Targeting soluble tumor necrosis factor as a potential intervention to lower risk for late-onset Alzheimer’s disease associated with obesity, metabolic syndrome, and type 2 diabetes
Insulin impairment and inflammation are two features common to type 2 diabetes and Alzheimer’s disease; however, the molecular and signaling interactions underlying this relationship are not well understood. Mounting evidence point to the associations between the disruption of metabolite processing in insulin impairment and neurodegenerative conditions such as Alzheimer’s. Although the brain depends partially on metabolites processed in the periphery, to date, little is known about how soluble tumor necrosis factor signaling (solTNF) impacts integrated peripheral immune and metabolic feedback signals in states of energy overload and insulin insensitivity.
C57Bl/6J mice were fed a high-fat high-carbohydrate diet (HFHC) for 14 weeks. The brain-permeant biologic XPro1595® was used to block solTNF-dependent pathways. Metabolic and immune alterations were evaluated in the gut, liver, and brain. Behavioral tests were performed. Untargeted metabolomics was carried out in the plasma and liver.
HFHC diet promotes central insulin impairment and dysregulation of immune-modulatory gene expressed in the brain. Alteration of metabolites associated with type 2 diabetes and Alzheimer’s such as butanoate, glutamate, biopterin, branched-chain amino acids, purines, and proteoglycan metabolism was observed in HFHC-fed mice. solTNF inhibition ameliorates hepatic metabolic disturbances and hepatic and intestinal lipocalin-2 levels, and decreases insulin impairment in the brain and behavioral deficits associated with HFHC diet.
Our novel findings suggest that HFHC diet impacts central insulin signaling and immune-metabolic interactions in a solTNF-dependent manner to increase the risk for neurodegenerative conditions. Our novel findings indicate that selective solTNF neutralization can ameliorate peripheral and central diet-induced insulin impairment and identify lipocalin-2 as a potential target for therapeutic intervention to target inflammation and insulin disturbances in obesogenic environments. Collectively, our findings identify solTNF as a potential target for therapeutic intervention in inflammatory states and insulin disturbances in obesogenic environments to lower risk for AD.
KeywordsInsulin Soluble tumor necrosis factor Metabolic inflammation Neuroactive metabolites Lipocalin-2 Purines Proteoglycans Gut Liver Metabolomics
Phospho-protein kinase B
Chemokine (C-C motif) ligand 20
Chemokine (C-C motif) receptor 4
Chemokine (C-C motif) receptor 7
Chemokine (C-X-C motif) ligand 11
Chemokine (C-X-C motif) ligand 9
Glyceraldehyde 3-phosphate dehydrogenase
High-fat high-carbohydrate diet
Interleukin 1 beta
Interleukin 1 receptor antagonist
Interleukin 23, alpha subunit p19
Insulin receptor substrate 1
Insulin receptor beta Tyr1150/1151
Phospho-protein kinase B serine 473
Hospho-protein kinase B threonine 308
IRS1 phosphorylation at tyrosine 608
IRS1 phosphorylation at serine 307
Peroxisome proliferator-activated receptor alpha
Suppressor of cytokine signaling 3
Soluble tumor necrosis factor
Sterol regulatory element-binding protein 1c
Type 2 diabetes
Toll-interleukin 1 receptor (TIR) domain-containing adaptor protein
Toll-like receptor 2
Toll-like receptor 4
Transmembrane tumor necrosis factor
Tumor necrosis factor
Tumor necrosis factor (ligand) superfamily, member 14
- ZO-1 (TJP1)
Tight junction protein 1
Insulin resistance (IR) affects an increasingly large population globally, and despite decades of intense research efforts, type 2 diabetes (T2D) remains an important public health problem throughout the world . Central IR disrupts memory and cognition and promotes disturbance in metabolic and inflammatory responses . Although IR and systemic inflammation have been identified as risk factors for Alzheimer’s disease (AD), the molecular and signaling mechanism underlying this relationship are not well understood. In the states of over-nutrition, intestinal microbiota-derived products and circulating food metabolites from intestinal and liver interactions can disrupt the regulation of insulin activities and immune balance.
