Homeostatic and pathogenic roles of GM3 ganglioside molecular species in TLR4 signaling in obesity

Abstract

Innate immune signaling via TLR4 plays critical roles in pathogenesis of metabolic disorders, but the contribution of different lipid species to metabolic disorders and inflammatory diseases is less clear. GM3 ganglioside in human serum is composed of a variety of fatty acids, including long-chain (LCFA) and very-long-chain (VLCFA). Analysis of circulating levels of human serum GM3 species from patients at different stages of insulin resistance and chronic inflammation reveals that levels of VLCFA-GM3 increase significantly in metabolic disorders, while LCFA-GM3 serum levels decrease. Specific GM3 species also correlates with disease symptoms. VLCFA-GM3 levels increase in the adipose tissue of obese mice, and this is blocked in TLR4-mutant mice. In cultured monocytes, GM3 by itself has no effect on TLR4 activation; however, VLCFA-GM3 synergistically and selectively enhances TLR4 activation by LPS/HMGB1, while LCFA-GM3 and unsaturated VLCFA-GM3 suppresses TLR4 activation. GM3 interacts with the extracellular region of TLR4/MD2 complex to modulate dimerization/oligomerization. Ligand-molecular docking analysis supports that VLCFA-GM3 and LCFA-GM3 act as agonist and antagonist of TLR4 activity, respectively, by differentially binding to the hydrophobic pocket of MD2. Our findings suggest that VLCFA-GM3 is a risk factor for TLR4-mediated disease progression.

Keywords: TLR4; chronic inflammation; ganglioside GM3; inflammation amplification loop; obesity.

Figure 1. Molecular species of ganglioside GM3 in human serum and the acyl‐chain structures

1. Biosynthetic pathway (schematic) of GM3, from ceramide, and to complex gangliosides.

2. TLC analysis of ganglioside species in human serum.

3. Quantification by densitometry of major ganglioside species GM3 and GD1a in human serum. Data expressed as mean ± SD, n = 6.

4. Detailed structures of GM3 species: sialyllactose head group, sphingoid base (d18:1), typical fatty‐acid lengths (LCFA, VLCFA), and acyl‐chain modifications (α‐hydroxylation, ω‐9 desaturation).

5. Quantification by LC‐MS/MS analysis of serum GM3 species with differing acyl‐chain structures. Total abundance of 10 species was defined as 1. Data expressed as mean ± SD, n = 6.

Figure 2. Alterations of relative abundance of GM3 species are involved in disease progression and chronic inflammation

1. A

   Heat map analysis of serum GM3 species in various pathological phases: control (n = 24), VFA (n = 38), lipidemia (n = 28), glycemia (n = 15), and lipidemia + glycemia (n = 17). Colors indicate fold change average of each species relative to control (defined as 1), as shown in key at right. Order of pathological phases corresponds to increments of HOMA‐IR and serum CRP.

2. B–D

   Properties of various GM3 species as a function of pathological phases: control (n = 24), VFA (n = 38), lipidemia (n = 28), glycemia (n = 15), and lipidemia + glycemia (n = 17). Data shown are relative abundances of total LCFA species (16:0, 18:0, 20:0) (B), total VLCFA species (22:0, 23:0, 24:0, h24:0) (C), and total unsaturated VLCFA species (22:1, 24:1, h24:1) (D) relative to total of ten major GM3 species (defined as 1) in each subject.

3. E–H

   Properties of various GM3 species as a function of BMI: LCFA‐GM3 (E), VLCFA‐GM3 (F), unsaturated VLCFA‐GM3 (G), and α‐hydroxy VLCFA‐GM3 (h24:0) (H). Colors indicate disease severity: light blue, no abnormal scores (n = 25); orange, early‐phase obesity (n = 74); purple, severe obesity (n = 23).

4. I, J

   Spearman's correlations for GM3 h24:0 vs. ALT (I) and vs. HOMA‐IR (J).

