Introduction
It has been widely accepted that inflammation is central to the development of diabetic nephropathy (DN). This disorder of glucose control in diabetics can cause inflammation and cellular stress, causing functional deterioration seen in DN [
1,
2]. Studies have shown that macrophage polarization occurs in renal inflammation and plays a decisive part in DN, and secretion of pro-inflammatory mediators by macrophages after activation contributes to deterioration of DN [
3]. Macrophages are subdivided into two subpopulations: the classically-activated or M1 phenotype macrophages and the alternatively-activated or M2 macrophages. M1 phenotype macrophages are observed to deliver higher levels of pro-inflammatory cytokines to deal with infections [
4,
5]. Macrophages can re-programme their metabolism and differentiate into pro- or anti-inflammatory phenotypes. Inflammatory macrophage activation involves metabolic adaptation towards glycolysis, videlicet, M1 macrophages mainly utilize glycolysis to provide energy [
6,
7].
Transforming growth factor β-activated kinase 1-binding protein 1 (TAB1) is a specific protein that interacts with transforming growth factor β-activated kinase 1 (TAK1). TAB1/TAK1 can activate nuclear factor κB (NF-κB) in high glucose (HG)-induced BMMs, regulating macrophage activation [
8‐
12]. Studies suggest that activation of the NF-κB signaling pathway upregulates HIF-1α activity and enhances glycolytic metabolism in an autonomous fashion, HIF-1α may be considered as a novel therapeutic target for DN [
13‐
16]. HIF-1α has been reported to be closely involved in energy metabolism and it increases the activity of glycolytic enzymes such as hexokinase-1 (HK1), 6-phosphofructo-2-kinase/fructose-2, 6-biphosphatase 3 (PFKFB3) and lactate dehydrogenase A (LDHA) [
17,
18]. Therefore, TAB1 may be involved in macrophage glycolysis and activation through NF-κB/HIF-1α, which is a principal link in the genesis and evolution of DN.
In this study, we used siRNA and lentiviral vector technology to knock down TAB1 in HG-induced macrophages and STZ-induced diabetic mice, then evaluated glycolysis, polarization of macrophages and levels of inflammation, thereby exploring the relationship between the TAB1/NF-κB/HIF-1α signaling pathway and glycolysis in polarization of macrophages in the DN microenvironment.
Materials and methods
Materials and reagents
The following materials and reagents were used in this study: mouse fibroblast cell line (L-929) was obtained from the cell bank of the Chinese Academy of Sciences; lipofectamine RNA iMAX transfection and Trizol reagent were obtained from Thermo Scientific (USA); HiScript II Q RT SuperMix for qPCR and AceQ qPCR SYBR Green Master Mix were obtained from Vazyme (Vazyme, Nanjing, China); A3500 Reverse Transcription System was obtained from Promega (Promega, Madison, USA); FITC anti-mouse F4/80, APC anti-mouse/human CD11b, FITC rat IgG2a κ isotype control and APC rat IgG2a κ isotype control were obtained from BioLegend (San Diego, California, USA); iNOS Monoclonal Antibody (CXNFT), APC were from eBioscience™ from Thermo Fisher Scientific (USA), anti-NF-κB p65 antibody and anti-NF-κB p65 antibody(ChIP) was obtained from Cell Signaling Technology (Beverly, MA, USA); anti HIF-1α, anti-HK1, anti-TAB1 and anti-inducible nitric oxide synthase (iNOS) were obtained from Abcam Biotechnology (Cambridge, UK); anti-PFKFB3, anti-LDHA and anti-P-P65 were from Wuhan Sanying Biotechnology Inc (Sanying, Wuhan, China); anti-CD68 and anti-TNF-α were from Arigo Biolaboratories Corp (Arigo, Taiwan, China); anti-IL-1β from ABclonal Technology (Wuhan, China) and ELISA kits for mouse IL-1β, MCP-1 and TNF-α were obtained from R&D (USA) and MULTI SCIENCES(Hangzhou, China). Glucose absorption kits were obtained from Biovision (USA), Lactic Acid assay kit from Nanjing jiancheng (Nanjing, China) and TAB1-siRNA from Ribobio (Guangzhou, China). TAB1 lentivirus was obtained from Genepharma (Shanghai, China). Immunohistochemical double-stain kit was obtained from Maxin (Fuzhou, China). Mouse microalbumin ELISA kits were from Abcam Biotechnology. Chromatin IP Kit were from Cell Signaling Technology (USA).
