Hyperglycemia amplifies TLR-mediated inflammatory response of M(IL4) macrophages to dyslipidemic ligands

Abstract

1. Introduction

Hyperglycemia (HG) is a key factor in diabetes pathology as it provides the biochemical environment for detrimental inflammatory responses that consequently lead to the development of diabetic vascular complications.¹ HG leads to the overproduction of reactive oxygen species (ROS), which activates proinflammatory transcription factors such as NF-κB and AP-1, responsible for the stimulation of proinflammatory chemokines/cytokines and adhesion molecules in monocytes/macrophages.² In addition, HG results in the elevated formation of advanced glycosylated end products (AGEs), triggering the activation of the receptor of AGEs, followed by the upregulation of the interferon pathway and the downstream pathways, which leads to cardiovascular disorders.³ Macrophages are responsible for the initiation, regulation, and resolution of the inflammatory response; they are key innate immune cells that promote the inflammatory reaction in diabetes pathology and diabetes-related micro- and macrovascular complications. Two major vectors of macrophage polarization are widely accepted: M1, an acute inflammatory macrophage stimuli, is IFNγ (M(IFNγ)), and M2, induced by a spectrum of stimulus, where the first identified prototype stimulus is IL4 (M(IL4)). M(IFNγ) macrophages initiate and amplify acute inflammation, while M(IL4) macrophages have a broad range of anti-inflammatory and healing activities.^4–6 M(IFNγ) macrophages utilize glucose, while the energy sources of M(IL4) are fatty acids.^7,8 HG, the key factor in diabetes pathology, promotes metabolic activity in macrophages, which also triggers obesity-associated insulin resistance.^9,10 It has been demonstrated that HG induces a mixed M(IFNγ)/M(IL4) cytokine profile, characterized by the elevated production of TNFα, IL1β, and IL1Ra.^11,12 The role of M(IL4) macrophages in the development of metabolic syndrome and diabetes is very complex and includes both beneficial and unfavorable effects since both diseases are heterogeneous in their phenotypes and functional activities compared to M(IFNγ).^13,14 Toll-like receptors (TLRs) are pattern recognition receptors that recognize pathogen-associated molecular patterns and nonmicrobe molecules, including endogenous ligands, responsible for tissue damage and inflammation.¹⁵ Several metabolic ligands of TLRs that belong to the categories of palmitic acids and free fatty acids are characteristic components of dyslipidemia, linking metabolic syndrome, diabetes, and cardiovascular events.^13,16,17 TLR-mediated cytokine responses to dyslipidemic ligands can amplify the detrimental inflammation in diabetic patients.¹⁸ Enhanced inflammatory cytokine production is critical for the development of type 1 diabetes as well as for the progression of type 2 diabetes.¹⁹ Increased expression of TLR4 has been linked to diabetes,^20,21 and saturated fatty acids are known to be recognized by TLR4 and can induce the short- and long-term TLR4-mediated production of inflammatory cytokines.¹³ Our study was designed to examine how HG affects the TLR system and how HG can amplify the inflammatory response of macrophage dyslipidemic ligands.

2. Materials and methods

2.1 Monocyte isolation and generation of macrophages

Human monocytes were isolated from buffy coats from individual donors. Buffy coats were provided by the German Red Cross Blood Service Baden-Württemberg–Hessen. Buffy coats were obtained from healthy donors after informed consent. Selection of monocytes occurred through selection by anti-CD14 antibodies and magnetic activated cell sorting (Milteny Biotech) as previously described.^11,22 The obtained monocytes were cultured at 1×10⁶ cells/mL in customized serum-free medium supplemented with 5 mM (normal glucose) and 25 mM (high glucose) glucose (Life Technology) at 7.5% CO₂ for the time periods up to 6 d without medium change. Macrophage colony-stimulating factor (M-CSF) at concentration 5 ng/mL was added for all macrophage cultures. IFNγ at a concentration of 100 ng/mL was used to generate the M(IFNγ) phenotype, and IL4 at a concentration of 10 ng/mL was used to generate the M(IL4) phenotype. All cytokines were purchased from Peprotech. Lipopolysaccharide (LPS) (cat. tlrl-3pelps; InvivoGen) was used at a concentration of 1 µg/mL, and Pam3CSK4 (cat. tlrl-pms; InvivoGen) was used at a concentration of 10 ng/mL. Stimulation with LPS and Pam3CSK4 was performed on day 6 of macrophage cultivation for 24 h.