A wealth of the literature suggests that elevated tumor necrosis factor (TNF) exerts central and peripheral effects on metabolic and immune pathways and contributes to IR and AD [3, 4, 5, 6, 7]. TNF is upregulated in the presence of obesity and impacts the expression of other multiple inflammatory factors such as IL-6 and LCN2 that promote, exacerbate, and sustain chronic systemic inflammation and insulin impairment [6, 8]. TNF is synthesized as a type I transmembrane protein (tmTNF) that is biologically active in innate immune defense against infection and in myelination . Once cleaved to a soluble (solTNF) form, it is able to mediate inflammatory processes [9, 10]. The overlap and synergistic effects of TNF on metabolic pathways can impact insulin sensitivity and diabetes comorbidities [4, 5].
Additionally, this pleiotropic cytokine is implicated in the central and hepatic interactions that control glucose metabolism and insulin functions [11, 12]. Therefore, maladaptive processes implicated in hepatic steatosis and liver inflammation are associated with the central effects of TNF on insulin dysregulation [3, 13]. In addition to its tissue direct effects on IR, TNF regulates pro-inflammatory markers, such as IL-6 and lipocalin-2 (LCN2) that are implicated in the pathogenesis of hepatic steatosis and T2D onset and progression [14, 15, 16]. Centrally, LCN2 promotes chemokine production in the brain in response to inflammatory insults and regulates glial cells activity and neuroinflammatory and neurodegenerative processes . Some immune-related effects of LCN2 include its deleterious effects in aging insulin insensitivity .
Despite the evidence linking anti-TNF strategies with the improvement of insulin sensitivity, it is not known how the selective neutralization of solTNF signaling can impact the deleterious metabolic-immune interactions present in obesity that affects IR [17, 18]. The hypothesis tested here is that solTNF drives metabolic and inflammatory changes in the gut-liver axis that contributes to insulin impairment and the systemic metabolic dysregulation that increases the risk for AD. We previously demonstrated that the brain-permeant dominant-negative solTNF-selective inhibitor XPro1595® decreases beta-amyloid plaque load in the 5XFAD animal model of AD-like pathology . In the current study, we used this agent to assess the effect of inhibition of central and peripheral solTNF activity in diet-induced insulin impairment. This biologic inhibits solTNF signaling by forming inactive heterotrimers with native solTNF to sequester it away from interacting with TNF receptors . Therefore, this approach leaves the host defense and the neuroprotective transmembrane TNF signal intact [7, 21]. Amyloid deposition was not assessed in our animal model because our previous studies and other groups demonstrate that C57BL/6J mice may present amyloid aggregates in older age (15 months or more) as a result of the aging process, and HFHC diet effects on amyloid deposition in our non-transgenic mice were not anticipated . The experimental design of this study explores a more physiologic approach to assess the metabolic and immune risks for AD associated with an obesogenic environment.
Materials and methods
Animals and diet intervention
C57Bl/6 male mice (n = 50, 6 weeks old, The Jackson Laboratory, Bar Harbor) were singly housed in a colony room (22–23 °C with a 12/12-h light-dark cycle). After 7 days of acclimation, mice received drinking water and standard chow diet (4% fat diet 7001, Envigo) or high-fat high-carbohydrate diet (HFHC) (42% kcal from fat, TD.88137, Envigo) plus 30% (w/v) fructose solution (F012, Sigma-Aldrich) available ad libitum for 14 weeks. Food and drink consumptions were measured twice a week. Mice were weighed weekly (Protocol # DAR-2003358-ENTRPR-N).
