Upper gut heat shock proteins HSP70 and GRP78 promote insulin resistance, hyperglycemia, and non-alcoholic steatohepatitis

Abstract

A high-fat diet increases the risk of insulin resistance, type-2 diabetes, and non-alcoholic steato-hepatitis. Here we identified two heat-shock proteins, Heat-Shock-Protein70 and Glucose-Regulated Protein78, which are increased in the jejunum of rats on a high-fat diet. We demonstrated a causal link between these proteins and hepatic and whole-body insulin-resistance, as well as the metabolic response to bariatric/metabolic surgery. Long-term continuous infusion of Heat-Shock-Protein70 and Glucose-Regulated Protein78 caused insulin-resistance, hyperglycemia, and non-alcoholic steato-hepatitis in rats on a chow diet, while in rats on a high-fat diet continuous infusion of monoclonal antibodies reversed these phenotypes, mimicking metabolic surgery. Infusion of these proteins or their antibodies was also associated with shifts in fecal microbiota composition. Serum levels of Heat-Shock-Protein70 and Glucose-Regulated Protein78were elevated in patients with non-alcoholic steato-hepatitis, but decreased following metabolic surgery. Understanding the intestinal regulation of metabolism may provide options to reverse metabolic diseases.

Fig. 1. Sites of intestinal heat shock proteins release and their kinetics and action.

a iTraQ-based analysis of proteins (n = 48) differentially expressed in the jejunal secretome in chow diet versus high-fat diet (n = 4 biologically independent animals). Downregulated proteins (n = 17) are colored in green while upregulated proteins (n = 31) are colored in orange. b Insulin-mediated glucose uptake of rat primary myocytes treated with HSP70 (15 ng/ml), GRP78 (3 ng/ml) alone or in combination with their respective mAbs (1 μg/ml). Rat myocytes were stimulated with 100 nM insulin for 10 min (n = 5 independent experiments). c, d Representative confocal microscopy images of FACS isolated rat enteroendocrine cells (c) stained for Claudin-4 (CLD4,red) and rat intestinal epithelial cells (d) stained for Integrin-β4 (ITB4,green) (n = 5 independent experiments). e HSP70 and GRP78 concentration in the culture medium of rat intestinal epithelial and enteroendocrine cells untreated or stimulated with 0.5 mM palmitic acid (n = 5 independent experiments). f Jejunal expression of key proteins involved in Unfolded Protein Response (UPR) in rats fed either chow diet (CD) or high-fat diet (HFD) (n = 5 biologically independent animals). g Expression of UPR proteins in rat intestinal epithelial cells untreated or stimulated with 0.5 mM palmitic acid (n = 5 independent experiments). h, i Time courses and fitting curves of recombinant (black circles) and native (red squares) HSP70 (h) and GRP78 (i) during and after a 2 h infusion of 2.5 ng/h of recombinant proteins and after an intragastric load of palm oil (n = 4 biologically independent animals). Magnification ×60. Scale bar: 50 μm Data are presented as mean value ± SEM. Statistical significances were calculated by unpaired two-tailed t-test and one-way Anova with Bonferroni’s correction for multiple comparisons, where appropriate. Source data are provided as a Source Data file.

Fig. 2. Effects of continuous infusion of HSP70/GRP78 recombinant proteins.

a Glucose disposition index (DI) metrics showing experimental data and fitting curves. Total β-cell responsivity to glucose (Φ) is a function of whole-body insulin sensitivity (SI). b, c Time courses of blood glucose and plasma insulin concentrations during an oral glucose tolerance test. d–f Hepatic phosphorylation of Akt on Ser473 (d), FoxO1 on Thr24 (e), and GSK3ab on Ser21/9 (f). g Gene expression of key enzymes involved in hepatic gluconeogenesis and glycolysis. h Representative images showing periodic acid Schiff (PAS) staining of liver section from rats fed chow diet (CD) and infused with recombinant HSP70 or GRP78 or saline solution. i–k Western blot analysis of Akt Ser473 (i) and GSK3αβ Ser21/9 (j) phosphorylation and GLUT4 (k) expression in skeletal muscle. l, m Representative images showing PAS (l) and Oil red O (ORO) (m) staining of skeletal muscle sections from rats fed CD and infused with recombinant HSP70 or GRP78 or saline solution. Magnification ×20. Scale bar: 0.10 mm. Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by Kruskal–Wallis test with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 3. Effects of continuous infusion of HSP70 or GRP78 on NASH hallmarks.

a–c Representative Hematoxylin and Eosin (H&E) (a), Oil Red O (ORO) (b), and Picro Sirius Red (c) staining of liver sections from rats under CD and infused with recombinant HSP70 or GRP78 or saline solution. d Gene expression of liver inflammation and fibrosis markers as well as key enzymes involved in hepatic de novo lipogenesis. e Plasma low-grade inflammation markers assessed by Luminex. f Gene expression of key genes involved in intestinal permeability. g Time course and AUC of in vivo intestinal permeability assessed by measuring Fluorescein IsoThioCyanate Dextran 4000 Da (FITC-DXT) in the systemic circulation after an oral gavage with FITC-DXT (500 mg/kg body weight). h–j Hepatic (h), skeletal muscle (i), and jejunal (j) expression of key proteins involved in Unfolded Protein Response (UPR). Magnification ×20. Scale bar: 0.10 mm. Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by one-way Anova with Bonferroni’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 4. Effects of native HSP70/GRP78 inhibition with mAbs on insulin sensitivity.

