In vitro osteoclast differentiation enhanced by hepatocyte supernatants from high-fat diet mice

Abstract

Non-alcoholic fatty liver disease (NAFLD) is associated with abnormal bone metabolism, potentially mediated by elevated levels of proinflammatory cytokines such as tumor necrosis factor alpha (TNF-ɑ) and interleukin 6 (IL-6). This study aims to investigate the direct regulatory effects of liver tissues on osteoblast and osteoclast functions in vitro, focusing on the liver-bone axis in NAFLD. Twelve-week-old C57BL/6 mice were fed either a control diet or a high-fat diet (HFD) for 12 weeks. Bone structural parameters were assessed using microCT. Primary hepatocyte cultures were established from control and HFD-fed C57BL/6 mice, as well as IL-6^−/− and TNF-α^−/− mice. The supernatants from these hepatocyte cultures were used to induce differentiation in bone marrow cell-derived osteoblasts and osteoclasts in vitro. Results showed that mice on a HFD exhibited increased lipid infiltration in liver and bone marrow tissues, alongside reduced bone mass. Moreover, the supernatants from hepatocyte cultures from mice on a HFD displayed elevated TNF-α and IL-6 levels. These supernatants, particularly those derived from HFD-fed and IL-6^−/− mice, significantly enhanced osteoclast differentiation in vitro. In contrast, supernatants from TNF-α^−/− mice did not significantly affect osteoblast or osteoclast differentiation in vitro. In conclusions, this current study suggested that fatty liver tissues may negatively impact bone metabolism. Additionally, knockout of TNF-α and IL-6 genes revealed distinct influence on osteoblast and osteoclast functions, highlighting the complex interplay between live pathology and bone health.

Highlights

• •Mice fed on a high-fat diet had increased lipid infiltration in liver and bone marrow tissues and decreased bone mass.
• •The supernatants from hepatocyte cultures from mice on a HFD displayed elevated TNF-α and IL-6 levels.
• •The hepatocyte supernatants from mice with fatty liver promoted osteoclast differentiation in vitro.
• •Knockout of TNF-α and IL-6 genes revealed distinct influence on osteoblast and osteoclast functions.

1. Introduction

Non-alcoholic fatty liver disease (NAFLD), which is the leading cause of chronic liver diseases, has become increasingly prevalent with the population aging and the rise in obesity. Currently, NAFLD affects over 25 % of the global population. In recognition of the close association of this disease with multiple metabolic disturbances such as obesity, diabetes and dyslipidemia, the International Expert Consensus Panel has proposed a new definition for NAFLD—metabolic dysfunction-associated fatty liver disease (MAFLD) [1,2]. Besides fatty infiltration of the liver and its associated complications such as hepatitis, fibrosis and cirrhosis, NAFLD is now considered by medical researchers and health professionals as a systemic disease, drawing increasing attention to its extrahepatic manifestations [[3], [4], [5]].

One such extrahepatic manifestation is impaired bone metabolism in individuals with NAFLD. We have found lower circulating sclerostin (SOST) levels in NAFLD subjects [6]. Polyzos et al. also observed a progressive decline in serum SOST levels, decreasing from controls (76.1 ± 6.8 pmol/L) to nonalcoholic simple steatosis (53.5 ± 6.4 pmol/L) and steatohepatitis (NASH) (46.0 ± 8.1 pmol/L) patients (p = 0.009) [7]. Given the positive correlation between circulating SOST levels and bone mass, decreased SOST levels in NAFLD subjects may be an indicator of reduction in bone mass in these individuals [[8], [9], [10]]. Additional studies have reported reduction in bone mineral density (BMD) in NAFLD subjects [11,12]. However, research on the relationship between NAFLD and bone mass has yielded inconsistent results [11,[13], [14], [15], [16]]. For instance, a study of the Chinese population reported a positive correlation between NAFLD and BMD [17], while other studies have found a negative relationship between NAFLD and BMD [13,18]. Sex-specific differences in the effects of NAFLD on BMD have also been documented. A study in Korea revealed that men with moderate to severe NAFLD presented reduction in femoral neck BMD, while postmenopausal women presented with increased lumbar spine BMD [19].