5. K

   Plots of α‐hydroxylation rate (h24:0/24:0) vs. serum CRP. Colors indicate range of CRP value (mg/dl): light blue, 0.01–0.02 (n = 21); orange, 0.03–0.09 (n = 56); gray, 0.10–0.29 (n = 29); red, 0.3–1.0 (diagnostically abnormal; n = 15).

6. L

   Association between serum GM3 species and progression of metabolic disorders (schematic).

Data information: Data shown are individual values and mean ± SD, analyzed by two‐tailed unpaired t‐test with Bonferroni's correction. *P < 0.05, **P < 0.01, and ***P < 0.001 for comparisons between indicated groups.

Figure EV1. Properties of various GM3 species as a function of clinical markers of metabolic disorders and chronic inflammation

1. A–D

   LCFA species (A), VLCFA species (B), unsaturated VLCFA species (C), and α‐hydroxy VLCFA‐GM3 (h24:0) (D). Colors indicate disease severity: light blue, no abnormal scores (n = 17); orange, early‐phase obesity (n = 80); purple, severe obesity (n = 25). Data shown are mean ± SD, analyzed by two‐tailed unpaired t‐test with Bonferroni's correction. *P < 0.05, **P < 0.01, and ***P < 0.001 for comparisons between indicated groups.

2. E–J

   Spearman's correlations for GM3 h24:0 vs. BMI (E), GM3 h24:0 vs. abdominal circumference (F), total of α‐hydroxy GM3 (h24:0 and h24:1) vs. HOMA‐IR (G), GM3 h24:0 vs. serum CRP (H), α‐hydroxylation rate (h24:0 to 24:0) vs. serum CRP (I), and α‐hydroxylation rate (h24:0 and h24:1 to 24:0 and 24:1) vs. serum CRP (J).

Data information: Sample sizes: (A–G), n = 122; (H‐J), n = 121.

Figure 3. Self‐organization map (SOM) analysis based on relative abundances of serum GM3 species

1. Procedure (schematic) for self‐organization map (SOM) analysis, a pattern recognition method using neural‐network‐type artificial intelligence. Complex patterns of multi‐dimensional information (in this case, expression patterns of the ten major GM3 species in human subjects) are mapped onto a 2D surface. Subjects having similar GM3 patterns are located proximal to each other and form several clusters (red arrows), whereas subjects having different GM3 patterns are located distal to each other (blue arrows).

2. SOM analysis of control and lipidemia subjects based on expression patterns of ten GM3 species.

3. Mapping of expression levels of ten GM3 species onto SOM in (B).

4. Identification of sub‐clusters having different GM3 patterns based on SOM in (B).

5. Metabolic pathways for fatty acids: elongation, desaturation, and α‐hydroxylation (α‐oxidation) (schematic). Sub‐clusters identified by SOM analysis are mapped on metabolic pathways.

6. Heat map analysis for GM3 species and clinical markers of six clusters. Sample sizes: sub‐clusters 1–3 (total), n = 22; cluster 4, n = 7; cluster 5, n = 9; cluster 6, n = 12.

7. Self‐organization map (SOM) based on four GM3 species as indicated at bottom.

8. ROC curve derived from Bayesian regularized neural networks (BRNNs) based on four GM3 species in (G).

Figure 4. Positive and negative regulation of innate immune responses by serum GM3 species in an acyl‐chain‐dependent manner

1. A

   Profiling of bioactivities of serum GM3 species in LPS‐mediated monocyte activation (schematic).

2. B, C

   GM3‐mediated enhancement and inhibition of proinflammatory cytokine production from LPS‐stimulated monocytes (LPS: 0.06, 0.13, 0.25 ng/ml). TNF‐α (B) production and IL‐6 (C) production in culture supernatant were measured by ELISA.

3. D, E

   Co‐stimulation of monocytes by LPS plus GM3 species or complex ganglioside species (1.5, 3.0, 4.5 μM). TNF‐α (D) production and IL‐6 (E) production were shown in heat maps.

4. F

   Co‐stimulation of monocytes by LPS plus GM3 species or precursor GSL species. TNF‐α production, IL‐6 production, and IL‐12/23 production were shown in heat maps.

