Chronic alcohol ingestion induces osteoclastogenesis and bone loss through IL-6 in mice

Abstract

To investigate the role of IL-6 in alcohol-mediated osteoporosis, we measured a variety of bone remodeling parameters in wild-type (il6(+/+)) or IL-6 gene knockout (il6(-/-)) mice that were fed either control or ethanol liquid diets for 4 months. In the il6(+/+) mice, ethanol ingestion decreased bone mineral density, as determined by dual-energy densitometry; decreased cancellous bone volume and trabecular width and increased trabecular spacing and osteoclast surface, as determined by histomorphometry of the femur; increased urinary deoxypyridinolines, as determined by ELISA; and increased CFU-GM formation and osteoclastogenesis as determined ex vivo in bone marrow cell cultures. In contrast, ethanol ingestion did not alter any of these parameters in the il6(-/-) mice. Ethanol increased receptor activator of NF-kappaB ligand (RANKL) mRNA expression in the bone marrow of il6(+/+) but not il6(-/-) mice. Additionally, ethanol decreased several osteoblastic parameters including osteoblast perimeter and osteoblast culture calcium retention in both il6(+/+) and il6(-/-) mice. These findings demonstrate that ethanol induces bone loss through IL-6. Furthermore, they suggest that IL-6 achieves this effect by inducing RANKL and promoting CFU-GM formation and osteoclastogenesis.

Figure 1

Effect of ethanol ingestion on whole-body and excised femoral BMD in wild-type and IL-6 gene knockout mice. Il6^+/+ and il6^–/– mice were fed either a control diet or 5% ethanol diet for 4 months. (a) Whole-body BMD was measured before and 4 months after initiation of the diet. (b) Femoral BMD was measured from excised left femurs 4 months after initiation of diet. Data are reported as (a) mean (± SD) percent change in whole-body BMD measured at 4 months relative to the prediet BMD or (b) mean (± SD) percent difference between femoral BMD in ethanol-fed mice compared with pair-fed matched control-fed mice. Data were analyzed using ANCOVA using bodyweight as a covariate and Fisher’s least significant difference for post hoc analysis. ^AP < 0.03 vs. control-fed il6^+/+ mice. Measurements were performed in eight to ten individual mice per group.

Figure 2

Effect of ethanol ingestion on urinary deoxypyridinolines in wild-type or IL-6 gene knockout mice. Il6^+/+ or il6^–/– mice were fed either a control diet or 5% ethanol diet for 4 months. The mice were then sacrificed, and urine was collected and evaluated for deoxypyridinolines (Dpd’s) and creatinine levels as described in Methods. Data are presented as mean (± SD) urinary Dpd corrected for creatinine levels. Data were analyzed using ANOVA and Fisher’s least significant difference for post hoc analysis. ^AP = 0.045 compared with control-fed il6^+/+ mice. Measurements were performed in six to eight individual mice per group.

Figure 3

Effect of ethanol ingestion on CFU-GM formation, osteoclastogenesis, and RANKL mRNA expression in the bone marrow of wild-type or IL-6 gene knockout mice. Il6^+/+ or il6^–/– mice were fed either a control diet or 5% ethanol diet for 4 months. The mice were then sacrificed, and bone marrow was collected from the femur. (a) CFU-GM formation was determined as described in Methods. Data are presented as mean (± SD) CFU-GM formed per 5.5 × 10⁵ nucleated bone marrow cells. (b) Osteoclast formation was determined in marrow cultures maintained for 9 days in the presence of 10 nM 1,25(OH)₂D3. Osteoclast numbers were determined by counting cells that bound I-125-calcitonin and were TRAPase-positive as described in Methods. Data are presented as mean (± SD) osteoclasts per 1.5 × 10⁶ nucleated bone marrow cells. (c) RANKL mRNA expression was determined by subjecting bone marrow cell total RNA to RT-PCR for RANKL and β-actin mRNA as described in Methods. Density of individual bands was determined. A representative gel is shown. Data are presented as mean (± SD) relative density of RANKL PCR product bands normalized with β-actin PCR product bands. (d)Data were analyzed using ANOVA and Fisher’s least significant difference for post hoc analysis. ^AP < 0.001 compared with control-fed il6^+/+ mice. ^BP = 0.033 compared with control-fed il6^+/+ mice. ^CP < 0.01 compared with control-fed il6^+/+ mice. Measurements were performed on ten individual mice per group for CFU-GM and osteoclast studies and four individual mice per group for RANKL mRNA determination.

Figure 4

Effect of ethanol ingestion of the ability of CFU-F to support osteoclastogenesis. Il6^+/+ or il6^–/– mice were fed either a control diet or 5% ethanol diet for 4 months. The mice were then killed, and bone marrow was collected from the femur. CFU-F cultures, from individual animals, were established in 96-well plates for 5-days; then single-cell suspension from spleens of il6^+/+ mice were added to the wells. The cultures were maintained for an additional 8 days and then stained for TRAP. Each well was then examined microscopically for the presence of CFU-F (>20 cells) and osteoclast-like cells (TRAP-positive). Data are presented as mean (± SD) CFU-F with osteoclasts. Data were analyzed using ANOVA and Fisher’s least significant difference for post hoc analysis. ^AP < 0.01 compared with control-fed il6^–/– mice. Measurements were performed on ten individual mice per group. OC, osteoclasts.

Figure 5

Effect of ethanol ingestion on osteoblast parameters in wild-type or IL-6 gene knockout mice. Il6^+/+ or il6^–/– mice were fed either a control diet or 5% ethanol diet for 4 months. The mice were then killed and bone marrow was collected from the femur. CFU-F, CFU-OB, and Ca. (a) To assess the marrow’s ability to support osteoblastogenesis, marrow cultures were maintained for 10 days in the presence of 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate. CFU-Fs were determined by counting ALP-positive colonies as described in Methods. (b and c) To assess for the marrow’s ability to support osteoblast function, marrow cultures were maintained for 25 days in the presence of 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate. (b) CFU-OBs were determined by counting von Kossa–positive colonies and (c) total calcium in the cultures was determined as described in Methods. Data are reported as mean (± SD). Data were analyzed using ANOVA and Fisher’s least significant difference for post hoc analysis. ^AP < 0.01 compared with control-fed il6^–/– mice. ^BP = 0.02 compared with control-fed il6^–/– mice. ^CP = 0.005 compared with control-fed il6^–/– mice. Measurements were performed on ten individual mice per group.