The relationship between NAFLD and BMD is intricate and influenced by multiple factors including abnormal lipid and vitamin D metabolism, reduced physical activity, chronic inflammation, and abnormal adipokines [13,20,21]. Particularly, abnormal lipid metabolism and chronic inflammation, characterized by sustained production of proinflammatory cytokines such as tumor necrosis factor alpha (TNF-ɑ) and interleukin 6 (IL-6), are considered key pathophysiological mechanism in NAFLD [21,22]. Our previous studies have shown that the knockout of TNF-ɑ and IL-6 enables resistance to high-fat diet (HFD)–induced bone loss in mice [23,24]. Therefore, investigating whether TNF-ɑ and IL-6 are involved in the link between fatty liver and abnormal bone metabolism warrants further investigation.

Currently, studies examining the relationship between NAFLD and bone metabolism are primarily limited to cross-sectional studies and animal studies. Research that directly observes the influence of fatty liver tissues on osteoblast and osteoclast functions is limited. Therefore, this study aims to investigate whether the fatty liver tissues have a direct regulatory effect on osteoblast and osteoclast differentiation in vitro.

2. Materials and methods

2.1. Animals and intervention

All animal experiments were conducted in accordance with the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978) and were approved by the Institutional Animal Care and Use Committee at the West China Hospital (Sichuan University). Every possible measure was taken to minimize any discomfort or suffering of mice. Male C57BL/6 mice, IL-6 knockout (IL-6^−/−) mice and TNF-ɑ knockout (TNF-ɑ^−/−) mice were maintained in a controlled environment, ensuring temperature and lighting conditions were regulated within a pathogen-free barrier facility, operating on a 12-h light–12-h dark cycle. IL-6^−/− mice and TNF-ɑ^−/− mice were generated on the C57BL/6 background and were purchased from the Jackson Laboratory (Bar Harbor, ME). They had ad libitum access to water and diets. At the age of 12 weeks, C57BL/6 mice were divided into 2 groups each consisting of 8 animals. They were fed either a standard chow diet (comprising 64 % carbohydrates, 10 % fat, and 26 % protein) or a HFD (28 % carbohydrate, 60 % fat, and 12 % protein) for a duration of 12 weeks.

2.2. Plasma and tissue sampling

At the conclusion of the experiment, the mice were humanely euthanized via intraperitoneal injection of sodium pentobarbital. Liver and adipose tissues were fixed in 4 % formalin, embedded in paraffin, and stained with Hematoxylin and Eosin (H&E) staining for histopathological analysis. Additionally, Oil Red O staining was performed on the frozen liver sections. The right femora were promptly immersed into 4 % paraformaldehyde, and these samples were used to further assess bone structural parameters. The left femora were flash-frozen in liquid nitrogen and stored at −80 °C for subsequent measurement of mRNA expression levels of bone-related genes. Tibias were fixed using paraformaldehyde and tissue sections were stained with H&E. The quantities of osteoblasts and bone marrow adipocytes (BMAs) were evaluated. Resistant acid phosphatase (TRAP) staining was performed on fixed tibia sections using a kit from Sigma Aldrich (Merck KGaA, Germany). BMAs were isolated following previously established methods [25].

2.3. Evaluation of bone structural parameters with microcomputed tomography

Bone samples from mice were imaged with microcomputed tomography (μCT) (MicroCT80, Scanco Medical AG, Bassersdorf, Switzerland) as previously described [26]. After reconstruction of μCT scans, quantitative analysis was performed to determine bone structural parameters such as trabecular bone volume (Tb·BV/TV), trabecular number (Tb·N), trabecular thickness (Tb·Th), trabecular connectivity density (Conn.D) and trabecular separation (Tb.Sp).

