese adipose tissue. We found that FetA knockdown in high fat diet (HFD) fed mice could significantly subdue the augmented MCP-1 expression and reduce adipose tissue macrophage (ATM) content thereby indicating that MCP-1 is being regulated by FetA. Additionally, knockdown of FetA in HFD mice impeded the expression of inducible nitric oxide synthase (iNOS) reverting macrophage activation from mostly proinflammatory to anti-inflammatory state. It was observed that the stimulating effect of FetA on MCP-1 and iNOS was mediated through interferon γ (IFNγ) induced activation of JAK2-STAT1-NOX4 pathway. Furthermore, we detected that the enhanced IFNγ expression was accounted by the stimulatory effect of FetA upon the activities of both cJun and JNK. Taken together, our findings revealed that obesity-induced FetA acts as a master upstream regulator of adipose tissue inflammation by regulating MCP-1 and iNOS expression through JNK-cJun-IFNγ-JAK2-STAT1 signaling pathway. This study opened a new horizon in understanding the regulation of ATM content and activation in conditions of obesity-induced insulin resistance.


ABSTRACT
In the context of obesity-induced adipose tissue inflammation, migration of macrophages and their polarization from predominantly anti-inflammatory to proinflammatory subtype is considered a pivotal event in the loss of adipose insulin sensitivity. Two major chemoattractants, monocyte chemoattractant protein-1 (MCP-1) and Fetuin A (FetA), have been reported to stimulate macrophage migration into inflamed adipose tissue instigating inflammation. Moreover, FetA could notably modulate macrophage polarization, yet the mechanism(s) is unknown. The present study was undertaken to elucidate the mechanistic pathway involved in the actions of Biochemical Journal. This is an Accepted Manuscript. You are encouraged to use the Version of Record that, when published, will replace this version. The most up-to-date-version is available at https://doi.org/10.1042/BCJ20210442

