Bone matrix quality in a developing high-fat diet mouse model is altered by RAGE deletion

Abstract

Overweightness and obesity in adolescents are epidemics linked to chronic low-grade inflammation and elevated fracture risk. The increased fracture risk observed in overweight/obese adolescence contrasts the traditional concept that high body mass is protective against fracture, and thus highlights the need to determine why weight gain becomes detrimental to fracture during growth and maturity. The Receptor for Advanced Glycation End products (RAGE) is a central inflammatory regulator that can influence bone metabolism. It remains unknown how RAGE removal impacts skeletal fragility in overweightness/obesity, and whether increased fracture risk in adolescents could result from low-grade inflammation deteriorating bone quality. We characterized the multiscale structural, mechanical, and chemical properties of tibiae extracted from adolescent C57BL/6J (WT) and RAGE null (KO) mice fed either low-fat (LF) or high-fat (HF) diet for 12 weeks starting at 6 weeks of age using micro-computed tomography, strength, Raman spectroscopy, and nanoindentation. Overweight/obese WT HF mice possessed degraded mineral-crystal quality and increased matrix glycoxidation in the form of pentosidine and carboxymethyl-lysine, with HF diet in females only showing reduced cortical surface expansion and TMD independently of RAGE ablation. Furthermore, in contrast to males, HF diet in females led to more material damage and plastic deformation. RAGE KO mitigated glycoxidative matrix accumulation, preserved mineral quantity, and led to increased E/H ratio in females. Taken together, these results highlight the complex, multi-scale and sex-dependent relationships between bone quality and function under overweightness, and identifies RAGE-controlled glycoxidation as a target to potentially preserve matrix quality and mechanical integrity.

Keywords: Bone quality; Glycation; Mineralization; Obesity; RAGE.

Fig. 1.

Experimental design and outcome measures. Figure was adapted from “Mouse High Fat Diet Experimental Timeline”, by BioRender.com (2022). Retrieved from https://app.biorender.com/biorender-templates

Fig. 2.

HF diet led to significant weight gain in mice. Body weight in HF females and HF males were significantly higher than LF controls at termination (18 weeks of age). WT HF mice also possessed elevated weight compared to KO HF mice. Notably, WT LF males exhibited higher weight than KO LF controls. WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat.

Fig. 3.

HF diet led to impaired cortical structure in females. HF diet in females led to significantly reduced cortical area (p = 0.006, A), cortical thickness (p < 0.001, B) and TMD (p < 0.001, C) compared to LF females and was independent of RAGE presence. RAGE loss was linked to lowered moment of inertia in HF females relative to WT HF controls (p = 0.013, D). Diet and RAGE loss had no effect on these properties in males. WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat.

Fig. 4.

HF diet in females led to altered mineral and organic matrix quality which were conserved by RAGE KO. Amide I mineral-to-matrix ratio was strongly heightened by RAGE loss in both sexes (p < 0.001, A). Amide III mineral-to-matrix ratio was lowered in WT HF females but remained constant in KO females (p = 0.03, B). WT HF females exhibited lowered crystallinity over LF controls and was not conserved by RAGE loss, while RAGE loss in LF males lead to decreased crystallinity over controls (p < 0.001, C). WT HF females exhibited high type-B carbonate over WT LF and KO HF groups (p < 0.001, D). WT HF males possessed lowered type-B carbonate levels compared to WT LF cohort, but RAGE loss reversed this difference. Pentosidine intensity was also strongly raised in WT HF females with respect to WT LF littermates, but RAGE weakly attenuated this difference (p = 0.04, E). CML intensity was greatly heightened in WT HF females over WT LF and KO HF females, while KO led to increased CML intensity in LF males that was suppressed in KO HF males (p < 0.001, F).

Fig. 5.

Glycoxidation markers measured in the bone matrix correlated to carbonate content. Levels of pentosidine in WT females (p = 0.01, A) and levels of CML in WT males (p = 0.01, B) were correlated to type-B carbonate content. Type-B carbonate content was significantly correlated to CML in KO females (p < 0.001, C), but there were no significant relationships between glycoxidation markers and mineral parameters in KO males (graphs not shown). WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat.

Fig. 6.

RAGE KO increased matrix rigidity and E/H ratio in females, and E/H ratio was predicted by parameters of mineral quality. KO females possessed elevated surface elastic modulus (p < 0.001, A) and elastic-to-plastic ratio (p < 0.001, B) relative to WT controls. KO LF males exhibited higher surface elastic modulus and elastic-to-plastic ratio relative to WT LF controls, but this pattern was reversed in the HF male groups. Amide I mineral-to-matrix was a moderate predictor for elastic-to-plastic ratio in both WT and KO females (C-D) while type-B carbonate was a negative predictor for elastic-to-plastic ratio in WT and KO males (E-F). WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat.

Fig. 7.

In females, diet-induced weight gain in WT mice led to impaired mineral crystal quality and increased post-translational modifications in the form of pentosidine and CML. Such impairment of organic matrix quality and mineral crystal state in WT HF females impacted the degree of mineralization ($v_2{\text{PO}}_4^{3-}$/Amide III mineral-to-matrix ratio and TMD in this work). Potential alterations in bone cell activity would most likely lead to the observed changes in morphology. Loss of RAGE in KO females mitigated the degree of post-translational modifications triggered by high-fat diet, but still led to impaired mineral-crystal quality. However, loss of RAGE led to increased nano-hardness that may stem from altered collagen structure, which manifested in greater collagen anisotropy at the microscale. This flowchart synthesizes all data generated in this work, while including several other notable metrics of bone quality not measured here, to give an illustration on how matrix-level and tissue-level results can occur within a single model and may interact with one another. The arrows in the figure, along with their directionality (up or down), illustrates how measured aspects of bone quality change with respect to a control (e.g., an upward red arrow besides PYD demonstrates that WT HF females exhibited higher PYD compared to WT LF females). WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat. Figure was created in BioRender.com (2022).

Fig. 8.

In males, diet-induced weight gain increased the degree of post-translational modifications and nano-hardness but had no other observable influence over the bone hierarchy proposed here. Of interest, the loss of RAGE in males, regardless of diet type, had a profound impact on mineral crystal status. This flowchart synthesizes all data generated in this work, while including several other notable metrics of bone quality not measured here, to give an illustration on how matrix-level and tissue-level results can occur within a single model and may interact with one another. The arrows in the figure, along with their directionality (up or down), illustrates how measured aspects of bone quality change with respect to a control (e.g., a downward green arrow besides PYD demonstrates that RAGE KO HF males exhibited lower PYD compared to RAGE KO LF males). WT = wildtype, KO = RAGE knockout, LF = low fat, HF = high fat. Figure was created in BioRender.com (2022).