2.4. Isolation and culture of mouse hepatocytes

Male C57BL/6 mice, HFD-fed mice, IL-6^−/−, and TNF-α^−/− mice all aged 24 weeks, were included in this study. After anesthesia induced by intraperitoneal injection of sodium pentobarbital, the mice were sterilized with 75 % ethanol. A ‘U’-shaped incision was made through the skin of the lower abdomen to the lateral aspect of the rib cage, the skin was folded back over the chest. After exposing the portal vein and the vena cava, a cannula was inserted into the vena cava, and perfusion medium was slowly perfused through the liver at a rate of 8–10 ml/min. When the liver was swollen and white, the portal vein was cut and the perfusion was continued for approximately 10–15 min. After removal of the gallbladder, the liver was removed and transferred to a 50 ml sterile tube containing 10 ml digestion buffer. Liver tissues were carefully dissected and sectioned into 1–2 mm segments using forceps in a sterile environment. The resulting suspension of liver tissues was further digested in a 37 °C incubator for 20 min. Subsequently, 4 ml α-MEM (Thermo Scientific, US) supplemented with 10 % fetal bovine serum (FBS) (Thermo Scientific, US) was added to halt the digestion process. The crude hepatocyte preparation was filtered with a 70 μm cell strainer, transferred to a sterile 50 ml tube, and centrifuged at 50×g for 5 min at 4 °C. After resuspending the hepatocytes in serum-free media and centrifuging the suspension for 5 min, the hepatocytes were counted and plated on collagen-coated cell culture plates. Perfusion medium and digestion buffers were prepared following previously reported protocols [27]. Hepatocytes were cultured in Willam's E media (Thermo Scientific, US) supplemented with 10 % FBS (Thermo Scientific, US), 2 mM l-glutamine, 1 % penicillin/streptomycin (P/S), and 1 % nonessential amino acids in a 37 °C incubator regulated at 5 % CO₂. Supernatants collected at 4 h, 24 h, and 48 h were mixed equally and TNF-ɑ and IL-6 levels were evaluated with enzyme-linked immunosorbent assays (ELISA) (Elabscience Biotechnology, China).

2.5. Influence of hepatocyte culture supernatants on osteoblast and osteoclast differentiation in vitro

Hepatocyte culture supernatants collected at 4 h, 24 h, and 48 h were stored at −80 °C prior to experiments. Before use, supernatants collected at different time points were mixed equally and were further mixed with the ordinary culture medium at a 1:1 ratio. Femur and tibia from 6-week–old male C57BL/6 mice were used for cell cultures following established methods [26]. Bone marrow cells were extracted using a 3 cm³ syringe and 25G needle. The collected cells were counted and plated on 24-well plates at a density of 1 × 10⁶ cells/well and incubated at 37 °C in α-MEM (Thermo Scientific, US) supplemented with 10 % FBS (Thermo Scientific, US), 1 % P/S. The day of plating was designated as day 0.

For osteoblast culture, on day 7, the culture medium was switched to conditional medium (comprising 50 % regular medium and 50 % supernatants from hepatocyte cultures) and supplemented with 50 μg/ml ascorbic acid and 10 mM beta-glycerophosphate. After 7 days of osteoblast induction, alkaline phosphatase (ALP) staining was performed using a kit from Sigma Aldrich (Merck KGaA, Germany). For osteoclast culture, bone marrow cells were collected as above and cultured at 37 °C in conditional medium (50 % regular medium and 50 % hepatocyte culture supernatants), supplemented with 25 ng/ml of human macrophage-colony stimulating factor (M-CSF) (R&D Systems, U.S.) and 30 ng/ml of human soluble receptor activator of NF-κB ligand (RANKL) (R&D Systems, U.S.). Cell media was replaced every 3 days. At day 7, TRAP staining was performed with a kit from Sigma Aldrich (Merck KGaA, Germany) to quantify osteoclast formation. TRAP-positive cells containing three or more nuclei were counted as osteoclasts.

2.6. qRT-PCR for gene expression

Total RNA was extracted from liver tissues, femur distal metaphyses, and bone marrow adipose tissues using Trizol reagent according to the manufacturer's protocol (Invitrogen, Frederick, USA). RNA was reverse transcribed into cDNA with PrimeScriptR RT reagent kit (TaKaRa Biotechnology Co., Ltd., Dalian, China) and the cDNA product was subsequently amplified and quantified (TaKaRa Biotechnology Co., Ltd., Dalian, China). The sequence of oligonucleotide primers used for reverse transcription is listed in Table 1. Reference gene Gapdh was used as an endogenous control. The relative mRNA expression levels were normalized to Gapdh levels and analyzed with the 2^−△△CT method.

Table 1.

Oligonucleotide primers for cDNA amplification.