INTRODUCTION
As sentinel cells of the body, macrophages play a pivotal role in host defense mechanisms. Armed with multiple pattern recognition receptors including toll-like receptors (TLRs), C-type lectins (CTLs), nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs) etc., macrophages can appropriately sense danger signals emanating from a variety of pathogens and damaged tissue (1). Not only that, these heterogeneous group of cells exhibit a wide range of plasticity and complex molecular program in coordinating efficient responses to an array of cues arising within the tissue microenvironment (2). The adipose tissue (AT) provides an outstanding case of study that was initiated by the seminal observations made independently by two groups about the accumulation of adipose tissue macrophages (ATM) with increasing adiposity in mice and humans (3,4) followed by reports on genetic or chemical ablation of ATM improving insulin sensitivity (5,6). Extensive research over the years has highlighted ATM as key mediator of inflammation, insulin resistance and AT dysfunction during progression of obesity and type 2 diabetes (7). ATM biology in an obese setting has come of age but nevertheless, it has made us appreciate the amazing heterogeneity, inherent complexities of macrophage activation, the bewildering array of metabolic cues that can potentially regulate ATM profile and importantly, cross-talk with adipocytes much of which is yet to be defined. A couple of recent excellent reviews in the subject have put forth that factors and/or metabolic triggers that initiate ATM activation and induce sterile inflammation within AT remain largely unknown (8,9).
Hypertrophied adipocytes secrete a glut of negative signals that can contribute to the onset of AT inflammation in conditions of excess energy. Monocyte chemoattractant protein-1 (MCP-1), a member of the C-C chemokine family, is a potent chemotactic factor for monocytes/macrophages (10). The augmented levels of adipocyte MCP-1 in obese subjects was reduced upon weight loss (11). Higher MCP-1 expression promoted increased macrophage infiltration into obese AT (12). An elegant experiment based on fluorescently-labeled monocyte tracking technique showed that in conditions of either MCP-1 or its receptor gene KO, the reduction in ATM content was upto 40% only (13)  adipokine that is worth mentioning in this context is Fetuin-A (FetA). It has been reported that FetA acts through TLR4 and induces a strong inflammatory response in hypertrophic adipocytes (14). We noticed that being upregulated in hyperlipidemic condition, FetA served as a potent chemoattractant leading to increased macrophage influx into inflamed AT in a synergistic manner with MCP-1, and also caused ATM polarization from anti-inflammatory M2 to proinflammatory M1 subtype in obese insulin resistant mice (15). The close link between MCP-1 and inflammation was revealed by adipose-specific Mcp-1 knockdown mice displaying lowered plasma concentrations of several proinflammatory adipokines viz. leptin, plasminogen activator inhibitor-1 and resistin compared to wild-type controls (16). We believe that the two chemotactic signals, FetA and MCP-1, copiously generated in lipid-enriched AT milieu, can significantly influence ATM content and activation. Considering the multifarious dimensions of FetA action (17), we hypothesized that FetA might regulate MCP-1 level in AT and the interplay between the duo is a critical determinant of AT inflammation.
The contribution of ATM to the pathogenesis of obesity-induced insulin resistance can be adjudged from the massive increase in ATM number (~25-fold) over normal representing upto one-half of total AT cell population in obese diabetic state (18). This is associated with altered phenotype as well. While the inflammatory ATM is characterized by M1-like classically activated subtype identified as F4/80 + /CD11b + /CD11c + and overproducing TNFα, IL-6, IL-1β, iNOS, C-C chemokine receptor 2 (CCR2), the receptor for MCP-1, and reactive oxygen species (ROS), metabolically healthy ATM present alternatively activated M2 profile with F4/80 + /CD11b + /CD206 + cells expressing Arginase-1, IL-10 and other type 2 effectors (9). In fact, there exists a continuum of phenotypes between the two extremes of proinflammatory M1 and anti-inflammatory M2 in response to diverse stimuli (19) and what tip-offs the balance is actually the decisive factor.
Interferon γ (IFNγ) is known to be the main cytokine responsible for M1 activation and acts through its receptor recruiting Janus kinase (JAK) adaptor proteins leading to dimerization and nuclear translocation of Signal transducers and activators of transcription 1 (STAT1) (20). The present study revealed the key role of FetA as an upstream regulator of MCP-1 and iNOS through the IFNγ-JAK2-STAT1 axis and Adult male Swiss albino mice (20-25 g) were acclimatized at ideal environment (25 ± 2 °C and relative humidity 55 ± 5% with 12 h alternate light and dark cycle) and provided with drinking water and standard diet (SD) ad libitum. Animal handling, acclimatization, the mice were divided into two groups: one group was continuously supplied with Standard Diet (SD) and the other group was provided with High Fat Diet (HFD) for a duration of 3 months as described in detail in our previous reports (14,,21,22). In vivo morpholino oligos specific for FetA was injected dissolved in physiological saline (25 nM) for 5 days in a row to generate FetA knockdown mice model (23). As its corresponding control, a subset of HFD mice was given control Vivo Morpholino (VMO) supplied by the manufacturer (Gene Tools, Philomath, USA).
On the 10 th day following the final injection, the mice were sacrificed. Furthermore, to mimic the effect of FetA in vivo, FetA was dissolved in sterile phosphate-buffered saline (PBS, 0.05M, pH 7.4) and was administered to SD mice at a dose of 0.7 mg/gram body weight for a period of 5 days (22,23). The control subset of mice received same volume of sterile PBS. Additionally, one subset of HFD mice were administrated subcutaneously with JNK inhibitor, SP600125 at a dose of 30 mg/kg body weight as described previously (24,25) for a period of 5 days to observe the effect of JNK in regulating FetA's impact on macrophage migration and polarization.
The mice were euthanized by CO 2 inhalation after completion of the respective treatments.