Soluble TNF neutralization
Multiplexed immunoassays and metabolic measurements
The trunk blood was collected in EDTA and centrifuged at 2000g for 15 min at 4 °C. The plasma was collected and stored at − 80 °C until measurements for LCN2 (Lipocalin-2/NGAL Quantikine ELISA Kit, R&D Systems), IL-1, IL-6 (Mouse Proinflammatory 7-Plex Ultra-Sensitive Kit, Meso Scale Discovery immunoassays), triglycerides ELISA (Charles River Lab), total cholesterol (Cholesterol Quantitation Kit, Sigma-Aldrich), and insulin and leptin (Mouse Metabolic Kit (Multi-spot Assay System, Meso Scale Discovery). Plates were processed in a SECTOR® Imager 6000 plate reader (Meso Scale Diagnostics, LLC). Data acquired using the Discovery Workbench software (v4.0; Meso Scale Diagnostics, LLC).
To gain further insight into the resultant metabolic and immune alterations present in the adipose tissue and in the gut-liver axis in diet-induced insulin impairment, we next assessed the expression of metabolites in plasma and liver tissue. Plasma and liver tissue samples were analyzed for untargeted metabolomics using established methods . For the analysis of plasma, 50 μL was treated with 100 μL of acetonitrile to precipitate proteins, and triplicate 10-μL aliquots were analyzed by C18 (Accucore, 100 mm × 2.1 mm, 2.6 μm; Thermo Scientific) chromatography with acetonitrile/formic acid gradient interfaced to a Q-Exactive HF high-resolution mass spectrometer (Thermo Scientific) operated in a positive and negative electrospray ionization (ESI) mode. Polar fractions collected from liver samples were analyzed using HILIC chromatography for comparison of tissue metabolites to plasma results. Mass spectral signals were detected using a Thermo Fusion high-resolution mass spectrometer operated at a resolution of 120,000 and scan range 85–1250. Raw data for both plasma and tissue metabolomics analysis was processed using apLCMS with modifications by xMSanalyzer, which provided a feature table of uniquely detected features consisting of mass-to-charge ratio (m/z), retention time, and peak intensity. Prior to statistical analysis, replicate injections were averaged, log2 transformed, and filtered to remove features with greater than 20% missing values. Following statistical analysis, altered pathways were identified using Mummichog and all m/z features meeting the false discovery rate (FDR) threshold.
Immunoblot analyses were performed as previously described . Flash-frozen samples were stored at − 80 °C until processing. Protein was isolated from the hypothalamus, PFC, and liver samples with RIPA buffer (1% Triton-X 100, 50 mM Tris HCL, 0.1% sodium dodecyl sulfate, 150 mM NaCL, pH 8.0). Intestinal proteins were isolated using TRIzol (Life Technologies #15596-018). RIPA samples were centrifuged at 12,000 rpm for 20 min at 4 °C. The supernatant was transferred to a new tube for bicinchoninic acid protein assay (Pierce Scientific #23225). TRIzol samples were resuspended in 1% SDS. Samples were diluted to 1 μg/μL in 4× sample buffer (BioRad #1610747) and boiled at 90 °C for 5 min). After BCA analysis, the membranes were probed overnight with ZO-1, OCLN, CLDN2, IRS-1, p-IRSTyr608, p-IRβTyr1150/1151, p-IRSSer307, p-AktSer473, p-AktThr308, AKT, or anti-β-actin primary antibodies (Additional file 4: Table S2). The membranes were exposed to species-appropriate horseradish peroxidase (HRP)-conjugated secondary antibody (1:1000). Bands were visualized by chemiluminescence, and protein band optical intensity was measured using densitometric analysis (Image Studio Lite). Values were normalized relative to β-actin levels from the same sample. The densities of the phosphorylated protein bands were measured relative to the targeted total protein levels.
RNA was isolated from the colon, small intestine, liver, hypothalamus, and hippocampus as published previously . Samples were homogenized in TRIzol reagent (Life Technologies). RNA was isolated using the RNeasy mini kit (QIAGEN), and reverse transcription of RNA was performed using SABiosciences RT2 First Strand Kit. qPCR was performed using an ABI Prism 7900HT Fast Detection System (Applied Biosystems). Primers were designed using Primer-Blast (www.ncbi.nlm.nih.gov/tools/ primer-blast/). qPCR was performed as published previously . Relative gene expression was measured by validated primers (Additional file 5: Table S3) for Lipocalin-2, tight junction protein 1, Occludin, Claudin-2, tumor necrosis factor, interleukin 1 beta, interleukin 6, Toll-like receptor 2, Toll-like receptor 4, suppressor of cytokine signaling 3, peroxisome proliferator-activated receptor alpha, and sterol regulatory element-binding protein-1c (Integrated DNA Technologies). Transcript abundance was quantified using the 2−ΔΔCt method.