Gene name	Forward	Reverse
GAPDH	TGCACCACCAACTGCTTAG	GGATGCAGGGATGATGTTC
Atgl	ACTGAACCAACCCAACCCTTT	GCCACTCCAACAAGCGGA
Cpt2	CCAGCTACATCTCAGGCCCC	CCAGGAGGTGTCTAGCCTTG
Dgat2	AACACGCCCAAGAAAGGTGG	GTTCTTCAGGGTGACTGCGT
Fas	TGCCCGAGTCAGAGAACCTA	AGGCTGGGTTGATACCTCCA
Hsl	GACTCTAACGCGACTCCTCAC	TGTGAGAACGCTGAGGCTTTGA
Lpl	AGGTGGACATCGGAGAACTG	TGTTTGTCCAGTGTCAGCCAGA
Plin2	CAGCTCTCCTGTTAGGCGT	CGGAGGACACAAGGTCGTAG
Col1α1	GCGAAGGCAACAGTCGCT	CTTGGTGGTTTTGTATTCGATGAC
Ctsk	GAAGAAGACTCACCAGAAGCAG	TCCAGGTTATGGGCAGAGATT
SOST	CCTCATCTGCCTACTTGTGC	GGTCTGGTTGTTCTCAGGAGG

3. Results

3.1. Metabolic parameters

After feeding on different diets for 12 weeks, C57BL/6 mice fed a HFD showed higher liver weight and serum total cholesterol (TC) levels, when compared with mice on a control diet. Body weight, serum triglyceride (TG) and fasting glucose levels were similar between the two groups (Fig. 1).

Fig. 1.

Comparison of metabolic parameters in mice. Liver weight, liver weight/body weight and serum TC levels were increased in mice fed a HFD. Body weight, serum TG levels and fasting glucose levels were similar between the two groups. CON, control; HFD, high-fat diet; TC, total cholesterol; TG, triglyceride.

3.2. Histological analyses of liver, adipose and bone tissues

In mice on a HFD, an increased average diameter of adipocyte cells, disorganized subcutaneous and visceral adipose tissues, larger quantities of lipid droplet accumulation in liver tissues, and elevated quantity of fat vacuoles in bone marrow were observed (Fig. 2). Furthermore, H&E and TRAP staining of tissue sections from mice on a HFD regimen revealed a decrease in viable osteoblasts and an increase in osteoclasts count (Fig. 3).

Fig. 2.

Histological analyses of subcutaneous and visceral tissues, bone, and liver tissues. (A–D) Hematoxylin and eosin staining showed larger fat cell size in subcutaneous adipose tissues of mice fed a HFD. (E–H) Larger fat cell size in visceral adipose tissues of mice fed a HFD. (I, J) More fat vacuoles in the bone marrow of mice fed on a HFD. (K, L) More lipid droplets accumulated in the liver tissues of HFD-fed mice with Oil Red O staining. CON, control; HFD, high-fat diet. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3.

Histological analyses of bone tissues. (A–C) Hematoxylin and Eosin staining found decreased number of osteoblasts in mice fed a HFD. (D–F) TRAP staining found increased number of osteoclasts in mice fed a HFD. CON, control; HFD, high-fat diet; TRAP, resistant acid phosphatase.

3.3. Bone mass and structure

Three-dimensional reconstruction of region of interest images of mice femora revealed reduced bone density, fractures, and a wider trabecula space in mice on a HFD (Fig. 4). μCT results revealed congruent trends. Tb.BV/TV (0.34 ± 0.01 vs. 0.45 ± 0.03, p = 0.000) and Tb·Th (0.071 ± 0.01 vs. 0.090 ± 0.01, p = 0.021) were significantly lower, while Tb. Sp (0.37 ± 0.05 vs. 0.26 ± 0.02, p = 0.000) and Conn. D (725.93 ± 201.30 vs.76.39 ± 22.57, p = 0.000) were significantly higher in mice on a HFD regimen, when compared with mice on a control diet (Fig. 4).

Fig. 4.

Decreased bone mass in mice fed a HFD. (A–H) Three-dimensional reconstruction of region of interest image of femur by μCT analysis software. (I) Trabecular bone volume (Tb. BV/TV). (J) Trabecular bone thickness (Tb·Th). (K) Trabecular connectivity density (Conn.D). (L) Trabecular separation (Tb·Sp). (M) Trabecular bone number (Tb. N). CON, control; HFD, high-fat diet.

