Bone regeneration is a delicate and complex physiological process (Clarke, 2008). Bone defects resulted from traumatic injury, tumor resection, and degenerative diseases are challenging problems in clinics (Burge et al., 2007). The gold standard in clinical practice is an autologous bone graft (Einhorn and Gerstenfeld, 2015). Unfortunately, the limited sources, the injury of the donor site and perioperative complications significantly restrict the employment of this approach (Baqain et al., 2009). Bone morphogenic protein 2 (BMP-2) is considered the most efficacious cytokine for bone repair and has been extensively studied for the treatment of various bone fractures and bone defects (Glassman et al., 2008; Long, 2011). However, supraphysiological dosage is necessary in clinical practice, causing undesirable side effects, including hollow bone formation, life-threatening tissue edema, and cancer (Cahill et al., 2009; Barbour et al., 2011; Skovrlj et al., 2015). In addition, the high cost for BMP-2 becomes a heavy financial burden for the health system. Therefore, developing alternative strategies are imperative for bone regeneration.

The close relationship between bone and fat formation was well acknowledged in the literature (Gimble et al., 2006). The bone marrow mesenchymal stem cells may take different pathways during their lifetime, to differentiate and transdifferentiate in response to changes in the microenvironment to bone or fat (Chen et al., 2016). The inverse relationship between bone and fat implied that agents inducing adipogenesis inhibited osteoblast differentiation and vice versa (Naot et al., 2017). These results coincided with classic pathological and epidemiological phenomena of increased marrow adiposity with aging and bone loss.

Human adiponectin is a 30-kDa adipose-derived secreted protein containing 244 amino acid residues, with an N-terminal signal sequence, a hypervariable region, a collagenous domain, and a globular domain (Scherer et al., 1995). Since it was first discovered in 1995, efforts have been devoted to unraveling the biological activities of adiponectin. Metabolic regulation and maintenance of whole-body energy homeostasis are recognized as the main physiological role of adiponectin (Wang and Scherer, 2016; China et al., 2018). Anti-inflammatory and antiapoptotic effects were demonstrated as major physiological activities of adiponectin as well (Ohashi et al., 2010). Adiponectin binds to two seven-transmembrane domain receptors, AdipoR1 and AdipoR2. Interestingly, unlike the well-known G-protein-coupled receptors, the N-terminus is located inside the cell, whereas the C-terminus faces outward for both AdipoR1 and AdipoR2. AdipoR1 was found abundant in skeletal muscle and the liver via ubiquitous expression, while AdipoR2 was isolated mostly from the liver (Kang et al., 2009). T-cadherin, highly expressed in endothelial and smooth muscle cells, was identified as a third adiponectin receptor (Matsuda et al., 2015).

In light of the inverse relationship between serum adiponectin levels and fat mass, the inverse relationship between bone marrow fat and bone mass inspired researchers to focus on the effects of adiponectin on bone regeneration. In vitro, adiponectin was reported to increase the mRNA expression of alkaline phosphatase (ALP) in preosteoblasts and promote the mineralization of the bone matrix (Williams et al., 2009; Naot et al., 2017). Furthermore, in a mouse model, the elevated adiponectin in the bloodstream significantly increased the volume of cancellous bone (China et al., 2017). Interestingly, an inverse correlation was demonstrated in clinical studies between serum adiponectin concentrations and bone mineral density (BMD) (Napoli et al., 2010). AdipoR1 and AdipoR2 were found to be expressed in primary human osteoblasts and in bone marrow macrophages, which could be the possible reasons for adiponectin playing a significant role in bone regeneration (Berner et al., 2004; Wu et al., 2019). Contradictory results in the literature were demonstrated as well; therefore, more evidence is needed to further clarify the physiological role of adiponectin in bone biology.

Although the endocrine effects of the secreted protein adiponectin from adipose tissue into the circulation account for the energy homeostasis, its local paracrine effects may play a pivotal role in bone regeneration (Martinez-Huenchullan et al., 2020). Here, we recombined human globular adiponectin (ADPN) and further characterized the pharmacokinetic behaviors and toxicity through medulla injection. A rat tibia fracture model was exploited to evaluate the capability of ADPN for bone repair. In addition, we attempted to unravel the underlying mechanism of adiponectin promoting bone regeneration.

Bone regeneration and bone fracture healing remain great challenges in daily clinical practice. Alternative therapies are being developed, including the FDA-approved BMP-2 and BMP-7 treatments (Jo et al., 2015), to overcome the adverse effects of autologous bone transplantation. Unfortunately, supraphysiological dosage, short half-life time, and extremely high cost significantly limit the applicability of these protein therapies.