RT2 PCR array
Because sustained central inflammation is considered a risk factor for neuroinflammation and neurodegenerative conditions such as AD , a RT2 PCR profiler was used to investigate the effects of HFHC consumption and solTNF signaling in the hippocampal gene expression. The hippocampus is a brain structure associated with a cognition that is sensitive to high-fat diet-induced insulin resistance . Hippocampal tissue was processed using Qiagen RNeasy mini kit as described previously . After mixing the cDNA template with the appropriate PCR master mix, reverse transcription of RNA was carried out using SABiosciences RT2 First Strand Kit and qPCR was performed using an ABI Prism 7900HT Fast Detection System (Applied Biosystems). Reactions were performed in the 384-well format mouse inflammatory response and Receptors RT2 Profiler PCR Array (PAMM-077Z, SABiosciences). Data analysis was based on the ΔΔCT method with normalization of the raw data to either housekeeping genes.
A three-chambered sociability apparatus (acrylic, 60 × 40 × 22 cm) (UGO BASILE s.r.l.) was used to assess social interaction as previously described . The total duration of active contact made by the tested mouse was recorded in an area of 3 cm around the mesh empty cup or a cup containing a novel mouse. EthoVision XT (Noldus) was used for behavior analysis. Preference for novel mouse was calculated as [(time spent exploring novel mouse)/(total time spent exploring empty cup and novel mouse)] × 100. Preference for novel object was calculated as [(time spent exploring empty cup)/(total time spent exploring empty cup and novel mouse)] × 100.
Marble burying test
A marble burying test was conducted as previously described  to determine whether HFHC diet and solTNF signaling impact anxiety-like behavior. Mice were placed in a plastic tub (50.5 × 39.4 × 19.7 cm) containing 5 in. of lightly pressed bedding. On top of the bedding, 20 marbles of uniform size and color were placed in 5 rows of 4 marbles each. Mice were placed in the containers and allowed to roam freely for 30 min. At the end of testing, the mice were placed back in home cages, and the number of marbles buried at least two thirds of their height was considered buried.
Open field test
In the open field test, a mouse that spends less time in or hesitates to re-enter the open center of the testing chamber is considered to be exhibiting anxiety-like behavior . During the light phase of the light/dark cycle, mice were placed into the open field (45 cm × 45 cm square box) and allowed to move freely for 20 min. Distance, velocity, center, and border statistics were measured using the Noldus/Ethovision software. Center was defined as the central 22.5 cm × 22.5 cm.
Stressful behavioral tests such as fear conditioning and Morris water maze were avoided in this study design because our previous study demonstrates the effects of stress on diet-induced insulin resistance .
Next, because hepatic lipid accumulation leads to insulin impairment and insulin insensitivity, we assessed how HFHC diet consumption and solTNF signaling impact ectopic lipid deposition in liver tissue  Liver tissue from the left lobe was fixed in 4% paraformaldehyde/PBS and cryoprotected in 30% sucrose solution. Tissue was frozen in OCT, sectioned (10 μm), and stained with Oil Red O (150678, Abcam) according to the manufacturer’s instructions. Images were obtained using a Nikon Eclipse 90i microscope with a DS-Fi1 (Nikon) camera and Nikon NIS-Elements AR 3.10 software, magnification × 40.