2.2 RNA isolation and cDNA synthesis

cDNA synthesis was performed using the SensiFAST cDNA Synthesis Kit from BIOLINE according to the manufacturer's instructions. The obtained cDNA was diluted 10 times and 1 μL was used for quantitative reverse transcriptase polymerase chain reaction (RT-qPCR). Levels of mRNA from IL1β and TLR1–9 were quantified using the TaqMan SensiMix II Probe Kit (Meridian Life Science) in the standard conditions. The obtained cDNA was diluted 1:10 with double distilled water, and 1 μL was used for RT-qPCR.

2.3 RT-qPCR

Preformulated TaqMan gene probes were purchased from Applied Biosystems for the following human genes: TLR1 (Hs00413978_m1), TLR2 (Hs00610101_m1), TLR4 (Hs00152939_m1), TLR5 (Hs00152825_m1), TLR6 (Hs00271977_s1), TLR7 (Hs00152971_ml), TLR8 (Hs00152972_m1), IL1β (Hs01555410_m1), CCR2 (Hs00704702_s1), and 18S (Hs99999901_s1). RT-qPCR was performed using the TaqMan SensiMix II Probe Kit (Bioscience, BIO-83005), and reactions were performed in duplicate using 96-well optical plates on a LightCycler 480 Real-Time PCR System (Roche). 18S was used as a reference. To calculate the gene expression change of selected genes, the ΔΔCT method was used. According to this method, the threshold cycle values (CT) for a specific mRNA expression in each sample were normalized to the CT values of 18S mRNA in the same sample. This provides ΔCT values that were used to calculate the changes in gene expression levels. Thereby, for each gene, the gene expression change in each condition is defined by the difference of the ΔCT value of other specific conditions divided by the ΔCT value of normal glucose (NG)/M0 (nonstimulated macrophages).

2.4 Flow cytometry

Cells were removed from the incubator (37°C, 7.5% CO₂) and placed on ice for 20 min. Cells were collected into FACS tubes using cell scrapers. After washing cells with phosphate-buffered saline (PBS) and FACS buffer (PBS, 1% bovine serum albumin, and 0.01% sodium azide), 200 µL 2% Paraformaldehyde (PFA) was used to fix the cells. After a 15-min fixation, PFA was removed, and cells were washed with FACS buffer 2 times and treated with a 1% saponin solution for 30 min to permeabilize cell membranes. Next, cells were washed with FACS buffer 2 times and resuspended in 100 µL FACS buffer supplemented by 10 µL FcR blocker (Milteny Biotech). The following antibodies were used: PE-TLR1-IgG1κ (cat. 12-9911-42; eBioscience), PE-TLR2-IgG2aκ (cat. 12-9922-42; eBioscience), PE-TLR4-IgG2aκ (cat. 12-9917-42; eBioscience), PE-TLR5-IgG2aκ (cat. 394504; BioLegend), PE-TLR6-IgG1κ (cat. 334708; BioLegend), PE-TLR8-IgG2aκ (cat. 395503; BioLegend), PE-CCR2-IgG2aκ (cat. 357206; BioLegend), PE-IgG1κ (cat. 12-4714-82; eBioscience), PE-IgG2aκ (cat. 14-4724-82; eBioscience), PE-IgG1κ (cat. 400112; BioLegend), and PE-IgG2aκ (cat. 400212; BioLegend). Incubation was for 1 h, at room temperature, in dark conditions followed by 2 times washing with 2 mL Cell Wash (BD Biosciences). Data were acquired with BD FACSDiva software 6.0 (BD Biosciences). Mean fluorescence intensity (MFI) level was acquired by the FlowJo software (TreeStar). For the cells stimulated by the same cytokines, the data for the TLR expression are presented as delta mean fluorescence intensity (ΔMFI): MFI obtained with isotope control (in NG or in HG conditions) is subtracted from MFI acquired from cells stained with a specific anti-TLR antibody, and the fold change equals ΔMFI(HG)/ΔMFI(NG).

2.5 Cytokine secretion assay

Concentration of secreted IL1β was determined in macrophage culture supernatants using enzyme-linked immunosorbent assays (ELISAs) from R&D Systems according to the manufacturer's instructions.

2.6 Statistics

Statistical analysis was performed using GraphPad Prism 9 software (GraphPad Software). Bar graphs show mean ± SEM. The significance of the data was analyzed using the ratio paired Student's t test. We considered a 2-tailed P value of less than 0.05 to indicate statistical significance (confidence level 95%); ns = nonsignificant, P < 0.05 (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001).

