Diedaqili Tablet Promotes Bony Fusion Accelerates Fracture Healing In Mice

Diedaqili tablet has been found effective for fracture healing in previous clinical studies. In a recent study, we investigated the effects of diedaqili tablet (DDQL) on bone fracture repair in mice. Adult C57BL/6 mice were subjected to transverse femoral fractures and administrated orally with a low dose (40.55mg/ ml), a high dose (162.18mg/ml) DDQL suspension and normal saline daily from day 1 after operation. The femur and blood of mice was analyzed by plain radiography, micro-computed tomography (Micro-CT), histology, biomechanical analysis, and serum Ca, P, and ALP test. The results demonstrated that DDQL can effectively improve the porosis of fracture end and bony union in the course of healing via Micro-CT and hematoxylin-eosin staining (HE staining) analysis. Consistent with morphological ndings, biomechanical properties of fracture healing have also been demonstrated. And Diedaqili tablet has a dose-dependent effect. DDQL augmented the release of alkaline phosphatase (ALP) and phosphorus (P) into blood, indicating that it promoted mineralization of hypertrophic cartilage and woven bone growth simultaneously during bone healing. In summary, the preliminary experiment revealed that DDQL can improve bone formation via promoting osteogenic capability, calcium (Ca) and P metabolism, and subsequently accelerates fracture repair and bony fusion.

Thus, this experimental study was designed to explore the effect of DDQL on fracture healing, so as to provide the theoretical basis for clinical effect of DDQL on fracture healing.

Experimental Animals
A total of 100 speci c pathogen-free (SPF) 8-week-old C57BL/6J male mice with a body weight of 20±2g were obtained from Beijing Vital River Laboratory Animal Technology Co. Ltd. The experimental animals were kept in the room which is SPF grade in Animal Experimental Center of Shanghai University of Traditional Chinese Medicine. The laboratory environment was as follows: indoor temperature 20-23℃, relative humidity 40-60%, noise control below 60 dB, light duration of 12 hours of light and 12 hours of night. The feeding room was disinfected once a day, and the pasteurized material was replaced twice a week. The feed was in accordance with the national standard of dry feed for rodents, and was fed  Fig.1 provides an overview that the animal model of fracture was established, and the process was as follows. The mice abstained from drinking and fasting the night before modeling. Mice were anesthetized with 10% chloral hydrate (Sinopharm Chemical Reagent Co. LTD) at a dose of 0.004 ml/g body weight. After shaving and sterilization of both legs, the shaft of the femur was exposed and cut off with a bone saw machine MDJ-110029 (MYDAJAR Co. LTD). The special 0.5mm intramedullary needle was inserted into the femoral bone marrow cavity through the intercondylar fossa of the femur until sensing loss sensibility which indicated the needle had entered the femoral bone marrow cavity. Finally, the wound was washed with normal saline. The mice after modeling were performed small animal X-ray imaging through Faxitron X-ray MX20 (Faxitron, Inc.) immediately to con rm the fracture line in good position and alignment indicating that the modeling was successful, and injected Gentamicin (Shandong Zhengmu Biological pharmaceutical Co. LTD) continuously for 3 days to prevent infection. In the sham group, only the middle femur region was exposed, and then the muscle, fascia and skin were directly sutured layer by layer with 4-0 line.

Drug Administration and Specimen Harvest
DDQL used in animal administration was purchased from Chongqing Xieran Pharmaceutical Co. LTD. The tablets were crushed into powder by a Chinese medicine crusher. The powder was blended with 0.3% sodium carboxymethyl cellulose solution (CMC-NA) and prepared into low dose (40.55mg/ml) and high dose (162.18mg/ml) DDQL tablet suspension, respectively. DDQL was given to mice by daily oral gavage from the day after modeling until euthanasia, and mice in the control group were fed with normal saline only.
Animals in each group were subject to serological detection, Micro-CT, histology and biomechanics. In week 1(W1), week 2(W2), week 3(W3) and week 4(W4) post-operation, arterial blood of mice was collected from abdominal aorta and stood at room temperature for 30 minutes. Blood was set in a centrifuge at 3000r/min for 20min to extract the yellowish serum from the upper layer and stored in a refrigerator at -80℃. Fractured and contralateral femurs were harvested at the end of experiments in W2 and W4. Soft tissues were removed from the operated and contralateral femurs. The specimens were xed within 4% paraformaldehyde solution for 24h. Internal xation needle was removed from the intramedullary cavity and the femur specimens were stored in 75% ethanol for Micro-CT and histologic processing.