Adiponectin, a protein hormone produced primarily in adipose tissue, is secreted into the bloodstream and is very abundant in plasma (5–10 mg/L), accounting for approximately 0.01% of all plasma proteins, compared with many other hormones (Nien et al., 2007). In the last 20 years, the physiological functions of adiponectin in whole-body energy homeostasis have been well documented, particularly the connections with obesity, diabetes mellitus, and atherosclerosis (Li et al., 2009; Yamauchi and Kadowaki, 2013). An interesting finding was the inverse relationship between adiponectin levels and fat mass, which distinguished adiponectin from other adipokines, such as leptin (Yadav et al., 2013). Fat and bone tissues can crosstalk with each other through hormones and cytokines to maintain their balance. In 2004, adiponectin and its receptors, AdipoR1 and AdipoR2, were reported to be present in human osteoblasts (Berner et al., 2004). In addition, the supplementation of culture medium with adiponectin enhanced cell proliferation of mice (Kanazawa et al., 2007). Thereafter, more attention has been garnered on the activity of adiponectin in bone. Numerous studies in vitro, in vivo, and in clinical settings were carried out to clarify the role of adiponectin in bone physiology (Williams et al., 2009). Adiponectin enhances osteoblast proliferation and differentiation concurrently with the inhibition of osteoclastogenesis in vitro (Yang et al., 2019); however, an inverse relationship between serum adiponectin concentrations and BMD was dominantly demonstrated in clinical studies (Richards et al., 2007). The most inconsistent results reported in the literature were obtained from different animal models. The variations are most likely due to the different forms of intercellular signaling, including paracrine effects of adiponectin produced in bone marrow adipocytes, endocrine effects of adiponectin released from white adipose tissue into the bloodstream, and second-order effects from the balance of whole-body energy (Naot et al., 2017).

We particularly focused on the relationship between adiponectin and bone fracture repair in this study, to extend our understanding on the paracrine effects of ADPN through medulla injection. First of all, a facile and efficient approach was designed to obtain purified recombinant human adiponectin. The endogenous glutamine synthetase gene in CHO-K1 cells was selected to be knocked out with the CRISPR technique to speed up the screening process. High purity protein (20 μg/ml) was yielded with our more optimal preparation method. The lyophilized protein powder in glass vials was easy for storage and further usage. The pharmacokinetic behaviors of the derived adiponectin were further evaluated. The C_max reached a maximum after 8 h via medulla injection of ADPN. Longer retention in the bone marrow of ADPN was achieved with intrabone marrow injection, which could provide extended bioavailability. Local administration of ADPN remarkably promoting the growth of callus and bone healing was confirmed in the rat tibia fracture model. With longer duration of local action, ADPN induced hyperplasia, which extended to the non-fractured fibula, indicating, to some extent, the osteogenic ability of local ADPN. This hyperplasia may result from the high dosage of local ADPN and the leakage of the local injection of ADPN to the fibula. This phenomenon warns that the proper dosage and duration of locally administrated ADPN are crucial factors for bone fracture treatment in practice. Biosafety and non-toxicity are of importance for every therapy. Hematology and serum biochemistry tests confirmed the biocompatibility of ADPN for medulla injection.

Local administration of ADPN promoting bone formation in our study mainly results from enhancing osteogenic differentiation. In BMSCs, the ADPN promoted osteogenic differentiation through its receptor AdipoR1 to increase the expression of osteogenic markers ALP, BMP-2, and OCN. In addition, the ADPN also presented the ability of enhancing proliferation up to 30 μg/ml in vitro. The addition of ADPN can increase the expression of AdipoR1 in BMSCs considerably, while knocking down of AdipoR1 with ADPN loses their osteogenic ability. Our observations coincide with the results reported in the literature that adiponectin could decrease the number of osteoclasts and improve bone healing via the OPG/RANKL pathway (Luo et al., 2006). A potential mechanism for bone regeneration promoted by recombinant human adiponectin was preliminary investigated; however, systemic signal transduction process should be thoroughly examined to unravel the underlying mechanism for the bioactivity of adiponectin in bone regeneration. Heterotopic ossification is the process by which bone tissue forms outside of the skeleton. Heterotopic hyperplasia was observed in the higher-dosage ADPN group in our study, which could have a certain extent of influence on the μCT data, however, it did not alter the conclusion of the ability to enhance osteogenesis and bone regeneration of ADPN. In our future experiment, this phenomenon of hyperplasia is planned to be further investigated.

In conclusion, we provided a facile method to obtain high-purity and high-yield ADPN for bone fracture treatment, which presented great biocompatibility as well as efficacy for improved bone healing via medulla injection. ADPN plays a potential significant role in stimulating the expression of ALP, BMP-2, and OCN at the fracture site through increasing the expression of its own receptor AdipoRl. A dosage of 1 mg/kg of ADPN was optimal to accelerate bone fracture healing in our experimental setting. Our findings demonstrated that local fracture treatment with ADPN could be a useful therapeutic option to shorten healing time and potentially be rapidly translated into the clinics.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

The animal study was reviewed and approved by the Ethics Committee of the Chinese People’s Liberation Army General Hospital, Beijing, China.

YG and Z-KC conceived the ideas for the experimental designs, analyzed the data, and wrote the manuscript. YG, YW, YQZ, LCW, LJW, and YAZ conducted all the experiments and helped with manuscript preparation. CL provided suggestions and revised the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.