Data are represented as the mean ± standard error of the mean (SEM). For statistical comparisons between the groups, two-way ANOVA followed by Turkey’s post hoc test was used where applicable. Metabolomic features were tested for differential expression using a Benjamini-Hochberg false discovery rate threshold of 5%; for all other analyses, 0.05 P value threshold was considered statistically significant. The association between variables was analyzed using the Pearson’s correlation coefficient (r). Analyses were performed using GraphPad Prism 6 except where otherwise specified. HRM profiling data were analyzed using R . Metabolites associated with HFHC diet, XPro treatment, and HFHC diet:XPro interations were evaluated using linear models for microarray data (LIMMA) based on two-way ANOVA analysis , as implemented in xmsPANDA . Using this approach, metabolites associated with HFHC were identified by comparison of all HFHC-fed mice and control diet mice, irrespective of XPro treatment; metabolites associated with XPro treatment were identified by comparison of all treated mice to saline-fed control; the influence of XPro on diet-associated metabolic changes was evaluated through post hoc tests for each metabolite. Pearson’s correlation coefficient (r) was used to analyze the associations between variables. PCR array data were analyzed using the RT2 Profiler TM PCR Array Data Analysis software on the SABiosciences website http://www.sabiosciences.com/pcrarraydataanalysis.php and are expressed as a fold regulation change.
solTNF neutralization decreases insulinemia in diet-induced metabolic inflammation
Soluble TNF neutralization reverts increased insulin plasma levels in presence of diet-induced metabolic inflammation
0.91 ± 0.13 a
1.89 ± 0.38 a
6.97 ± 1.63 b
2.76 ± 0.62 a
4.80 ± 0.64 a
6.90 ± 1.32 a
84.46 ± 24.84 b
67.45 ± 16.10 b
88.88 ± 22.79 a
79.00 ± 5.23 a
234.80 ± 35.87 b
183.81 ± 24.91 b
38.53 ± 7.41 ab
57.35 ± 6.70 a
31.49 ± 4.65 b
29.77 ± 2.77 b
1.80 ± 0.57 a
1.94 ± 0.41 a
7.82 ± 1.48 b
5.22 ± 0.93 ab
56.01 ± 7.03 ac
39.11 ± 2.53 a
80.92 ± 6.75 b
72.18 ± 6.36 bc
HFHC diet and solTNF signaling impact neuroactive metabolites processing
An untargeted plasma and liver metabolomics was used to investigate the effect of HFHC diet on metabolic profiles and the extent to which solTNF neutralization reversed any of those alterations. The results for all significant metabolites and tissues are provided in Additional files 2, 4, and 5: Tables S4, S5, and S6. HFHC diet promoted significant changes in the plasma metabolic profile, with 329 m/z features differentially expressed between the control and HFHC diet (Fig. 1b). Metabolic pathway enrichment identified 20 altered pathways associated with HFHC consistent with proteoglycans, amino acids, lipids and inflammatory lipid pathways, β-oxidation, oxidative stress, and pentose/glucuronate metabolism (Fig. 1c). Comparison of anti-solTNF and saline-treated mice detected 44 m/z features associated with solTNF neutralization. Pathway enrichment identified 3 altered pathways, including prostaglandin formation from arachidonate, glycerophospholipid metabolism, and C21-steroid hormone biosynthesis and metabolism. Thirty-nine m/z features showed an interaction between HFHC diet and solTNF neutralization. Annotated metabolites included oxidized fatty acids, sterols, bilirubin, and chondroitin 4-sulfate, a metabolite related to proteoglycan synthesis. No specific pathways were associated with the interactions between HFHC diet and solTNF neutralization.
