Fig 1.
HSPG core proteins and HS rapidly decline in the pancreatic islets of db/db mice.
(A) Representative images show the distribution of HSPG core proteins and HS in the islets of male db/db mice and wt mice at 4 and/or 6 weeks of age as demonstrated by immunohistochemical staining. (B-E) Bar graphs for each time point show morphometric analysis of the % islet stained for the HSPG core proteins (B) COL18, (C) SDC1, (D) CD44 and HS (E) in pancreases of lean control male mice (wt, db/+; open bars) and db/db mice (shaded bars). Data show mean ± SEM for 3–6 pancreases/age group with n = 21–72 islets examined/group for HSPG core proteins and n = 19–52 islets/group for HS. *p<0.05, **p<0.01 and ***p<0.0001, Mann-Whitney test (HSPGs) and Unpaired t test (HS). Scale bar = 100 μm.
Fig 2.
Intra-islet COL18 core protein colocalizes with insulin not glucagon staining in wt pancreas and is progressively lost in db/db pancreas.
Immunofluorescence staining of (A, F, K,) COL18, (B, G, L) insulin (INS), (C, H, M) insulin and COL18 merged, (D, I, N) glucagon (GLUC) and COL18 merged in 9 wk wt (A-D), 9 wk db/db (F-I) and 4 wk db/db (K-N) pancreas. (E, J, O) Nuclei were stained with DAPI. Scale bar = 20 μm.
Fig 3.
ER stress markers in the islets of male db/db mice at different ages.
Relative transcript levels for ER stress genes (Bip, p58, Chop and Atf3) in islets of (A) young (4.0–4.4 week) lean controls (db/+; open bars) and normoglycemic db/db mice (shaded bars) and (B) 5–8 week lean controls (wt; open bars/~baseline) and db/db mice (hyperglycemic; shaded bars). The values were normalized to the house-keeping gene Gapdh. The data represent mean ± SEM for 4 independent experiments (120–743 islets/group/experiment). *p<0.05, Mann-Whitney test.
Fig 4.
ER stress, intra-islet HSPG core protein and HS expression in Ins2WT/C96Y islets.
Fold change in the expression of ER stress-related genes (A) Bip, p58 and (B) Chop, Atf3 in 5–12 week male Ins2WT/WT control islets (open circles) and Ins2WT/C96Y islets (black squares) was normalized to the house-keeping gene Gapdh. Morphometric analysis of immunostained intra-islet HSPG core proteins and HS in Ins2WT/WT (open bars) and Ins2WT/C96Y (shaded bars) pancreases at 4, 5, 6 and 9 weeks (HSPGs) and 6 weeks (HS). Data is presented as the % islet area with positive staining for (C) COL18, (D) SDC1, (E) CD44 and (F) HS. Data represent mean ± SEM for (A-B) n = 4 independent experiments (250–350 islets/group/experiment) and (C-F) n = 3 pancreases/age group with n = 15–31 islets examined/group; each symbol denotes an individual islet. *p<0.05, **p<0.001 and ***p<0.0001, Mann-Whitney test.
Fig 5.
Intracellular levels of HSPG core proteins, HS and HPSE are decreased in ER stressed MIN6 cells.
(A) MIN6 cells were cultured with 6 mM glucose (open bars), 50 nM thapsigargin (striped bars) or 2 μM tunicamycin (grey bars) for 20–24 hours. Transcript levels for Bip, p58, Chop and Atf3 were normalized to the house-keeping gene Ube2d1 and were quantified as a fold change compared to 6 mM control mRNA levels which were assigned a value of 1. (B-F) MIN6 cells were cultured for 3 days with 6 mM glucose-DMEM (control cells, C), 50 nM thapsigargin (Tg) or 2 μM tunicamycin (Tm) and intracellular HSPG core protein (B) COL18, (C) SDC1 and (D) CD44, (E) HS and (F) HPSE levels were assessed by flow cytometry and expressed as GMFI. Data represent mean ± SEM for (A) n = 5 independent experiments and (B-F) n = 11 experiments. *p<0.05, **p<0.01 and ***p<0.001, One-way ANOVA with Fischer’s unprotected LSD test.
Fig 6.
Intracellular expression of HSPGs, HS and HPSE in wt and db/db isolated islet cells.