Isolation of stromal vascular fraction (SVF) from adipose tissue
Mice epididymal and retroperitoneal fat was collected in normal saline, cleaned and homogenized in PBS, pH 7.4 at 50 mM supplemented with 0.5% BSA. The tissue suspensions were centrifuged at 500g for 5 min. and subjected to digestion with 1 mg/ml collagenase at 37°C in a shaking water bath for 30 min. The cell suspension was then filtered through a 100 µm sieve and re-centrifuged at 300g for 5 min to isolate floating adipocytes from the stromal vascular fraction (SVF) pellet which were used for further experimental procedures (26).

Cell culture and treatments
For primary culture, isolated SVF was washed in PBS, suspended in fetal bovine serum (FBS)-free culture media in 6-well plates and kept in 5% CO 2 incubator for 4 h at 37 °C. RAW 264.7 cells were cultured in DMEM supplemented with penicillin (100 U/ml), streptomycin (100 μg/ml) and FBS (10%) at 37°C in a humidified chamber with 5% CO 2 . Isolated SVF and RAW 264.7 cell lines were treated without (Con) or with (50-200 μg/ml) FetA conjugated with 0.25 mM saturated fatty acid, palmitate in serum-free media for specified time period following our previous publication (22).

Blood and serum analyses
At the onset of treatment and throughout the treatment tenure, blood glucose level was monitored regularly with Accu-Chek glucometer (Roche, Basel, Switzerland).
Oral glucose tolerance test (OGTT) was done by measuring blood glucose levels before and after oral glucose gavage at a dose of 1 g/kg body weight. The efficiency of whole-body insulin functioning was evaluated by insulin tolerance test (ITT) injecting insulin at a dose of 0.75 IU/kg body weight. The sera of mice were used for quantification of insulin and MCP-1 levels.

Immunoblotting
The isolated cells were suspended in RIPA lysis buffer (Sigma-Aldrich, St. Louis, MO, USA) and protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA), sonicated on ice and centrifuged at 10,000g for 20 min at 4°C. The supernatant was collected and protein concentration was determined by Lowry's method (27). Equal amount of denatured protein samples were loaded in 7.5% or 10% polyacrylamide Downloaded from http://portlandpress.com/biochemj/article-pdf/doi/10.1042/BCJ20210442/923885/bcj-2021-0442.pdf by guest on 03 November 2021 Biochemical Journal. This is an Accepted Manuscript. You are encouraged to use the Version of Record that, when published, will replace this version. The most up-to-date-version is available at https://doi.org/10.1042/BCJ20210442 gels and then transferred to polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA, USA). Non-fat skimmed milk (5%) was used to block protein free regions on the membrane after which the membrane was dipped in desired primary antibody solution. Following overnight incubation at 4°C, alkaline phosphatasetagged secondary antibody was applied for the detection of protein bands using 5bromro-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT) as substrate of the tagged enzyme. ImageJ software was used to quantify the intensity of protein bands on the membrane.

Quantitative RT-PCR
Initially, RNA was isolated from the cells using TRI reagent (Sigma-Aldrich, St. Louis, MO, USA) and then cDNA was synthesized with the help of Revert Aid First-Strand cDNA Synthesis kit (Thermo Scientific, MA, USA). Customized primers listed in Table 1 were used for SYBR Green-based amplification of desired genes in real-time thermal cycler (Applied Biosystems, Foster City, CA, USA). The value of gene expression was quantified by measuring the fluorescence of double stranded DNA binding dye SYBR Green with respect to the reference gene EF1α. Melting curve analysis was done to ensure the specificity of product.

Immunohistochemistry
Paraformaldehyde fixed-paraffin embedded adipose tissue sections of SD, HFD, and HFD+FetA KD mice were incubated firstly with CD11c primary antibody followed by its corresponding horseradish peroxidase (HRP)-tagged secondary antibody for detecting the intensity of M1 macrophage marker.  Software). Similar procedure was followed for analysis of FITC-tagged CD86 and CD206 markers in SVF primary culture incubated with FetA in presence or absence of JNK inhibitor SP600125. In addition to M1 and M2 macrophage specific markers, all treatment groups were treated with rhodamine-conjugated antibody against F4/80 which is a common marker of macrophage.