solTNF inhibition decreases hepatic LCN2 in the presence of diet-induced insulin impairment and liver inflammation
The hepatic inflammatory profile after HFHC diet treatment was confirmed by an increase in hepatic Il-1b (P < 0.0001) and Il-6 (P = 0.0036) mRNA expression (Fig. 3d, e). Diet treatment raised Tlr2 (P = 0.022) and decreased Ppara (P = 0.034) mRNA expression compared to the CD groups (Fig. 3f, g). XPro decreased Tlr2 expression in HFHC diet-treated mice to levels statistically indistinguishable from the CD group. XPro reduced Ppara levels in the CD group. There was no significant difference in hepatic Tlr4 between the experimental groups (Additional file 3: Figure S2F). The assessment of insulin signaling in isolated liver tissue revealed decreased IRS1 phosphorylation at tyrosine 608 (TYR 608) in the HFHC diet/saline compared to the CD group (P = 0.0372); a diet effect decreased the expression of IRS1 (P = 0.0075) in the liver (Fig. 3h, i). A positive correlation between hepatic and colonic LCN2 (r2 = 0.4650, P < 0.0002) and hepatic and plasma LCN2 levels was observed LCN2 (r2 = 0.4168, P < 0.0001) (Fig. 3j, l).
solTNF inhibition decreases colonic LCN2 and tight junction protein alterations associated with HFHC diet
HFHC consumption and solTNF neutralization impact the expression of immunomodulatory genes in the brain
Peripheral injections of a solTNF inhibitor revert central insulin signaling impairment and behavioral deficits in DIO
Evidence suggest that systemic and central energy balance is regulated by hypothalamic insulin, which partially occurs by hypothalamic-liver interactions controlling the glucose metabolism [12, 26]. The impact of DIO and solTNF inhibition on central insulin signaling in the hypothalamus and the pre-frontal cortex (PFC) was investigated to evaluate the impact of these metabolic interactions.
Elevated levels of TNF have been implicated in IR through multiple mechanisms that include the inhibition of the IRS1 through its serine phosphorylation, inhibition of insulin receptor auto-phosphorylation, and impairment of insulin signaling by the activation of phosphatidylinositol 3-kinase Akt mammalian target of rapamycin pathway [13, 39]. Although the effects of anti-TNF therapies on circulating insulin and glucose levels in patients with severe chronic inflammation are well established [18, 40], past attempts at restoring insulin sensitivity using non-selective TNF antagonists that block both membrane-bound and solTNF failed in the presence of metabolic syndrome or obesity conditions [17, 41]. Importantly, our novel findings indicate that selective solTNF neutralization decreases systemic insulin levels without any significant impact on adiposity or weight gain, an outcome that is opposite to that achieved with several non-selective anti-TNF biologics .
Here, we speculate that the impact of selective solTNF neutralization on the plasma insulin levels in DIO may be partially attributable to the XPro effects on hepatic metabolites processing and the hepatic and intestinal decreases in LCN2. IR is attenuated by LCN2 deficiency in animal models, and human studies show a positive association between insulin dysregulation, systemic inflammation, and LCN2 levels [14, 15, 42]. Our findings strongly suggest that solTNF and LCN2 interactions are a potential point of intervention for targeting inflammatory tissue-specific responses in obesity. Hepatic LCN2 is frequently associated with IR and liver alterations in states of positive energy balance [14, 43]. We previously reported that hepatic Lcn2 overexpression, and increased circulating levels of LCN2 are associated with hepatic steatosis and insulinemia in DIO . The current study obtains evidence that colonic LCN2 is increased by the HFHC diet. These data broadly support the contribution of intestinal inflammation to the state of chronic immune activation present in obesity . Additionally, we are the first to demonstrate that selective solTNF signaling with XPro can regulate the hepatic and intestinal LCN2 levels in the presence of hepatic steatosis and metabolic inflammation in DIO. These findings are consistent with the results of human studies that report a positive impact of anti-TNF therapies in reducing NGAL/LCN2 levels in the presence of intestinal alterations [45, 46].