3.4. Gene expression profiles in liver and bone tissues

Abnormal lipid metabolism and chronic inflammatory conditions have been theorized to be one of the primary underlying mechanism in the pathogenesis of fatty liver disease [22,28]. Our study revealed that mice on a HFD exhibited elevated expression levels of Lpl (lipoprotein lipase) and Atgl (adipose triglyceride lipase) in liver tissues, while reduced Fas (fatty acid synthase) levels were observed. Conversely, the expression levels of Dgat2 (diacylglycerol acyltransferase 2), Cpt2 (carnitine palmitoyltransferase 2), Plin2 (perilipin 2), and Hsl (hormone sensitive lipase) remained consistent between the two groups. Additionally, mice on a HFD displayed increased TNF-α and IL-6 expression levels in liver tissues (Fig. 5). The HFD regimen resulted in a reduction in Col1a1(alpha-1 type I collagen) and SOST expression levels in bone tissues. The expression levels of Ctsk (cathepsin K) in bone tissues and Plin2 expression levels in BMAs were not different between the two groups (Fig. 5). These data indicated that mice subjected to a HFD exhibit enhanced triglyceride synthesis and hydrolase, increased accumulation of lipid storage droplets, elevated inflammatory factor expression in liver tissues, and impaired bone formation.

Fig. 5.

Relative mRNA expression levels shown by qRT-PCR. (A) Atgl expression levels were higher in liver tissues in mice fed a HFD. (B–D) Cpt2, Plin2 and Dgat2 expression levels were similar in liver tissues between the two groups. (E, F) Lpl expression levels was higher, while Fas expression levels were lower in liver tissues in mice fed a HFD. (G) Hsl expression levels were similar in liver tissues between the two groups. (H, I) HFD induced higher expression levels of TNF-α and IL-6 in liver tissues. (J–L) Mice fed a high-fat diet had lower Col1a1 and SOST expression levels in bone tissues, while Ctsk expression levels were similar. (M) Plin2 expression levels were similar in BMAs. CON, control; HFD, high-fat diet; Atgl, adipose triglyceride lipase; Cpt2, carnitine palmitoyltransferase 2; Plin2, perilipin 2; Dgat2, diacylglycerol acyltransferase 2; Lpl, lipoprotein lipase; Fas, fatty acid synthase; Hsl, hormone sensitive lipase; TNF-α, tumor necrosis factor alpha; interleukin 6, IL-6; Col1a1, alpha-1 type I collagen; SOST, sclerostin; Ctsk, cathepsin K; BMAs, bone marrow adipocytes.

3.5. TNF-ɑ and IL-6 levels in supernatants of primary hepatocyte culture

TNF-ɑ and IL-6 levels were increased in supernatants of primary hepatocyte culture from mice fed a HFD, while were decreased in supernatants from TNF-ɑ^−/− and IL-6^−/− mice (Fig. 6).

Fig. 6.

Supernatants of primary hepatocyte culture from mice fed a HFD had elevated TNF-ɑ and IL-6 levels, while supernatants from TNF-ɑ^−/− and IL-6^−/− mice showed decreased TNF-ɑ and IL-6 levels. CON, control; HFD, high-fat diet; IL-6, interleukin 6; TNF-α, tumor necrosis factor alpha.

3.6. Conditional media induced osteoblast and osteoclast differentiation in vitro

Assays performed on bone marrow cell cultures containing conditioned media revealed that the supernatants from HFD-fed mice and IL-6^−/− mice enhanced osteoclastogenesis. Supernatants from TNF-ɑ^−/− mice exhibited a similar influence on osteoblast and osteoclast differentiation in vitro with respect to those of control mice (Fig. 7). Conditioned media induced similar levels of osteoblast differentiation among the experimental groups.

Fig. 7.

Differences in the ability of osteoblastogenesis and osteoclastogenesis induced in vitro by conditional media containing primary hepatocyte culture supernatants from different strains of mice. (A–H) ALP staining suggested that osteoblast differentiation in vitro was similar among groups. (I–P) TRAP staining showed that supernatants from HFD-fed mice and IL-6^−/− mice enhanced osteoclast differentiation in vitro. CON, control; HFD, high-fat diet; ALP, alkaline phosphatase; TRAP, resistant acid phosphatase; IL-6, interleukin 6; TNF-α, tumor necrosis factor alpha.