Intracellular levels of (A, D) HSPG core proteins COL18, SDC1 and CD44, (B, E) HS and (C, F) HPSE in wt (open bars) and db/db (shaded bars) isolated islet cells on day 0 and after culture for 2 days (day 2). Data show GMFI ± SEM, n = 7 experiments/group with n = 2–4 male donors/experiment. *p<0.05, **p<0.01 and ***p<0.001, compared to corresponding control wt cells, Mann-Whitney test.
Fig 7.
HS replacement protects db/db beta cells from dying in culture and from acute hydrogen peroxide-induced damage.
The viability of (A) wt and (C, E, G) db/db beta cells was analyzed by flow cytometry using the fluorescent dyes Calcein (Cal) and PI after culture for 2 days without (open bars) or with (shaded bars) the highly sulfated HS analogue heparin (50 μg/ml). Cell death/damage due to acute treatment with hydrogen peroxide was measured by uptake of the fluorescent dye Sytox Green. Donors were (A, B) wt; (C, D) normoglycemic db/db (bg<10 mmol/l); (E, F) mildly hyperglycemic db/db (bg = 10–15 mmol/l) and (G, H) severely hyperglycemic db/db (bg>15 mmol/l). Beta cells were identified as viable (Cal+PI-), damaged (Cal+PI+) or dead (Cal-PI+) and expressed as % total cells; alternatively, damaged/dead cells were identified as Sytox Green+ve. Data represent mean ± SEM for n = 3–11 experiments/group; n = 2–4 male donors/experiment. *p<0.05, **p<0.01 and ***p<0.0001, ANOVA with Fisher’s unprotected LSD post-test.
Fig 8.
Effect of TUDCA treatment of db/db mice on glycemic control, ER stress-related gene expression and intra-islet COL18 and HS.
4 week old male db/db mice were treated with saline (black line) or TUDCA (150 mg/kg/day; broken line) i.p. for 28 days. (A) The non-fasting blood glucose levels and (B) body weight were monitored 3x/week. (C) HbA1c levels were measured at termination of treatment. (D) ER stress-related gene transcripts were analyzed in islets isolated from saline-treated (open bars) and TUDCA-treated (shaded bars) db/db mice. Fold-change refers to mRNA expression relative to gene expression in wt kidney (broken line) which was assigned a value of 1 and normalized to the house-keeping gene Gapdh. The in situ expression of (E) HSPG core protein (COL18), (F) HS and (G) insulin was determined by immunohistochemistry and evaluated as % islet area stained. (A-B) Data represent mean weekly measurements ± SEM with n = 12–14 male mice/group, n = 3 measurements/mouse; (C) mean ± SEM for n = 12–13 mice/group; (D) mean ± SEM, each data point represents an independent experiment, n = 4 mice/group (100–140 islets/donor); (E-G) n = 4–9 pancreases/group with n = 86–88 (COL18), n = 38–44 (HS) and n = 92–105 (insulin) islets examined/group. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001, Mann-Whitney test.
Fig 9.
TUDCA treatment restores the colocalization of intra-islet COL18 core protein and insulin in db/db pancreas.
Immunofluorescence staining of (A, F) COL18, (B, G) insulin (INS), (C, H) insulin and COL18 merged, (D, I) glucagon (GLUC) and COL18 merged in (A-D) saline-treated control and (F-I) TUDCA-treated db/db pancreases. (E, J) DAPI was used to identify nuclei. Scale bar = 20 μm.
Fig 10.
The proposed model: ER stress inhibits the synthesis of HSPG core proteins and HS in T2D-prone beta cells.
In pre-T2D pancreatic beta cells, ER stress initiates the unfolded protein response (UPR) which acts to alleviate the stress by increasing molecular chaperones for protein folding, increasing ERAD and reducing the translation of mRNAs to a range of “general” proteins. As a consequence, HSPG core protein synthesis is impaired, which in turn severely reduces the synthesis of HS, a constitutive non-enzymatic antioxidant in beta cells. Depletion of beta cell HS increases intracellular ROS and elevates oxidative stress. Oxidative stress induces the expression of secondary antioxidant enzymes to help maintain beta cell homeostasis [57]. Failure to compensate for this stress results in UPR failure, the preferential expression of apoptotic genes and beta cell apoptosis/death. ERAD, ER-associated degradation; T2D, Type 2 diabetes; HS, heparan sulfate; HSPG, heparan sulfate proteoglycan; ROS, reactive oxygen species.