Cell migration assay
Cell migration studies were undertaken in SVF isolated from SD, HFD, HFD+FetA KD and HFD+SP600125 mice using QCM 24-well colorimetric cell migration assay kit in an ELISA plate reader at 560 nm wavelength following the manufacturer's protocol.

Statistical analyses
The data were subjected to Student's t-test (for single-factorial designs) and one-way ANOVA (for multi-factorial designs). The means were compared by performing Tukey's post hoc multiple range test using Prism version 8 for Windows (GraphPad Software, Inc.). All values were expressed as means ± standard error of the mean (SEM) and minimal statistical significance was considered at P < 0.05.

FetA silencing reduces MCP-1 and iNOS, and lowers M1/M2 ratio within inflamed adipose tissue
To discern the effect of FetA on macrophage migration, we knocked down FetA in high fat diet (HFD) mice using vivo morpholino. The efficiency of FetA knockdown was assessed by immunoblotting followed by densitometry of the protein expression of FetA in lysates of hepatocytes, adipocytes and SVF isolated from SD, HFD and HFD+FetA KD mice ( Supplementary Fig. S1). No significant change in body weight and/or fat mass between HFD mice and HFD+FetA KD mice was noticeable (data not shown). However, it was interesting to note that HFD+FetA KD mice displayed much reduced M1 macrophage load compared to that of HFD mice as evident from immunohistochemistry with anti-CD11c (Fig. 1A). Further, immunofluorescence of adipose tissue (AT) clearly revealed that knockdown of FetA in HFD mice drastically reduced the M1 surface marker CD86 but augmented the M2 marker CD206 ( chemokines known to regulate macrophage migration is monocyte chemoattractant protein-1 (MCP-1) (10), we checked the effect of FetA knockdown upon MCP-1. It was striking to note that serum MCP-1 protein expression as well as concentration were markedly lowered in HFD+FetA KD mice (Fig. 1D,E). These data provided an insight that FetA effect upon ATM is possibly by regulating MCP-1. To ascertain FetA's direct involvement in this event, we injected FetA in SD mice and examined the level of MCP-1 by ELISA. Fig. 1F demonstrated that FetA injection elevated MCP-1 level in SD mice while knockdown of FetA considerably reduced MCP-1 in SVF lysate. This was accompanied by notable rise in levels of proinflammatory cytokines TNFα and IL-6 in FetA-injected SD mice as well as in HFD mice whereas decreased cytokine values were recorded in HFD+FetA KD condition (Fig. 1G,H). On the contrary, the concentration of IL-10, the key anti-inflammatory cytokine followed the opposite trend (Fig. 1I).

FetA tightly regulates IFNγ expression and FetA action is obliterated in IFNγ-silenced macrophages
Administration of high fat diet to mice resulted in robust expression of IFNγ mRNA and protein in SVF; however, it was interesting to note that FetA knockdown could lower IFNγ expression at both the levels of transcription and translation ( Fig. 2A,B).
Presence of this type II interferon can induce MCP-1 expression (28,29). It is also known to be a potent stimulator of the inducible isoform of nitric oxide synthase (iNOS) (30,31). Intriguingly, at a concentration of 50 and 100 µg/ml in vitro, FetA upregulated IFNγ, MCP-1 and iNOS proteins in SVF in a dose-dependent manner (Fig. 2C). Not only for protein, it held true for gene expression as well. FetA (100 µg/ml) was shown to enhance significantly MCP-1 and iNOS transcripts in SVF (Fig.   2D). We performed time kinetics experiments after incubating SVF with 100 µg/ml FetA for 2 h, 4 h and 6 h, respectively. Maximal effect of FetA upon IFNγ, MCP-1 and iNOS expression was detected after 4 h (Fig. 2E). This was found to correlate strongly with dose-dependent and time-dependent responses of MCP-1 concentration to FetA (Fig. 2F,G). In order to clearly delineate the involvement of IFNγ, we silenced IFNγ in murine macrophage RAW 264.7 cell line by using siRNA that was confirmed by immunoblotting followed by densitometry and significantly  Figure S2). Our hypothesis was validated as the raised MCP-1 and iNOS protein levels upon FetA treatment were significantly downregulated when IFNγ was silenced (Fig. 2H). This was reflected in FACS analyses showing obliterated FetA action in IFNγ-silenced cells having reduced CD11c + cell population, a potent M1 marker (Fig. 2I) while Arginase 1 + cell population, an indicator of M2 phenotype was restored (Fig. 2J).