Isolation and culture of BMMs
BMMs were isolated from C57BL/6 J mice (6–8 weeks, male). Mice were killed with ether and bone marrow collected from tibias and femurs. Cells were washed in cold phosphate buffered saline (PBS) + 2% heat-inactivated fetal calf serum (FBS) and then centrifuged at 500 g for 5 min. Cell pellets were resuspended in red blood cell lysis buffer and washed again. Resuspended cells were maintained in DMEM medium supplemented with 15% L-929 cell supernatant, 10% FBS and 1% penicillin/streptomycin. They were cultured at 37° C and 5% CO2 and medium was replaced on day 3. The adherent cells were BMMs and were used for subsequent experiments on day 7.
Flow cytometry (FCM) analyses
For surface marker analysis, BMMs in PBS were blocked with anti-CD16/CD32 to prevent non-specific binding and then washed in PBS and centrifuged for 10 min, the supernatant was decanted and 0.5 ml PBS/tube containing FITC-conjugated anti-mouse F4/80 and APC-conjugated anti-mouse CD11b was added—the samples were incubated in the dark at room temperature for 30 min. For intracellular cytokine staining, the cells were fixed (4% paraformaldehyde) and permeabilized (3% triton × 100), after blocked with anti-CD16/CD32, stained with APC-conjugated iNOS Monoclonal Antibody and FITC-conjugated anti-mouse F4/80. The macrophages were then washed. Finally, 1 ml of PBS was added to the resuspended macrophages, which were analyzed with a BD FACS Calibur flow cytometer.
Transfection with TAB1 siRNA
Following the manufacturer’s instructions, TAB1 siRNA and the negative control (NC) were transfected into BMMs by lipofectamine RNA iMAX transfection for 24 h and were selected for the optimal transfection band by Western blot. The sequences of TAB1 siRNA were: (1) 5′-GAGGAACTTTGGCTATCCA-3′; (2) 5′-GTGGATGGGTTACAGGTTA-3′; (3) 5-GGATTACAAGGTCAAATAT-3′.
Experimental groups
First, the stimulation time for d-glucose was optimized. According to previous work of our group, we used 30 mmol/L glucose as a stimulating factor. The BMMs were grouped into: (1) mannitol control group (D-mannitol), (2) negative control siRNA (control siRNA), (3) blank control (control), (4) hyperglycemic stimulation (HG), (5) TAB1 siRNA (TAB1 siRNA), (6) TAB1 siRNA HG stimulation group (TAB1 siRNA + HG).
ELISA
The levels of Inflammatory cytokines were tested by ELISA. BMMs were treated with TAB1 siRNA or not for 24 h and then stimulated with HG for 24 h; the kidney tissue of each mice was cuting to pieces and then ultrasonically broken with PBS. Supernatants and tissue homogenate were assayed following the manufacturer’s instructions, OD values were measured at 450 nm and actual concentrations calculated using the standard curves.
Lactic acid production and glucose absorption
BMMs were pre-treated with TAB1 siRNA or not for 24 h, then exposed to HG for 24 h. Their lactic acid production and glucose absorption were detected using the lactic acid production kit and glucose absorption kit.