3. Results and discussion

The effect of HG on the expression of TLR (1–9) in M(NC), unstimulated control macrophages; M(IFNγ), stimulated with INFγ; and M(IL4), stimulated with IL4 was quantified by RT-qPCR. In our study, we keep the nomenclature of macrophages as M(NC), M(IFNγ), and M(IL4) to avoid confusion due to the number of other stimuli that can lead to M1 and M2 phenotypes. We have already analyzed the phenotypes of macrophages under stimulation with IFNγ and IL4 in normoglycemic and hyperglycemic conditions in previously published studies.^11,22 Here, the phenotypes of M(IFNγ) and M(IL4) were additionally controlled by the analysis of TNFα and IL12/23 (for M1) and CD206 and CCL18 (for M2) (Supplementary Fig. 1). TLR expression levels were examined on day 6 of monocyte-to-macrophage differentiation in a macrophage-specific, serum-free medium supplemented by a macrophage colony-stimulating factor (M-CSF). M(NC), M(IFNγ), and M(IL4) macrophages have been differentiated from CD14⁺ monocytes of healthy individuals in the NG (5 mM glucose) and HG (25 mM glucose) conditions for 6 d. Figure 1A demonstrates that in each macrophage subtype, HG elevated mRNA levels for the specific sets of TLRs (TLR1, 2, 4, 5, 6, 8). HG statistically significantly increased mRNA levels of TLR1 in M(NC) and M(IFNγ). In M(IFNγ), mRNA levels of TLR2 and TLR6 (which act as a heterodimer) were upregulated by HG. In M(IL4), HG upregulated mRNAs for TLR4 and TLR5, which worked as single-chain receptors (see Supplementary Table 1). Figure 1B summarizes the effects of HG in percentages of the increase compared to the mRNA expression level of TLRs in NG (values in NG conditions were set up as 100%, and effects of HG were calculated as a percentage increase or decrease of values in NG conditions). HG upregulated mRNA levels of TLR1 and TLR8 in M(NC) by 82% and 134% (Fig. 1B, Supplementary Table 1A). In M(IFNγ), HG upregulated mRNA levels of TLR1 by 59%, TLR2 by 53%, and TLR6 by 50%. In M(IL4), HG upregulated mRNA levels of TLR4 by 79% and TLR5 by 92% (Fig. 1B, Supplementary Table 1). Previously, in a murine model, the expression of TLR5 was also found to be upregulated by glucose in the isolated islets of Langerhans, resulting in the elevated production of ROS, inflammatory cytokines, and heat shock proteins and the reduction of insulin secretion.¹⁹ It has been also reported that TLR5 knockout mice developed a condition similar to the human metabolic syndrome involving hyperlipidemia, hypertension, insulin resistance, obesity, and altered gut microbiota.^23,24

Fig. 1.

Hyperglycemia affects TLR gene expression in human differentially activated macrophages. Monocyte-derived macrophages were cultured for 6 d in NG (5 mM) and HG (25 mM) conditions and stimulated with either IFNγ or IL4 or without further cytokines (M(Control)). (A) RT-qPCR analysis of TLR expression in NG and HG in M(NC), M(IFNγ), and M(IL4). Bar graphs show mean ± SEM of 6 donors for M(NC) and M(IFNγ) and 8 donors for M(IL4). (B) Increase in percentage (showed as %increase on the y-axis) of TLRs’ mRNA expression level induced by HG compared to NG. Only statistically significant TLRs that were upregulated by HG are demonstrated on the graph. The bars are calculated from the ratio of mean values for M(NC) and M(IFNγ) and M(IL4). (C) The significant response (NG or HG in M(IL4)) presented as 8 individual donors. The significance of the data was analyzed using ratio paired Student's t test.

Open in new tabDownload slide

The most significant effect of HG was the upregulation of the TLR4 mRNA expression in M(IL4) (Fig. 1A, B). Fig. 1C represents individual responses (n = 10) of the TLR4 expression in NG and HG conditions in M(IL4). Fold changes of the HG effect for all TLRs (1–9) are summarized in Supplementary Table 1. We identified that the strong stimulating effect of HG on TLR4 mRNA can explain the mechanism of the essential role of TLR4 in diabetic complications, mentioned in earlier published materials by other groups. Thus, TLR4 was shown to promote tubular inflammation in diabetic nephropathy.²⁵ Also, the expression of TLR4 is significantly increased in wounds of diabetic patients.²⁶