4. Discussion

Since NAFLD has become the leading cause of chronic liver diseases, accumulating researches have been made to investigate the correlation between NAFLD and bone metabolism. Among those studies, most are cross-sectional studies, a few are retrospective cohort studies, while prospective studies are missing [19,[29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42]]. Regarding the relationship between NAFLD and BMD, although negative, neutral or positive association has been obtained from a wide range of studies including participants with different gender, age, race, and concomitant diseases, most studies tend to conclude that fatty liver showed disadvantageous effects on bone metabolism, causing reduced bone mass or increased fracture risk [19,[29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42]]. Regarding the possible underlying mechanism, vitamin D and several endocrine factors and metabolites secreted by liver have been speculated to mediate the bone-liver link, including insulin-like growth factor-1 (IGF-1), fibroblast growth factor 21 (FGF21), IGF binding protein 1 (IGFBP1), fetuin-A, TNF-α, and osteopontin (OPN) [14,43]. However, only correlative rather than causal relationship could be drawn from previous clinical studies and animal experiments and it is unknown whether NAFLD has a direct regulatory effect on bone metabolism. Therefore, this current study was designed. To current date, our study is the first to provide evidence of the primary hepatocyte culture supernatants from NAFLD mice enhances osteoclast formation in vitro when compared with supernatants from control mice.

Obesity has a complex effect on bone metabolism. On one hand, the increased mechanic load and elevated bioavailable estradiol levels associated with obesity contribute to higher bone density and reduced fracture risks. On the other hand, obesity-related metabolic disturbances, such as impaired glucose and lipid metabolism, increased fat infiltration in liver and bone marrow, dysregulated adipokine secretion, and chronic inflammation have deleterious effects on bone metabolism [[44], [45], [46], [47]]. The combined effects of obesity on BMD likely result from a balance between these favorable and unfavorable pathophysiologies. Our previous study suggested that NAFLD may serve as a critical point in fat-liver-bone link. Evidence supporting this notion includes the following: in individuals without NAFLD, no significant association were found between circulating SOST levels and metabolic indicators [6]. On the contrary, in individuals with NAFLD, circulating SOST levels were negatively correlated with various physical and metabolic indicators including waist circumference, urea, hepatic enzyme, gamma-glutamyl transpeptidase, and triglyceride levels, while such correlations were not significant in control subjects [6]. Furthermore, NAFLD mice exhibited reduced bone mass and decreased SOST expression levels in bone tissues, both of which were positively correlated [6]. Based on these data, we speculate that in patients with NAFLD, the detrimental effects of obesity on bone metabolism may outweigh the beneficial effects, resulting in reduced bone mass.

NAFLD is closely associated with systemic chronic inflammation, with the progression from simple steatosis to nonalcoholic steatohepatitis driven by chronic liver inflammation and characterized by elevated levels of proinflammatory cytokines, TNF-ɑ, IL-6, and IL-1β as a result of low-grade, persistent inflammation [22,48,49]. Our study also revealed elevated IL-6 and TNF-ɑ expression levels in fatty liver tissues of mice. Notably, persistent chronic inflammatory status and elevated circulating TNF-ɑ and IL-6 levels have also been reported to be closely associated with the development of osteoporosis [43,50]. It has been speculated that TNF-ɑ and IL-6 may be involved in liver-bone link. In our previous animal studies, we found that gene knockout of IL-6 and TNF-ɑ could mitigate HFD–induced bone loss, although the underlying mechanism was not completely the same [23,24,51,52], indicating antagonizing the effects of TNF-ɑ and IL-6 may be beneficial for bone metabolism. However, our current study did not find that hepatocyte supernatants from TNF-ɑ and IL-6 knockout mice had a beneficial effect on bone metabolism. On the contrary, our results showed that supernatants from IL-6^−/− mice promoted osteoclastogenesis, as evidenced by elevated quantities and diameters of osteoclasts observed through TRAP staining, whereas supernatants from TNF-ɑ^−/− mice exhibited a similar influence on osteoclast and osteoblast activity in vitro as those from control mice. These differences may be secondary to the different degree of lipid infiltration in liver tissues between the two strains of mice. Our previous research has identified different metabolic responses in TNF-ɑ^−/− and IL-6^−/− mice dependent on diet (Supplemental Fig. 1) [26,28]. The expression profiles of lipid-related genes in liver and adipose tissues in IL-6^−/− mice, whether on a control diet or a HFD, resembled those of C57BL/6 J mice on a HFD, resulting in significant lipid accumulation in liver tissues of IL-6^−/− mice regardless of dietary regimen. Conversely, TNF-ɑ^−/− mice on a control diet did not exhibit lipid accumulation in liver tissues; however, on a HFD regimen, TNF-ɑ^−/− mice developed a substantial accumulation of lipid droplets in the liver tissues in quantities surpassing those of IL-6^−/− and control mice. Therefore, the similarity in the cell response induced by supernatants from both control mice on a HFD and from IL-6^−/− mice may be attributed to enhanced lipid infiltration in the liver tissues of these mice, which further suggests that fatty liver itself shows an obvious effect on osteoclastic function in vitro.