Confocal microscopy analysis
BMMs were fixed using paraformaldehyde (4%) at room temperature for 30 min. 5% bovine serum albumin + 0.2% Triton X-100 was used to block for 2 h and BMMs were then incubated with specific primary antibodies at 4 °C overnight. FITC anti-mouse F4/80 and corresponding fluorescent secondary antibody were added in the dark at 37 °C. After washing three times with PBS, the nuclei were stained with 40, 60-dimidyl-2-phenylindole (DAPI) for 10 min and observed under a Leica TCS SP5 laser confocal microscope. The semi-quantitative analysis of laser confocal was performed by ImageJ.
Western blotting
Renal samples and BMMs were treated with Radio-Immunoprecipitation Assay and phosphatase inhibitor mixed buffer. Proteins were separated by SDS-PAGE electrophoresis and transferred to polyvinylidene difluoride (PVDF) or nitrocellulose (NC) membranes. These were incubated overnight at 4 °C with specific primary antibodies and then stained with horseradish peroxidase-labeled secondary antibody. Detection was performed with enhanced chemiluminescence (ECL) reagent and visualized with the chemiluminescence GE 600 system. Results were expressed as relative ratios and quantified by Image J.
RNA extraction and qRT- PCR
Following manufacturer’s instructions, total RNA was extracted from the BMMs using Trizol reagent. 1 μg RNA was reverse transcribed to cDNA using the Reverse Transcription System. The cDNA was amplified by real-time PCR with AceQ qPCR SYBR Green Master Mix. Primers to detect mRNA were as follows: GAPDH, as the internal control: Forward primer: 5′-ACCCCAGCAAGGACACTGAGCAAG-3′, Reverse primer: 5′-GGCCCCTCCTGTTATTATGGGGGT-3′; HK1: Forward primer: 5′-TGCCATGCGGCTCTCTGATG-3′, Reverse primer: 5′-CTTGACGGAGGCCGTTGGGTT-3′; PFKFB3: Forward primer: 5′-AGCCAGCTACCAGCCTCTTG-3′, Reverse primer: 5′-AATTCGGCTCTGGATGTGGT-3′; LDHA: Forward primer: 5′-CAAAGACTACTGTGTAACTGCGA-3′, Reverse primer: 5′-TGGACTGTACTTGACAATGTTGG-3′; MCP-1: Forward primer: 5′-TTGACCCGTAAATCTGAAGCTAAT-3′, Reverse primer: 5′-TCACAGTCCGAGTCACACTAGTTCAC-3′; IL-1β: Forward primer: 5′- GCCTCGTGCTGTCGGACCCATAT-3′, Reverse primer: 5′-TCCTTTGAGGCCCAAGGCCACA-3′. The relative mRNA levels were calculated by the 2-△△Ct method with values normalized to the GAPDH reference gene.
Mice
The study was performed according to the international, national and institutional rules concerning animal experiments and the Animal Ethics Committees of the Faculty of Anhui Medical University approved all experimental protocols, in accordance with “Principles of Laboratory Animal Care and Use in Research” (Ministry of Health, Beijing, China). 24 C57BL/6 J littermate male mice (6–8 weeks) were obtained from the Model Animal Research Center of Nanjing University and were maintained in Specific Pathogen Free conditions at 24 °C and 60% humidity, light and dark for 12 h each day. Mice were housed in separate cages and were free to eat and drink. Mice were given STZ (Sigma Chemical Co., St. Louis, MO, USA) at 50 mg/kg body weight daily for 5 consecutive days to establish diabetes. Animal groups (n = 6 each group) were: C (control), STZ, STZ + Negative control (NC) lentivirus, STZ + TAB1 lentivirus. TAB1 lentivirus (5′-GTGGATGGGTTACAGGTTA-3′) was administered weekly to the STZ + TAB1 lentivirus group mice through the tail vein at a dose of 1 × 109, 50 µl. NC lentivirus was administered at the same dose as the STZ + NC lentivirus group. Other groups of mice were injected with the same amount of normal saline via the tail vein (C group and STZ group).