Next, we quantified the effect of HG on the expression of TLRs at the protein level using flow cytometry (Fig. 2 and Supplementary Fig. 2). HG statistically significantly increased protein levels of TLR1 (fold change 1.25) and TLR8 (fold change 2.47) in M(NC), TLR2 (fold change 1.58) and TLR6 (fold change 1.34) in M(IFNγ), and TLR4 (fold change 1.51) and TLR5 (fold change 1.59) in M(IL4) (Fig. 2). Here, for the first time, we identified that HG alone, without any other factors, upregulates TLR4 expression in M(IL4) macrophages. Our data show that M(IL4), which is prevalent in the healing tissue, can respond to HG by the elevation of TLR4 expression. In patients with metabolic syndrome or untreated diabetes, monocytes in the circulation, as well as their precursors in the bone marrow, are exposed to HG. Hyperglycemic epigenetic memory in monocytes and macrophages has been suggested earlier to contribute to the pathology of diabetes-associated cardiovascular disease.²⁷ We have recently identified one of the mechanisms of such memory and have shown that HG can induce epigenetic change in human primary monocyte-derived macrophages by enhancing the durable presence of activating histone marks on the promoters of genes responsible for endothelial inflammation.²²

Fig. 2.

Hyperglycemia enhances TLR protein expression in human differentially activated macrophages. Flow cytometry analysis was used to quantify protein expression levels of TLRs. Analysis was done on day 6 of macrophage differentiation in NG (5 mM) and HG (25 mM) conditions. The ΔMFI is used to present the data for TLR expression. n = 3 for TLR1, TLR2, TLR5, TLR6, TLR8; n = 4 for TLR4.

Open in new tabDownload slide

TLR4 specifically recognizes bacterial LPS, and its activation leads to the production of proinflammatory cytokines. To assess the capacity of TLR4 to be activated, its prototype agonist, LPS, was used. We used IL1β, known to be produced upon the LPS-driven TLR4-activated NF-κB pathway,²⁸ as a readout to detect TLR4-mediated response. M(IL4) macrophages were generated for 6 d in NG and HG conditions, and on day 6, M(IL4) was stimulated with LPS for 24 h. RT-qPCR and ELISA were used to measure the expression and secretion of IL1β. Consistent with the biological function of LPS, stimulation of M(IL4) with LPS resulted in the upregulation of IL1β gene expression by 7.47 times in NG conditions. HG alone also stimulated IL1β expression (fold change 2.2). In HG conditions, LPS increased IL1β gene expression 13.7 times (Fig. 3A). The combination of HG and LPS had an additive effect and significantly increased IL1β gene expression (fold change 39.97, by comparison with NG conditions). The combination of HG and LPS demonstrated also an additive effect on the levels of secreted IL1β (fold change 1.5) compared to LPS alone (1.32) or HG alone (1.1) (Fig. 3B). Thus, we identified that HG cooperates with TLR4 agonists and promotes further upregulation of IL1β expression. We have previously found that HG, independently, can induce elevated production of IL1β in M(IFNγ) macrophages.¹¹ Here, for the first time, we demonstrate that HG potentiates the ligand-mediated TLR response in M(IL4) macrophages. The proinflammatory effect of HG on M(IL4) was further confirmed by the decreased levels of IL10 release in LPS-stimulated M(IL4), while IL10 was an established negative feedback loop mediator needed to suppress acute inflammation (Fig. 3C). The production of TNFα, IL12/23, and TGFβ was not statistically significantly affected by HG, suggesting the selectivity of its effect (Supplementary Fig. 3).

Fig. 3.

Hyperglycemia amplifies ligand stimulated TLR responses in M(IL4) macrophages. Monocyte-derived macrophages were cultured for 6 d in NG (5 mM) and HG (25 mM) conditions in the presence of IL4 (M(IL4)) and when stimulated with LPS or Pam3CSK4 for 24 h. (A) RT-qPCR analysis of IL1β expression. Bar graphs show mean ± SEM of n = 9 individual donors. (B) IL1β secretion analysis by ELISA in macrophage conditioned medium; n = 4 for M(IL4) without LPS and n = 6 for M(IL4) with LPS. (C) IL10 secretion analysis by ELISA in macrophage conditioned medium; n = 4. (D) RT-qPCR analysis of TLR4 expression in NG and HG in M(IL4) stimulated with PAM3CSK4 (n = 5) or with LPS (n = 7). (E) IL1β secretion analysis by ELISA in macrophage conditioned medium; n = 4 for M(IL4) without PAM3CSK4, n = 6 for M(IL4) with PAM3CSK4. (F) RT-qPCR analysis of CCR2 expression in NG and HG in M(IL4) macrophages stimulated with PAM3CSK4 (n = 6). Statistical significance of the data was analyzed using ratio paired Student's t test. (n) represents the number of individual donors.