a Glucose disposition index (DI) metrics showing experimental data and fitting curves. Total β-cell responsivity to glucose (Φ) is a function of whole-body insulin sensitivity (SI). b, c Time courses of blood glucose and plasma insulin concentrations during an oral glucose tolerance test. d–f Hepatic phosphorylation of Akt on Ser473 (d), FoxO1 on Thr24 (e), and GSK3ab on Ser21/9 (f). g Gene expression of key enzymes involved in hepatic gluconeogenesis and glycolysis. h Representative images showing periodic acid Schiff (PAS) staining of liver section from rats fed HFD and infused with monoclonal antibodies against HSP70 or GRP78 or antibody isotype. i–k Western blot analysis of Akt Ser473 (i) and GSK3αβ Ser21/9 (j) phosphorylation and GLUT4 (k) expression in skeletal muscle. l, m Representative images showing PAS (l) and Oil red O (ORO) (m) staining of skeletal muscle sections from rats fed HFD and infused with monoclonal antibodies against HSP70 or GRP78 or antibody isotype. Magnification ×20. Scale bar: 0.10 mm. Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by Kruskal–Wallis test with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 5. Effects of native HSP70/GRP78 inhibition with mAbs on NASH hallmarks.

a–c Representative Hematoxylin and Eosin (H&E) (a), Oil Red O (ORO) (b), and Picro Sirius Red (c) staining of liver sections from rats fed HFD and infused with monoclonal antibodies against HSP70 or GRP78 or antibody isotype. d Gene expression of liver inflammation and fibrosis markers as well as key enzymes involved in hepatic de novo lipogenesis. e Plasma low-grade inflammation markers assessed by Luminex. f Gene expression of key genes involved in intestinal permeability. g Time course and AUC of in vivo intestinal permeability assessed by measuring Fluorescein IsoThioCyanate Dextran 4000 Da (FITC-DXT) in the systemic circulation after an oral gavage of FITC-DXT (500 mg/kg body weight). h–j, Hepatic (h), skeletal muscle (i), and jejunal (j) expression of key proteins involved in Unfolded Protein Response (UPR). Magnification ×20. Scale bar: 0.10 mm. Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by one-way Anova with Bonferroni’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 6. Effects of HSP70 or GRP78 infusion in Duodenal-Jejunal Bypass (DJB) on insulin sensitivity.

a Glucose disposition index (DI) metrics showing experimental data and fitting curves. Total β-cell responsivity to glucose (Φ) is a function of whole-body insulin sensitivity (SI). b, c Time courses of blood glucose and plasma insulin concentrations during an oral glucose tolerance test. d–f Hepatic phosphorylation of Akt on Ser473 (d), FoxO1 on Thr24 (e), and GSK3ab on Ser21/9 (f). g Gene expression of key enzymes involved in hepatic gluconeogenesis and glycolysis. h Representative images showing periodic acid Schiff (PAS) staining of liver section from rats fed HFD that underwent Sham-operation or DJB and infused with recombinant HSP70 or GRP78 or saline solution. i–k Western blot analysis of Akt Ser473 (i) and GSK3αβ Ser21/9 (j) phosphorylation and GLUT4 (k) expression in skeletal muscle samples. l, m Representative images showing PAS (l) and (ORO) (m) staining of skeletal muscle sections from rats fed HFD that underwent Sham-operation or DJB and infused with recombinant HSP70 or GRP78 or saline solution. Magnification ×20. Scale bar: 0.10 mm Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by Kruskal–Wallis test with Dunnett’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 7. Effects of the infusion of HSP70 and GRP78 in Duodenal-Jejunal Bypass (DJB) on NASH hallmarks.

a–c Representative Hematoxylin and Eosin (H&E) (a), Oil Red O (ORO) (b), and Picro Sirius Red (c) staining of liver sections from rats fed HFD that underwent Sham-operation (Sham-op) or DJB and infused with recombinant HSP70 or GRP78 or saline solution. d Gene expression of liver inflammation and fibrosis markers as well as key enzymes involved in hepatic de novo lipogenesis. e Plasma low-grade inflammation markers assessed by Luminex. f Gene expression of key genes involved in intestinal permeability. g Time course and AUC of in vivo intestinal permeability assessed by measuring Fluorescein IsoThioCyanate Dextran 4000 Da (FITC-DXT) in the systemic circulation after an oral gavage of FITC-DXT (500 mg/kg body weight). h, i Hepatic (h) and skeletal muscle (i) expression of key proteins involved in Unfolded Protein Response (UPR). Magnification ×20. Scale bar: 0.10 mm. Data are presented as mean value ± SEM of n = 10 biologically independent animals. Statistical significances were calculated by one-way Anova with Bonferroni’s correction for multiple comparisons. Source data are provided as a Source Data file.

Fig. 8. Human studies.

a, b Fasting serum levels of HSP70 (a) and of GRP78 (b) in people with Non-Alcoholic Steato-Hepatitis (NASH) (n = 47 biologically independent samples) and in healthy participants (n = 39 biologically independent samples). c, d Serum levels of HSP70 (c) and of GRP78 (d) before and after Roux-en-Y Gastric Bypass (RYGB), Biliopancreatic diversion (BPD) (n = 15 biologically independent samples), Vertical Sleeve Gastrectomy (VSG) and Low-Fat Diet (LFD) (n = 10 biologically independent samples). e, f Serum HSP70 (e) and GRP78 (f) time course after a high-fat meal test in healthy participants and participants with obesity (n = 5 biologically independent samples). Data are presented as mean value ± SEM. Statistical significances were calculated by unpaired two-tailed t-test and one-way Anova with Bonferroni’s correction for multiple comparisons, where appropriate. Source data are provided as a Source Data file.