5. G

   Inhibitory effect of LCFA and unsaturated VLCFA‐GM3 on VLCFA‐GM3 species. IL‐6 production in culture supernatant was measured by ELISA.

Data information: Data shown are mean ± SD (n = 3), analyzed by Tukey's multiple comparison test. **P < 0.01 for comparisons between indicated groups.

Figure EV2. Positive and negative regulation of innate immune response by GM3 gangliosides

1. A

   GM3‐mediated enhancement and inhibition of IL‐12/23 production from LPS‐stimulated monocytes (measured by ELISA).

2. B

   Co‐stimulation of monocytes by LPS plus GM3 species or complex ganglioside species. IL‐6 production in culture supernatant was measured by ELISA.

3. C–E

   Co‐stimulation of monocytes by LPS plus GM3 species or precursor GSL species. The production of IL‐6 (C), TNF‐α (D), and IL‐12/23 p40 (E) in culture supernatant was measured by ELISA.

4. F, G

   Inhibitory effect of LCFA and unsaturated VLCFA‐GM3 on VLCFA‐GM3 species. The production of TNF‐α (F) and IL‐12/23 p40 (G) in culture supernatant was measured by ELISA.

Data information: Data shown are mean ± SD (n = 3) analyzed by Tukey's multiple comparison test. **P < 0.01 for comparison with LPS stimulation without GM3 species (A), or co‐stimulation by LPS and proinflammatory GM3 (22:0, 24:0) (F, G).

Figure 5. VLCFA‐GM3 species synergistically and selectively control human TLR4/MD‐2 activation

1. A

   Co‐stimulation of human monocytes by GM3 species plus various TLR ligands: LPS (0.13 ng/ml), TLR4/MD‐2, Pam3CSK4 (0.5 μg/ml), TLR1/2, Flagellin (50 ng/ml), TLR5, R848 (0.5 μg/ml), TLR7/8, MALP‐2 (1.0 ng/ml), and TLR2/6. IL‐6 production in culture supernatant was quantified by ELISA (shown in a heat map).

2. B

   Co‐stimulation of monocytes by GM3 species (16:0, 24:0, 24:1) plus synthetic TLR4 ligands LA506 (15 ng/ml), LA505 (150 ng/ml), or LA504 (150 ng/ml).

3. C, D

   Production of IL‐6 (C) and TNF‐α (D) in culture supernatant following co‐stimulation of monocytes by GM3 species plus human HMGB1.

4. E

   Overexpression of hTLR4, hTLR4/hMD‐2, and hTLR4 (P714H) /MD‐2 in HEK293T cells, and co‐stimulation by GM3 22:0 with LPS (5 ng/ml). TLR4 activation was monitored by NF‐κB luciferase reporter assay (termed “Relative Luc Activity” on y‐axis).

5. F, G

   Co‐stimulation of hTLR4/hMD‐2 by GM3 species plus LPS (5 ng/ml) (F) and further addition of soluble human CD14 (1 μg/ml) (G).

6. H

   Stimulation of Mal‐overexpressing HEK293T cells by GM3 species.

7. I

   Co‐stimulation of hTLR4/hMD‐2 by LPS (5 ng/ml) plus various mixtures of GM3 species.

8. J

   Regulation of hTLR4/hMD‐2 by GM3 species balance (schematic).

Data information: Data shown are mean ± SD (A–D and F‐I, n = 3; E, n = 4) analyzed by Tukey's multiple comparison test. *P < 0.05 and **P < 0.01 for comparisons between indicated groups.

Figure EV3. VLCFA‐GM3 species synergistically and selectively control human TLR4/MD‐2 activation

1. A

   Co‐stimulation of human monocytes by GM3 species plus various TLR ligands: LPS (0.13, 0.25 ng/ml), TLR4/MD‐2, Pam3CSK4 (0.25, 0.5 μg/ml), TLR1/2, Flagellin (10, 50 ng/ml), TLR5, R848 (0.25, 0.5 μg/ml), TLR7/8, MALP‐2 (0.5, 1.0 ng/ml), and TLR2/6. IL‐6 production in culture supernatant was quantified by ELISA.