In colonic epithelial cells, LCN2 expression is induced by IL-17A, IL-22, and TNF . In addition to the TNF effects on LCN2, IL-1β, which is also upregulated in the intestine and liver in our animal model, has been associated with LCN2 release by induction of the transcription factor nuclear factor kappa-light-chain enhancer of activated B cells (NFkB) . Our results indicate that elevated hepatic LCN2 and IL-1β are closely associated with hepatic insulin impairment, hepatic steatosis, and excessive lipid circulation, consistent with previous reports [43, 48]. The beneficial effect of solTNF neutralization in reducing intestinal inflammation in obesity is particularly important because intestinal immune alterations are a recognized contributor to metabolic syndrome comorbidities such as glucose and insulin impairment and CNS neurodegenerative processes [49, 50]. We previously demonstrated that the same HFHC used in the current study increases Lcn2 mRNA expression in the hippocampus in the presence of diet-induced insulin resistance . This previous finding in addition to the present results is relevant because LCN2/NGAL is associated with the pro-inflammatory signals that impact AD . LCN2 and its receptors are found in a different human postmortem of the brain regions and in the plasma of AD patients [51, 52, 53]. Notably, this TNF-induced molecule sensitizes neurons to toxic effects of amyloid, promotes phenotypic changes in glia, induces CNS chemokines production, and has been implicated in cognitive deficits [51, 52, 53]. Previous studies suggest that LCN2 may exacerbate insulin resistance in the brain of aging and cognitive-declined subjects [14, 54].
Central insulin impairment is associated with inflammation, oxidative stress, protein deposition, and alterations in synaptic plasticity. Several of these processes are dysregulated in neurodegenerative disorders [36, 55, 56]. Interestingly, HFHC diet impacts insulin and AKT signaling in the hypothalamus and PFC in the presence of metabolic inflammation, and peripheral injections of a solTNF inhibitor were effective at reducing the impact of HFHC diet on insulin signaling in the PFC and hypothalamic Socs3 mRNA expression of mice. Our results provide new and important insight into the role of solTNF in central-peripheral insulin interactions in the states of metabolic inflammation.
While the underlying mechanisms that promote sporadic neurodegenerative disease pathogenesis remain elusive, mounting evidence point to the associations between the disruption of metabolites processing in insulin impairment and neurodegenerative conditions such as AD [57, 58]. Distinct altered metabolic pathways affected by a high-energy diet in this study have been implicated in insulin impairment and the pathogenesis of T2D and neurodegenerative disorders [59, 60]. Additionally, disturbance of the complex gut-liver interactions can impact brain processes and associated central-peripheral energetic balance [61, 62, 63]. In this regard, recent human and animal studies revealed an association between purine dysregulation and brain inflammatory alterations and AD [57, 64]. Purines act as extracellular messengers and are involved in energetic pathways, signal transduction, immune regulation, neurotrophism, and neurotransmission. Moreover, neurodegeneration and diabetes progression can be hastened by disturbances in purine signaling. The brain depends partially on metabolites processed in the periphery. For instance, the transport of nucleotides synthesized de novo in the liver from the blood into the neurons and glia is an essential prerequisite for its central metabolic utilization . Cumulative evidence suggest that several environmental factors can affect the intestinal microbiome and the complex regulation of the brain-gut axis (hypothalamic-pituitary-adrenal axis, vagal modulation, and bacteria-derived metabolites) that ultimately may affect neurodegenerative diseases . Among the intestinal bacterial metabolites disturbed here by HFHC diet, propanoate, a short-chain fatty acid, is involved in gluconeogenesis  that centrally causes inhibition of energy metabolism in brain GABAergic neurons . Another significant aspect of the impact of the HFHC diet on metabolic pathway alterations relevant to insulin metabolism and neurodegeneration is the alteration in the proteoglycan pathways. Heparan sulfate has been identified as an important key factor in neuroinflammation and in the formation of a neurofibrillary tangle in AD [69, 70]. Additionally, in vitro and in vivo studies report that chondroitin sulfate groups are able to modulate insulin amyloid aggregation and protect the brain against amyloid and advanced glycation product-induced toxicity [71, 72]. Together, these findings demonstrate that HFHC diet promotes the formation of metabolites frequently associated with IR and neurodegeneration [60, 73]. Therefore, our results raise intriguing questions regarding the extent to which DIO impact insulin signaling, energy balance, and immune-metabolic interactions in a solTNF-dependent manner to increase the risk for neurodegeneration.