In our previous animal studies, we observed that a high-fat diet had a dual impact on bone health, impacting both osteoblasts and osteoclasts. This resulted in compromised bone formation and heightened bone resorption [26]. However, our recent in vitro analysis indicates that supernatants from fatty liver only stimulate osteoclast differentiation without significantly affecting osteoblast differentiation. This discrepancy suggests that in vivo, the effects of a high-fat diet on bone metabolism are likely the result of complex systemic metabolic dysregulation involving numerous organs and tissues. Therefore, more fine-designed in vivo and in vitro studies are needed to ascertain whether fatty liver primarily influences osteoclastic functions.

Some limitations exist in this current study. First, regarding the direct regulatory effects of fatty liver on bone metabolism, only in vitro study was performed. Secondly, the conditioned medium used in this study contained half of the ordinary culture medium, and osteoblastogenic and osteoclastogenic factors may outweigh the influence of the conditioned medium, which may be one of the reasons that no effects were observed in TNF-ɑ^−/− mice. At last, since the expression of Gapdh may be affected by high-fat diet, choosing Gapdh as reference gene may show some influence on data analysis.

5. Conclusion

In conclusion, our study identified a correlation between HFD and a reduction in bone mass in mice, accompanied by lipid infiltration in liver and bone marrow tissues. The conditional media containing hepatocyte culture supernatants from mice with fatty liver enhanced osteoclast formation in vitro, indicating liver tissues may have a direct regulatory effect on bone metabolism, especially on osteoclast differentiation.

Funding

This work was also supported by the 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (Grant No. ZYGD18022 to HM Tian). The study was also supported by the National Natural Science Foundation of China (Grant No. 82200641 to FL Zhou), China Postdoctoral Science Foundation (Grant No. 2021YFS0198 to FL Zhou), Chengdu Science and Technology Program (Grant No. 2022-YF05-01497-SN to FL Zhou), West China Hospital, Sichuan University (Grant No. 2020 HXBH028 to FL Zhou), science and technology department of Tibet, the central government guides the local science and technology development fund project (Grant No. XZ202102YD0026C to YH Wu), and National Natural Science Foundation of China (Grant No.82300987 to YJ Li).

Ethical approval

This study was approved by the Ethics Committee of the West China Hospital.

CRediT authorship contribution statement

Yan Wang: Project administration, Methodology, Investigation. Fangli Zhou: Project administration, Methodology, Investigation, Funding acquisition. Siyi Shu: Project administration, Methodology, Investigation. Yunhong Wu: Project administration, Funding acquisition, Formal analysis. Haoming Tian: Writing – review & editing, Project administration, Funding acquisition. Yujue Li: Project administration, Methodology, Funding acquisition. Xiang Chen: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Formal analysis.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Supplemental Fig. 1.

Differences in hepatic lipid droplet distribution induced by different diets in TNF-ɑ^−/− and IL-6^−/− mice. IL-6, interleukin 6; TNF-α, tumor necrosis factor alpha; CON, control; HFD, high-fat diet. (Xiang Chen et al., J Interferon Cytokine Res. 2016 Oct; 36:580–588, With permission from Xijie Yu)

Data availability

Data will be made available on request.