Physical and biochemical analyses
Body and kidney weight were measured and blood glucose was assessed when the mice were sacrificed after 12 weeks. Twenty-four-hour urine was collected through metabolic cages and urine volume calculated. Urinary albumin levels were assayed using a mouse microalbumin ELISA kit according to the manufacturer’s instructions.
Histology and Immunohistochemistry
All mice were sacrificed after 12 weeks. Then, 4% paraformaldehyde-fixed and paraffin-embedded fresh renal tissue sections were cut into sections of around 3 µm thickness. After deparaffinization, periodic acid–Schiff (PAS) staining was used to assess pathological damage to the kidney. The glomerular mesangial expansion index and the tubulointerstitial injury index were evaluated and graded using 10 randomly selected visual fields.
After antigen retrieval, tissue sections were blocked with serum for 30 min at 37 °C and then incubated with primary antibody (overnight at 4 °C). After incubation with secondary antibody for 1 h at 37 °C, they were stained with 3, 3-diaminobenzidine and hematoxylin. Staining for TAB1, CD68, TNF-α, IL-1β, HK1, PFKFB3 and LDHA was detected using the Image J System. According to the double immunohistochemical staining kit instructions, incubation and staining with the two primary antibodies was performed separately and the results observed under a standard microscope.
Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) assay was performed using the e SimpleChIP® Enzymatic Chromatin IP Kit (Magnetic Beads) #9005 (Cell Signaling Technology, USA) according to the manufacturer’s instructions. Chromatin was precipitated with NF-κB p65 antibodies and samples were analyzed by PCR. The murine HIF-1α promoters were amplified with the primer pairs: Forward primer: 5′-TGTGGCTTCCAACTGGGTGTT-3′, Reverse primer: 5′-AGGGTCAGCAGAATACCCCT-3'.
Statistical analysis
Statistical analysis was conducted using SPSS software version 17.0 (SPSS Inc. Chicago, IL, USA). All data were analyzed by one-way analysis of variance (ANOVA) and presented as mean ± standard deviation (SD). The difference between groups was tested by least significant difference (LSD) and the Levene method was used for homogeneity test of variance. p values < 0.05 were considered to be statistically significant.
Discussion
DN is a frequent microvascular complication in diabetes but current treatments, such as blood glucose and blood pressure control and renin–angiotensin–aldosterone system inhibitors, remain suboptimal. Increasingly, research has suggested that inflammation is a critical factor activated by metabolic, hyperglycaemic, AGEs [
19] and haemodynamic derangements. Macrophages may be pivotal to this process and unexplained inflammatory cells are known to mediate kidney inflammation [
20‐
22]. Research has shown that macrophage infiltration is closely involved in regulation of serum creatinine, proteinuria, and interstitial fibrosis [
23].
Macrophages have two major subtypes: the M1 macrophage (or classically activated) is involved in inflammation by secreting pro-inflammatory cytokines in response to pathogens or other immune stimuli (including but not limited to DN). Conversely, the M2 macrophage exhibits anti-inflammatory properties [
24]. In vivo, M1 macrophage polarization and proinflammatory cytokine expression increase, and M2 macrophage polarization and anti-inflammatory cytokine expression decrease in DN mice. The imbalance of M1/M2 macrophages reflects the pathogenesis of DN. In vitro, DN microenvironmental products, such as high glucose and AGEs tend to differentiate macrophages into the M1 phenotype, secreting more iNOS and pro-inflammatory cytokines including IL-1β, TNF-α and MCP-1 [
25,
26]. Our study suggests that expression of iNOS in HG-induced BMMs is markedly enhanced. The mRNA expression and secretion of inflammatory cytokines MCP-1 and IL-1β, were significantly elevated in the HG group. The result that HG induces polarization of BMMs towards the M1 phenotype is consistent with recent studies [
25]. In addition, we found that TAB1-siRNA can suppress the polarization and inflammation caused by HG—levels of INOS, MCP-1 and IL-1β in macrophages under high glucose were inhibited by TAB1 siRNA. CD68 is used as a valuable cytochemical marker for immunostaining monocytes/macrophages in histochemical analysis of inflammation of tumor tissues [
27]. Our in vivo results showed that CD68 and iNOS were overexpressed in the kidney after STZ treatment of mice. Inhibition of TAB1 by lentivirus lead to a reduction in infiltration and activation of M1 macrophages. In addition, TAB1 lentivirus markedly attenuated the upregulation of IL-1β and TNF-α induced by STZ. Additionally, the results of our study indicated that TAB1 lentivirus noticeably prevented the progression of DN—TAB1 lentivirus reduced the albuminuria excretion, tubulointerstitial damage index and glomerular mesangial expansion index in diabetic mice. These results indicate that TAB1 regulates the occurrence of macrophage polarization and inflammation, improving DN. However, the mechanisms behind this remain unclear.