Our results suggest that the HFHC diet impacts central insulin signaling and immune-metabolic interactions in a solTNF-dependent manner to increase the risk for neurodegenerative conditions. Our novel findings indicate that selective solTNF neutralization can ameliorate peripheral and central diet-induced insulin impairment and identify lipocalin-2 as a potential target for therapeutic intervention to target inflammation and insulin disturbances in obesogenic environments. The present study provides evidence that solTNF neutralization is associated with a reduction in downstream pro-inflammatory signaling decreasing the risk for the immune and metabolic dysregulation present in obesity.
Collectively, our findings identify solTNF as a potential target for therapeutic intervention in inflammatory states and insulin disturbances in obesogenic environments to lower risk for AD. Studies are underway to confirm and extend published observations that an obesogenic diet can accelerate AD-like phenotypes  and to directly test the protective effects of solTNF inhibition in vivo against the development of insulin resistance in the brain. An immunomodulatory approach that selectively targets solTNF is likely to have a positive therapeutic impact on broad metabolic and immune interactions in DIO and, if delivered during mid-life, would positively impact patients with metabolic syndrome and/or obesity who are at risk for developing systemic and neurodegenerative conditions later in life.
We thank Dr. Miriam Vos, R.J. Tesi, and the members of the Tansey Laboratory for the productive discussions. Additional thanks to David Szymkowski at Xencor Inc. (Monrovia, CA) and R.J. Tesi at INmune Bio Inc. for providing the XPro1595 biologic.
MESR, MGT, and CJB conceptualized and designed the research. MESR, MCH, DIW, DPJ, and JJ performed the experiments. DIW and DPJ contributed new reagents/analytic tools and performed the metabolomics studies. MESR, MCH, DIW, and CJB analyzed the data. MESR and MGT wrote the paper. MESR, MGT, MCH, CJB, and DIW reviewed and edited the manuscript. All authors read and approved the final manuscript.
Partial funding for this work was provided by a grant from the National Institute of Aging (NIA) at the National Institutes of Health (1RF1AG051514 and 1RF1AG057247, MGT). The Emory Multiplexed Immunoassay Core (EMIC) is subsidized by the Emory University School of Medicine and is one of the Emory Integrated Core Facilities. Additional support was provided by the National Center for Georgia Clinical & Translational Science Alliance of the National Institutes of Health under Award Number UL1TR002378. The content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institutes of Health.
Ethics approval and consent to participate
All experiments were approved by the Institutional Animal Care and Use Committee of Emory University and followed the guidelines established by the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Consent for publication
All authors have consented to publication.
MG Tansey is an ex-employee of Xencor, Inc. and is a co-inventor on the dominant-negative TNF patent but does not hold a significant financial stake in Xencor, Inc. or INmune Bio, Inc. CJ Barnum is an employee of INmune Bio. The other authors declare that they have no competing interests.
- 29.R Core Team (2018). R: a language and environment for statistical computing. R Foundation for statistical computing, Vienna, Austria. Available online at https://www.R-project.org/. Accessed 5 Apr 2018.
- 30.Smyth GK. Limma: linear models for microarray data. In: Gentleman R, Carey V, Dudoit S, Irizarry R, Huber W, editors. Bioinformatics and computational biology solutions using R and Bioconductor. New York: Springer; 2005. p. 397–420.Google Scholar
- 31.Uppal, K. xMSPANDA. (2019). GitHub Repository, https://github.com/kuppal2/xmsPANDA. Accessed 5 Apr 2018.
- 38.Kaidanovich-Beilin O, Lipina T, Vukobradovic I, Roder J, Woodgett JR. Assessment of social interaction behaviors. J Vis Exp. 2011;(48):2473. https://doi.org/10.3791/2473.
- 47.Karlsen JR, Borregaard N, Cowland JB. Induction of neutrophil gelatinase-associated lipocalin expression by co-stimulation with interleukin-17 and tumor necrosis factor-alpha is controlled by IkappaB-zeta but neither by C/EBP-beta nor C/EBP-delta. J Biol Chem. 2010;285(19):14088–100.PubMedPubMedCentralGoogle 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. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.