Proinflammatory stimulus of FetA is mediated through IFNγ-JAK2-STAT1 pathway
To get a deeper understanding of the IFNγ signaling pathway, we looked into the Janus kinase (JAK)/Signal transducers and activators of transcription (STAT) pathway which is activated by IFNγ (32). This pathway is also shown to regulate NADPH oxidase 4 (NOX4) (33). Administration of FetA to RAW 264.7 cells markedly enhanced the expression of pJAK2, pSTAT1 and NOX4 proteins (Fig. 3A). However, when given to IFNγ-silenced cells, FetA effect was markedly subdued (Fig. 3B). In order to validate STAT1's direct involvement in the pathway, we treated SVF with a potent inhibitor of STAT1, fludarabine (34,35). In presence of fludarabine, FetA failed to upregulate the protein expression of NOX4 and iNOS but not of IFNγ (Fig. 3C).
Inhibition of STAT1 also suppressed MCP-1 level even in presence of FetA (Fig.   3D). Not only this, the proinflammatory stimulus of FetA was significantly inhibited when fludarabine was administered as shown by lowered TNFα, IL-6 and IL-1β levels ( Fig. 3E-G) while anti-inflammatory IL-10 concentration was recovered to a significant extent (Fig. 3H). In a similar vein, inactivating STAT1 resulted in suppression of CD86 and CD11c M1 markers (Fig. 3I) while restoring CD206 and Arginase 1 M2 markers in SVF even in presence of FetA (Fig. 3J). Additionally, FetA failed to augment MCP-1 level when NOX4 was silenced in RAW 264.7 cells (Fig.   3K).

Inhibition of JNK attenuates FetA action and reverts ATM polarization
It was fascinating to observe that in primary culture of SVF isolated from SD mice AT, treatment of FetA in vitro could trigger JNK and cJun activation (Fig. 4A).
Furthermore, we treated SVF isolated from SD mice with a strong inhibitor of JNK, SP600125 (24,25) and noticed that protein expression of all the downstream molecules pcJun, IFNγ, pJAK2, pSTAT1, iNOS and NOX4 were strikingly reduced even in presence of FetA (Fig. 4B) that was reflected in MCP-1 level as well (Fig.   4C). These results clearly depict FetA action being governed by JNK. Inhibition of JNK lowered FetA-induced CD11c while elevating Arginase 1 expression (Fig. 4D).
The polarity shift was evident from FACS data displaying lesser M1 phenotype marker CD86 + cells and increased M2 marker CD206 + cell population in presence of SP600125 (Fig. 4E,F).

