Inhibition of HMGB1 Promotes Osseointegration under Hyperglycemic Condition through Improvement of BMSC Dysfunction

Abstract

High mobility group box 1 (HMGB1) participates actively in oxidative stress damage and the latter relates closely to diabetic complications, including poor implant osseointegration. This article is aimed at investigating the effects of HMGB1 on dysfunction of bone marrow stromal cells (BMSCs) and impaired osseointegration under diabetic environment. In vitro, BMSCs were treated with normal glucose (NG), high glucose (HG), and HG+glycyrrhizin (HMGB1 inhibitor, HG+GL). Cell proliferation, osteogenic behaviors, and oxidative stress were determined. In vivo, 8-week-old Sprague-Dawley rats were categorized to control, streptozotocin-induced diabetic, and diabetic-GL groups. Rats received GL (50 mg/kg, i.p.) or vehicle treatment daily after titanium implants were planted into the tibiae. After 4 and 8 weeks, plasma lipoperoxide detection, μCT analysis, and histomorphometric evaluation were conducted. By these approaches, we demonstrated that inhibiting HMGB1 by GL significantly attenuated HG-induced upregulation of HMGB1, HMGB1 ligand receptor for advanced glycation end products (RAGE) and their interaction, relieved oxidative stress, and reversed the downregulation of osteogenic markers, resulting in improved osteogenic differentiation. In diabetic rats, GL administration suppressed the upregulation of HMGB1, attenuated the lipoperoxide, and ameliorated the impaired trabecular structure and osseointegration. Taken together, inhibiting HMGB1 can be an effective approach to relieve BMSC dysfunction and enhance osseointegration under diabetic environment.

Figure 1

The design of in vivo study. (a) Time schedule of in vivo study. (b) Design graph of the screw-shaped titanium implant. (c) Implant in the tibia. (d, e) VOI of the peri-implant trabecular for μCT evaluation. VOI was defined as a hollow cylinder from 1.0 mm below the tibia cortex to 150 slices towards the bone marrow, extending with a radius of 350 μm from the implant body surface, which is 150 μm from the threads, between the green and purple circles in the graph.

Figure 2

GL suppressed HMGB1 upregulation in HG-treated BMSCs. (a) Relative gene expression of HMGB1 at 3 d, 7 d, and 14 d in different groups. Representative band (b) and quantification (c) of protein expression of HMGB1 after treatment for 14 days. All data were normalized to β-actin. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001; n = 3.

Figure 3

Inhibiting HMGB1 by GL relieved BMSC dysfunction under HG condition. (a) Cell viability evaluated by the CCK-8 assay at 3 d, 7 d, and 14 d, n = 6. (b) ALP staining of BMSCs after an osteogenic induction of 3 d, 7 d, and 14 d. (c) Alizarin red staining of BMSCs after an osteogenic induction of 21 days and deposited calcium stains red. (d) Quantification of mineralization nodules in different groups, n = 3. Relative expressions of Runx2 (e), ALP (f), OCN (g), and RANKL/OPG (h) detected at 3 d, 7 d, and 14 d by RT-qPCR analysis, n = 3. Representative band (i) and quantification (j) of protein levels of OPG and RANKL at 14 d, n = 3. All data were normalized to β-actin. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001.

Figure 4

HMGB1-RAGE interaction is involved in ROS accumulation. Representative images (a) and quantification of fluorescence intensity (b) of DCFA staining after an incubation of 72 h. ROS stains are green while nuclei stain is blue. Scale bar = 200 μm. Relative expression of mRNA (c) and protein level (d, e) of HO-1. Changes of mRNA (f) and protein level (g, h) of RAGE. All data were normalized to β-actin. HMGB1-RAGE binding activity (i, j) was determined by performing coimmunoprecipitation assay after incubation for 3 days. n = 3. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001.

Figure 5

HMGB1 upregulated in diabetic rats. (a) Plasma concentration of HMGB1, n = 6. (b) Immunochemistry evaluation of HMGB1 in trabecular around the implant. Scale bar = 100 μm. Representative band (c) and quantification (d) of HMGB1 protein levels of rat bone tissue, n = 3. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001.

Figure 6

Inhibiting HMGB1 relieved oxidative stress in diabetic rats. Plasma concentration of glucose (a), MDA (b), GSH-PX (c), and SOD (d), n = 6. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001.

Figure 7

Inhibiting HMGB1 rescued impaired osseointegration in diabetic rats. (a) Quantification of BV/TV, Tb.N, Tb.Th, and Tb.Sp within the VOI. (b) Morphology of peri-implant trabeculae. Four weeks after implantation, the trabecular thickness and number were quite similar in three groups. After another healing period of 4 weeks, however, the trabeculae became sparser and thinner among rats in the diabetic group. While in the control group, the trabeculae got denser and well organized. In the diabetic-GL group, the reduction in trabecular number caused by diabetes was relieved, but not eliminated. (c) Morphology of peri-implant trabecular separation. In the diabetic group, the intertrabecular space is filled with macro yellow-to-red bubbles, indicating large space and big separation. While in the control group, the space is filled with micro green-to-blue bubbles, representing thin space and small separation. (d) Representative images of BIC by μCT evaluation. The pink area represents bone in direct contact with the implant. (e) Quantification of BIC by μCT evaluation. Representative images (f) and quantification (g, h) of BIC under histomorphometry, showing the newly formed bone tissue adjacent to the implant surface (Masson trichrome staining). n = 6. ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.001.

Figure 8

Immunochemistry evaluation of RAGE and HO-1 of the trabecular around the implant and effect of HMGB1 inhibiting on the expression of OPG, RANKL, Runx2, and OCN. The brown color indicates positive cells. Scale bar = 100 μm.