2. B–D

   Production of IL‐6 (B), TNF‐α (C), and IL‐12/23 p40 (D) in culture supernatant following co‐stimulation of monocytes by GM3 species plus LA506 (synthetic TLR4 ligand) (3, 10, 30 ng/ml).

3. E

   Relative luciferase reporter activities of NF‐κB, AP‐1, and ISRE in response to LPS (+, 5 ng/ml; ++, 1 μg/ml), sCD14 (1 μg/ml), GM3 22:0 (5 μM), and their combinations. Relative luciferase activity of control condition was defined as 1 for every reporter gene.

Data information: Data shown are mean ± SD (n = 3, A–D; n = 4, E) analyzed by Tukey's multiple comparison test (A) or by two‐tailed unpaired t‐test (B–D). *P < 0.05 and **P < 0.01 for comparisons between indicated groups.

Figure 6. GM3 species selectively modulate mouse TLR4/MD‐2 signaling

1. A

   Co‐stimulation of RAW macrophages by GM3 species plus various TLR ligands: LPS (0.5, 1.0 ng/ml), bovine thymus HMGB1 (0.25 μg/ml), Pam3CSK4 (50 ng/ml), Poly I:C (10 μg/ml), R848 (4 ng/ml), and CpG type B (20 nM). TNF‐α production in culture supernatant was quantified by ELISA.

2. B

   Co‐stimulation of RAW macrophages by low‐ and high‐dose LPS (0, 5, 50 ng/ml) plus GM3 22:0 (5 μM). Time course of TNF‐α production in culture supernatant was quantified by ELISA.

3. C

   Co‐stimulation of BMDMs from C3H/HeN or C3H/HeJ mice by GM3 species plus LPS (0.5 ng/ml), and TNF‐α production in culture supernatant.

4. D

   Co‐stimulation of mTLR4/mMD‐2‐expressing HEK293T cells by GM3 species plus LPS (2.5 ng/ml), and further addition of soluble mouse CD14‐Fc fusion protein (1 μg/ml).

5. E, F

   Co‐stimulation of BMDMs from C3H/HeN mice by LPS plus GM3 species and complex ganglioside species (E; 2.5, 5.0, 10 μM), or by LPS plus GM3 species and precursor GSL species (F; 2.5, 5.0, 10 μM) (shown in heat maps).

6. G

   Structure–bioactivity relationships of GM3 species with human or mouse TLR4.

Data information: Data shown are mean ± SD (A and B, n = 3; C, E, and F, n = 4; D, n = 6) analyzed by Tukey's multiple comparison test (A, C, and D) or two‐tailed unpaired t‐test (B). *P < 0.05 and **P < 0.01 for comparison with stimulation by TLR ligand without GM3 species.

Figure EV4. GM3 ganglioside in adipose tissue showed increased abundance in early‐phase obesity and short‐term HFD

1. Body weight, blood glucose, and epididymal fat pad weight of 6‐week‐old control C57/BL6 mice and ob/ob mice (n = 3).

2. Ganglioside species in epididymal fat pad were analyzed by TLC.

3. Body weight, weight gain, blood glucose, and epididymal fat pad weight of normal diet (ND) and high‐fat diet (HFD) C57/BL6 mice (n = 4).

4. Ganglioside species in epididymal fat pad were analyzed by TLC.

5. Body weight, blood glucose, and epididymal fat pad weight of C3H/HeN mice (ND, HFD) and C3H/HeJ mice (ND, HFD) (n = 4).

6. Full‐size images of Fig 7C. TLC analysis of acidic GSL fraction (equivalent to 0.1 mg protein) from epididymal fat pads of C3H/HeN and C3H/HeJ mice on ND or HFD for 8 weeks.