HFD mice record similar IFNγ signaling, macrophage migration, cytokine and marker profiles in conditions of either FetA knockdown or JNK inhibition
To substantiate that FetA is regulating macrophage polarization through JNK-cJun-IFNγ-JAK2-STAT1 signaling cascade, we checked the total pathway in vivo by treating HFD mice with JNK inhibitor, SP600125 and in another subset of HFD mice, knockdown of FetA was accomplished. It was captivating to observe that conditions of either FetA knockdown or JNK inactivation in vivo elicited somewhat similar regulation of the pathway in HFD mice (Fig. 5A). Similar trend was also observed for the downstream MCP-1 (Fig. 5B). Next, we tried to analyze the cell-to-cell communication by using Boyden chamber system where incubation media of SVF isolated from SD, HFD, HFD+FetA KD and HFD+SP600125 mice were kept at the bottom and RAW 264.7 cells were kept in the upper chamber. This incubation was carried out for 4 h following which the upper chamber was taken under bright field microscope for photography. Fig. 5C-D demonstrated that macrophage migration and staining intensity was least in HFD mice with FetA knockdown and less in JNK inhibited-HFD mice compared to their only HFD fed counterparts. Additionally, both these conditions could significantly revert the cytokine profile of HFD mice (Fig. 5E-H) and the expression of relevant molecular markers (Fig. 5I).

FetA knockdown and JNK inhibition revives insulin sensitivity in HFD mice
As both FetA knockdown and JNK inhibition effected inflammatory status in AT, we then sought to test whole-body insulin sensitivity by performing OGTT and ITT. Significant revival of insulin sensitivity was possible in HFD+FetA KD mice that was followed by JNK inhibition in HFD mice (Fig. 6A,B). Serum insulin level also revealed the same pattern (6C). Taken together, the data depicted that both FetA knockdown and JNK inhibition could restore insulin sensitivity and asserted that FetA action in governing AT inflammation is mediated through JNK.

TLR4 suppression significantly blocks the proinflammatory effects of FetA
We next concentrated on the known signaling of FetA involving TLR4; in obese mice, FetA acts as an endogenous ligand for TLR4 and through TLR4-mediated pathway mounts strong inflammatory responses in obese HFD mice (14). TLR4-silenced RAW 264.7 cell line was utilized for this purpose and the extent of TLR4 knockdown was validated (Fig. 7A). The data clearly revealed that in TLR4-knockdown RAW 264.7 macrophages, FetA could not stimulate the inflammatory pathway controlling macrophage migration and polarization as evident from expression of different connecting signaling molecules (Fig. 7B). MCP-1 level was significantly decreased in TLR4-silenced cells even in presence of FetA (Fig. 7C). The involvement of TLR4 observed in vitro was substantiated in adipose-derived SVF prepared from SD mice. TLR4 activity was inhibited by CLI-095 and primary culture of these cells displayed low abundance of key inflammatory molecules namely, pJNK, IFNγ, pJAK2, pSTAT1, iNOS and MCP-1 even in presence of FetA as revealed by western blot analyses (Fig. 7D). Moreover, in CLI-095-treated cells even in presence of FetA, MCP-1 level was decreased (Fig. 7E). These results altogether imply the crucial involvement of TLR4 in FetA-mediated AT inflammation.