Open in new tabDownload slide

Next, we aim to analyze how HG can cooperate with dyslipidemia on the level of TLR activation. We used PAM3CSK4 (tripalmitoylcysteinylseryl-(lysyl)4), which is a classically applied ligand to examine TLR1/TLR2 responsiveness in patients with diabetes and atherosclerosis and can mimic TLR activation in dyslipidemic conditions.^21,29,30 We applied the TLR1 agonist PAM3CSK4 in M(IL4) and measured the TLR4 expression levels. Under PAM3CSK4 stimulation, HG statistically significantly increased TLR4 expression in M(IL4) macrophages, while stimulation with LPS did not induce upregulation of TLR4 expression (Fig. 3D). This finding shows that stimulation of TLR1/TLR2 with the dyslipidemic ligand in hyperglycemic conditions can sensitize macrophages to TLR4-mediated responses already by elevation of its expression. PAM3CSK4 also significantly amplified the expression of IL1β and tended to suppress IL10 in M(IL4), thus switching on the inflammatory program in M(IL4) in HG conditions (Figs. 3C and 3E). Considering that IL1β is an essential mediator of endothelial cell inflammation, a combination of HG and dyslipidemia can create a highly pathological environment for the development of diabetic vascular complications.

CCR2, the key receptor responsible for monocyte/macrophage migration toward inflammatory chemokines, plays a critical role in the onset of numerous diabetic complications. For instance, CCR2 can promote the recruitment of macrophages, leading to diabetic cardiomyopathy.³¹ We decided to examine the effect of HG on the upregulation of CCR2 expression upon activation of TLR1 and TLR4 receptors in healing macrophages. Consequently, stimulation with LPS (TLR4 agonist) did not affect CCR2 expression. However, in hyperglycemic conditions, activation with PAM3CSK43 (TLR1 agonist) resulted in the upregulation of CCR2 gene expression in M(IL4) (Fig. 3F).

In summary, our data demonstrate that HG differentially enhances TLR expression profiles in M(NC), M(IFNγ), and M(IL4) macrophages. In M(IL4) macrophages, HG enhances the inflammatory response of M(IL4) to TLR1/2 and TLR4 agonists. HG can cooperate with dyslipidemic ligand PAM3CSK43 by amplifying the production of inflammatory cytokines responsible for vascular inflammation. In parallel, HG, together with the TLR1/2 ligand, can enhance the migratory potential of macrophages by elevating CCR2 expression.

Our findings show for the first time that HG drives an inflammatory TLR-mediated response in M(IL4) macrophages, which can lead to the initiation and progression of vascular complications in patients with metabolic syndrome and diabetes. Our study opens a new area of interest in the biology of macrophages exposed to the HG environment in the context of TLR activation. We demonstrated a clear effect of HG on the stimulation of TLR signaling, which promoted the activation of the inflammatory program in M(IL4) macrophages. The mechanism of the selective upregulation of several TLRs in differentially activated macrophages is of great interest and requires to be studied in the future. One possibility is that hyperglycemia selectively changes the profile of the histone code on the TLR promoters, similar to the epigenetic changes we have previously found for the S100A9 and S100A12 genes in human macrophages.²²

Acknowledgments

The study was supported by the Deutsche Forschungsgemeinschaft IRTG DIAMICOM 1874/2 and by Tomsk State University Development Programme (Priority-2030). We thank the technical support of the team FlowCore facility for the help in the organization of flow cytometry experiments. Quan Liu received scholarship funding from the China Scholarship Council (CSC File No. 202108320068).

Supplementary material

Supplementary materials are available at Journal of Leukocyte Biology online.

References

Abbreviations

 
• AGE

  advanced glycosylated end product

 
• CCR2

  C-C motif chemokine receptor 2

 
• HG

  hyperglycemia (25 mM)

 
• IFN

  interferon

 
• IL

  interleukin

 
• LPS

  lipopolysaccharide

 
• M-CSF

  macrophage colony-stimulating factor

 
• MFI

  mean fluorescence intensity

 
• M(NC)

  macrophages nonstimulated with cytokines

 
• M(IFNγ)

  macrophages stimulated with IFNγ

 
• M(IL4)

  macrophages stimulated with IL4

 
• NG

  normal glucose (5 mM)

 
• ROS

  reactive oxygen species

 
• TLR

  toll-like receptor

Author notes