Data information: Data shown are mean ± SD analyzed by two‐tailed unpaired t‐test (A, C) or by Tukey's multiple comparison test (E). *P < 0.05 and **P < 0.01 for comparisons between indicated groups.

Figure 7. α‐hydroxy VLCFA‐GM3 species in adipose tissue showed increased abundance in obesity

1. GM3 molecular species of 6‐week‐old control C57/BL6 mice and ob/ob mice were analyzed, respectively, by LC‐MS/MS (n = 3).

2. GM3 molecular species of normal diet (ND) and high‐fat diet (HFD) C57/BL6 mice were analyzed by LC‐MS/MS (n = 4).

3. TLC analysis of acidic GSL fraction (equivalent to 0.1 mg protein) from epididymal fat pads of C3H/HeN (A) and C3H/HeJ mice (B) on ND or HFD for 8 weeks.

4. GM3 molecular species of C3H/HeN mice (ND, HFD) and C3H/HeJ mice (ND, HFD) were analyzed by LC‐MS/MS (n = 4).

5. Comparison of increased GM3 species in human serum and mouse adipose tissue.

6. Feedback loop mediated by TLR4 and GM3 species, promoting disease progression (schematic).

Data information: Data shown are mean ± SD analyzed by two‐tailed unpaired t‐test (A, B) or by Tukey's multiple comparison test (D). *P < 0.05 and **P < 0.01 for comparisons between indicated groups.

Figure 8. GM3 recognition by TLR4/MD‐2 induces receptor oligomerization

1. A, B

   Surface electrostatic potentials of reported crystal structure of human TLR4/MD‐2/LPS complex (3FXI), and mapping of putative GM3‐binding pocket (A). Candidate basic residues and a hydrophobic pocket recognizing sialic acid and ceramide structure of GM3 are indicated (schematic) (B).

2. C, D

   Alanine scanning for basic residues involved in signal transduction via VLCFA‐GM3 (n = 5) (C), and combinations of effective mutations (n = 6) (D). Signal transduction was monitored by NF‐κB reporter assay.

3. E–G

   Comparative inhibitory effects of GM3 16:0 on mTLR4/mMD‐2 (E), mTLR4/hMD‐2 (domain‐swapped complex) (F), and hTLR4/hMD‐2 (G) (n = 3).

4. H

   Cross‐linked SDS–PAGE analysis of recombinant mTLR4 (extracellular domain)/mMD‐2 complexed with GM3 18:0, GM3 species enhancing mTLR4 activation.

Data information: Data shown are mean ± SD analyzed by Tukey's multiple comparison test. *P < 0.05, **P < 0.01 for comparisons between indicated groups.

Figure 9. Ligand‐macromolecular docking analysis implicates species‐specific GM3 recognition by TLR4/MD‐2

1. A, B

   Docking model of GM3 24:0 (A) and 16:0 (B) binding to human TLR4/MD‐2 complex (3FXI). Basic residues of TLR4 are colored in blue.

2. C–E

   Superposition of GM3 24:0 (in docking model) vs. Ra‐LPS (in 3FXI) (C), GM3 24:0 vs. GM3 16:0 (in docking model) (D), and GM3 16:0 vs. lipid IVa (in 2E59) (E). Basic residues and the dimer interface are indicated schematically.

3. F, G

   Working model for hTLR4 activation enhanced by VLCFA‐GM3 species (F) and reduced by LCFA‐GM3 (G). Basic residues contributing to GM3 recognition are colored in blue. Residues of dimer interface are colored in red.

Figure EV5. Binding model of VLCFA/LCFA‐GM3 species on hTLR4/hMD‐2 and the comparison to LPS and lipid IVa

1. A–C

   Docking model of GM3 24:0 (A), Ra‐LPS (in 3FXI) (B), and superposition of GM3 24:0 vs. core structure of Ra‐LPS (lipid A) (C).

2. D–F

   Docking model of GM3 16:0 (D), lipid IVa (in complex with hMD‐2 in 2E59) (E), and superposition of GM3 16:0 vs. lipid IVa (F). Basic residues of hTLR4 are colored in blue.