It has been reported that many cancers have an altered metabolism, called the Warburg effect. These reports propose a neoteric form of metabolism. Cancer cells preferentially metabolize glucose through glycolysis, thereby producing lactic acid as an end product without using oxygen [
28]. Under this transformation, tumor cells survive with optimized utilization of glucose. Increasing evidence suggests that the Warburg effect occurs in other situations. Activated inflammatory immune cells also exhibit metabolic features similar to glycoclastic cancer cells [
29,
30]. M1 macrophage, which have a high energy demand, obtain energy from glycolysis to apidly provide ATP to promote inflammation. So macrophages activated by pro-inflammatory factors switch towards glycolysis, ignoring oxygen. For example, as a classical activator of M1 phenotype macrophages, lipopolysaccharide (LPS) has been reported to be involved in metabolic reprogramming of macrophages through increasing translation of HIF-1α mRNA, increasing expression of u-phosphofructokinase2 and inhibiting expression of Adenosine 5‘-monophosphate (AMP)-activated protein kinase, which contribute to macrophage glycolysis. Conversely, M2 phenotype macrophages appear to be more inclined towards oxidative phosphorylation—numerous processes that activate the glycolytic process in pro-inflammatory macrophages are decreased in M2 macrophages, which increasingly flux through OXPHOS [
31,
32]. Interestingly, imbalance in glycolytic metabolites, such as succinate and lactate, can also interferes with inflammatory phenotypes via activating signalling pathways, post-translational mechanisms, epigenetic landscape and pos-transcriptionally regulate [
33]. As the rate-limiting glycolytic enzymes, PFKFB3, HK1 and LDHA play a critical role in glycolysis and are over-expressed in glycoclastic cells [
34‐
36]. Recent studies report that M1 macrophages express high levels of glycolytic enzymes and that HK1 regulates NLRP3 inflammasome activation, which subsequently regulates the production of IL-1β. PFKFB3 participates in the anti-viral response of macrophages. Current research also suggests that LDHA may be a key metabolic enzyme involved in regulation of inflammation [
37]. Glycolysis promote the inflammation through multiple mechanism, metabolic reprogramming of macrophages promises to be a new target for macrophage-mediated inflammatory diseases. In our study, lactic acid production and glucose absorption test results showed that lactate production and glucose uptake in the HG group were significantly higher when compared with the control group. Compared with the HG group, lactic acid production and glucose absorption capacity of the TAB1 siRNA + HG group were decreased, and this difference was statistically significant. In the HG group, we also observed stronger staining for HK1, PFKFB3 and LDHA when compared with controls, and HG-induced BMMs treated with TAB1 siRNA showed a reduction in HK1, PFKFB3 and LDHA staining. Results of Western blots and RT-PCR also showed the same trend. The protein levels and mRNA of glycolytic enzymes HK1, PFKFB3 and LDHA in HG-induced BMMs were distinctly higher than the levels in the control group. In the TAB1 siRNA + HG group, they were significantly weaker than those in the HG group, which proved that the glycolysis of HG-induced M1 macrophages was enhanced and expression of glycolytic enzymes HK1, PFKFB3 and LDHA were up-regulated in an HG environment. TAB1 silencing inhibited the enhancement of glycolysis under HG conditions. Similarly, in vivo, immunohistochemical and Western blotting demonstrated that STZ-induced diabetic mouse kidneys showed higher levels of HK1, PFKFB3 and LDHA, and these STZ-induced glycolytic enzymes were notably inhibited by TAB1 lentivirus. Focusing on macrophages in renal tissue, immunohistochemical double-staining revealed that TAB1 inhibition decreased STZ-induced glycolysis of renal macrophages. Therefore, Macrophages enhance glycolysis upon activation by glucose dysregulation, and silencing of TAB1 inhibits macrophage glycolysis while reducing inflammation in vivo and in vitro.