DISCUSSION
The present study was performed in stromal vascular fraction (SVF) isolated from adipose tissue (AT). A heterogeneous population of varied cells viz. mesenchymal stem cells, preadipocytes, fibroblasts, endothelial cells and immune cells comprise AT-derived SVF wherein macrophages are the highly dynamic and dominant cell type (18). Additionally, we have used the murine macrophage cell line RAW 264.7.
The role of the hepatoadipokine Fetuin-A (FetA) in regulating a plethora of events during inflammation is well recognized (36,17). Based on the premise that adipocytederived FetA could promote macrophage migration into inflamed AT and cause polarity shift of ATM from predominantly M2-like to M1-like phenotype (15), the present study attempts to define the mechanism by which FetA could possibly exert this effect.
In conditions of obesity, the accumulation and polarization of macrophages within lipid-enriched AT is instrumental in generating an inflammatory milieu (3,4,37,38).
Inflamed AT copiously secretes monocyte chemoattractant protein-1 (MCP-1) the deficiency of which not just reduced macrophage infiltration into AT but consequentially improved insulin sensitivity (12). In contrast, overexpression of adipose MCP-1 was reported to exacerbate ATM recruitment thereby worsening insulin resistance (39). In our preliminary experiments, both immunoblotting and ELISA confirmed the presence of significantly higher MCP-1 levels in mice fed HFD which was expected; however, strikingly enough, MCP-1 protein was substantially Macrophages display an astonishing range of plasticity in responding to various endogenous cues arising within the tissue microenvironment (2). Compared to normal lean physiology, the dynamics of ATM in obesity is radically altered under the influence of a host of cytokines and chemokines as it becomes increasingly skewed towards the classically activated proinflammatory M1-like phenotype. A chief ATderived inflammatory factor implicated in metabolic abnormalities inherent in obesity is the type II interferon, IFNγ (40). It provides the primary stimulus for M1 activation inducing multiple gene transcripts including MCP-1 (28,29), NADPH oxidases (NOX) Downloaded from http://portlandpress.com/biochemj/article-pdf/doi/10.1042/BCJ20210442/923885/bcj-2021-0442.pdf by guest on 03 November 2021 Biochemical Journal. This is an Accepted Manuscript. You are encouraged to use the Version of Record that, when published, will replace this version. The most up-to-date-version is available at https://doi.org/10.1042/BCJ20210442 and inducible nitric oxide synthase (iNOS) (41). A number of studies have shown potent induction of MCP-1 at gene and protein levels in response to IFNγ stimulation in several cell types (42)(43)(44)(45)(46). Earlier, preponderance of M2-like ATM was observed in SVF isolated from obese IFNγ-KO mice with modestly attenuated systemic inflammation relative to obese wild-type control littermates (47). On a similar note, we found that silencing IFNγ not only downregulated MCP-1 and iNOS, but shifted macrophage polarity from M1 to M2 even in the presence of FetA which is known to cause M2 to M1 phenotypic switch in obesity (15). The present study revealed the Furthermore, FetA also seemed to regulate macrophage polarization through IFNγ-JAK2-STAT1-iNOS axis. A signature M1 marker iNOS is responsible for selectively regulating M1 macrophage gene expression and dedifferentiation (54).
In our pursuit of the probable mechanism by which FetA could be regulating IFNγ, we considered cJun because it has been shown previously that cJun binds to the proximal promoter of IFNγ gene and enhances its expression (55). JNK is reported to be responsible for stimulating the activity of cJun (56). This was confirmed by the fact that FetA treatment could potentially exacerbate JNK and cJun activation; suppression of JNK by its inhibitor SP600125 eventually blocked FetA's effect.
Several reports have demonrated that disrupting JNK1 function, which is activated in HFD mice, restores insulin sensitivity by reducing inflammation (57)(58)(59). We treated HFD mice with JNK inhibitor, SP600125 and interestingly enough, noted that it in FetA-null mice and how it protects against insulin resistance associated with aging (60,61). There is evidence of partial protection from HFD-induced insulin resistance in Tlr4-knockout mice that was ascribed to downregulated inflammatory gene expression in liver and fat tissue (62). Since FetA is known to activate TLR4 signaling thereby generating inflammatory responses in hyperlipidemic condition (14), we conducted experiments to look into this aspect. When TLR4 was suppressed in RAW 264.7 cells or adipose-derived SVF, the pJNK, IFNγ, pJAK2, iNOS and MCP-1 levels were markedly inhibited even in the presence of FetA indicating that FetA is working through TLR4 in regulating macrophage migration as well as polarization to induce AT inflammation.
In summary, our data ascertained FetA as an upstream master regulator of ATM content and activation; FetA is tightly regulating MCP-1 and iNOS through IFNγ. This mechanistic insight attains credence in the light of recent evidences that complete disruption of IFNγ signaling restored insulin sensitivity and metabolic homeostasis in obese mice (51). Future endeavors towards targeting FetA might present a promising therapeutic option to resolve lipid-induced AT inflammation and thereby improve insulin sensitivity.
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.