NF-κB is a recognized downstream signaling target molecule of TAB1/TAK1. After Toll-like receptor signal transduction, TAK1/TAB1 mediates NF-κB p65 nuclear entry through IKKα/β, inducing further inflammation [
9]. Emerging evidence points to NF-κB control over the HIF signaling pathway and interaction between the HIF signaling pathway and the NF-κB signaling cascade [
14]. It has been confirmed that NF-κB controls HIF-1α transcription in macrophages activated by LPS through ChIP [
38]. High expression of specific transcription factors, principally HIF-1α, NF-κB and Octamer Binding Transcription Factor 1, contribute to the Warburg effect. In other words, HIF-1α is involved in regulating the energy metabolism of cells and makes cells more inclined to utilize glucose through glycolysis. Simultaneously, studies have shown that patients with DN have higher HIF-1α levels and sodium-dependent glucose transporters 2 inhibitors, recently used to treat DN, also target HIF-1α [
12]. In the classical inflammatory cells, HIF-1α is involved in glycolysis and has a role in inflammation. But whether HIF-1α can be driven by high glucose via NF-κB has not been studied in depth. Therefore, it is worth exploring thoroughly whether TAB1 regulates macrophage glycolysis through TAB1/NF-κB/HIF-1α and mediate inflammation in M1 phenotype macrophage. In our study, we observed that the green fluorescence of NF-κB p65 in the control group was mainly located in the cytoplasmic region. In the HG group, the fluorescent of NF-κB p65 was mostly found in the nuclear region, and treatment with TAB1 siRNA significantly inhibited the presence of green fluorescence in the nucleus. Western blotting revealed that the levels of TAB1, NF-κB p65 and HIF-1α were increased significantly in the HG group. In the TAB1 siRNA + HG group, TAB1, NF-κB and HIF-1α protein expression both decreased, in contrast to the HG group. In vivo, the expression of TAB1, p-p65 and HIF-1α were measured in mouse kidneys. Western blot results showed that although the protein level of p65 remained unchanged, the expression of p-p65 was increased in STZ-induced diabetic mice, while TAB1 lentivirus decreased its expression significantly—this is consistent with the results of P65 nuclear expression in vitro. The same is true of the expression trends for TAB1 and HIF-1α. Results from ChIP showed specific binding of NF-κB to HIF-1α in HG-induced macrophages, confirming the potency of TAB1/NF-κB regulating HIF-1α. The above results suggest that TAB1 is involved in the regulation of NF-KB and HIF-1α to a certain extent, and this may be closely related to glycolysis and activation of macrophages.
At present, new developments in the treatment of DN include treatment options such as MCP-1 inhibitors that inhibit macrophages and inflammation, but more precise treatment targets are still urgently needed. Our study demonstrates that glucose dysregulation can initiate macrophage glycolysis and M1 polarization, causing the release of inflammatory factors. These effects are closely related to aggravation of DN. Downregulation of TAB1 signaling reverses glycolysis and M1 polarization of macrophages, reducing inflammation, improving the clinical and pathological manifestations of DN. In summary, our study proposes that the TAB1/NF-κB/HIF-1α signaling pathway mediates the glycolysis and polarization of macrophages to promote inflammation in DN.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.