Serum content test
The collected serum samples were restored to room temperature and placed in Beckman Coulter AU5800 Automatic biochemical instrument (American Beckman Coulter Co., LTD) to detect the contents of Ca, P and ALP in serum.

Histology
Fixed bone specimens were decalci ed with calcium chelating solution (0.5 M EDTA/NaOH, pH 7.5) for 2 weeks. Decalci ed bones were then dehydrated and embedded in para n wax using Leica EG Embedding Center (Leica Microsystem, Wetzlar, Germany). Para n blocks were sectioned into 5μm slices and mounted on glass slides. The sections were de-para nized and stained with hematoxylin and eosin (HE) for the bone tissue. The histomorphometric analysis was performed by a blinded observer using Olympus BX43 microscope (Olympus Corporation,Japan).

Biomechanical
The femur specimens of mice were subjected to Instron 5543 type material mechanics tester (Instron Shanghai Material Testing Machine Co., LTD,China) for biomechanical three-point bending test. The program is set as temperature 23℃, humidity 60%, speed 5.00mm/min, span 14mm. The detection index is Maximum Load (N), which refers to the maximum bending load borne by the whole bone before damage, indicating the strength of the healed bone directly re ect the biomechanical properties [[ii]].

Micro-CT
The femur specimens were scanned in Skyscan1172 type Micro-CT instrument (Belgium Bruker Technology Co., LTD) for 3D imaging analysis. The scanning parameters were as follows: voltage 56KV, current 171μA, scanning row number 666, scanning column number 1000, object to source 59.970mm,camera to source 214.620mm, vertical object position 39.110mm, exposure time 360ms, image pixel size 9.92μm.
After 3d reconstruction by Micro-View software, the region of interest (callus region) was selected on the femoral cross-section. The region of interest was centered on the fracture line and extended 5mm to the proximal and distal ends of the femoral axis. Bone volume fraction (BVF), that is to say BV/TV, in the region of interest was calculated and analyzed, which referred to the ratio of the volume of mineralized callus to the volume of total callus in the area of interest, re ecting the content of callus tissue at the fracture end [[iii]].

Statistical methods
Statistical analysis was conducted using GraphPad Prism 5.0 (GraphPad Software, CA, USA) software and Excel (Microsoft, CA, USA). All data was obtained from 5 individual mice in each group. Mean and standard deviation values (mean±SD) were calculated for all statistically analyzed parameters. The differences between groups were analyzed using ANOVA with Turkey's post hoc test or unpaired Student's t tests. The p value less than 0.05 was considered statistically signi cant.

Micro-CT analysis
To study the effect of DDQL on fracture repair in mice, the fracture morphology of mice in W2 and W4 were monitored and analyzed via Micro-CT 3D imaging. As is shown in Fig.2, Micro-CT allows for visual evaluation of callus at the fracture end through 3D reconstruction, and quantitative comparison through the software analysis. In W2, the fracture end was surrounded by newly formed soft callus in each group, among which the callus volume was signi cantly less in the model group than both DDQL groups. In comparison with the model group, the low dose group had more minor fracture end area and fresh callus than the model group, but the fracture end was still not connected by brous tissue. The high-dose group had smaller gap at the fracture end, more newly formed callus, which revealed a current of bone junction.
In W4, though the new callus of the model group joined the fracture end, the callus quantity was less and uncalci ed. Compared with the model group, low dose group had more calluses and slight calci cation, and the fracture ends were completely surrounded by calci ed bone and connected surrounding bone cortex in the high dose group. Although the three-dimensional image can only show the size of callus, the bone volume fraction can directly verify the quanti cation of callus by DDQL. As shown in Fig 3 and Table 1,in W2,the BVF of the low-dose group was higher than that of the model group, but there was no signi cant difference (P 0.05).The BVF of the high-dose group was higher than that of the model group and the low dose group, and the difference between the high-dose group and the model group was statistically signi cant (P 0.05).There was no signi cant difference between the high dose group and the low dose group (P 0.05). In W4,the BVF of the low dose and high dose groups were both higher than that of the model group, and the difference was statistically signi cant (P 0.05). The high dose group was higher than that of the low-dose group, and the difference was statistically signi cant (P 0.05).

HE staining
Since radiographic evaluations were not able to ascertain cartilage and other soft tissues, histological sections of fractured bone were labeled with both cartilage and bone and histomorphometric analysis were performed [[i]]. As shown in Fig 4, the fracture ends of the model group were not connected which was same as Micro-CT analysis, there were gaps surrounded by a large number of necrotic bone fully unabsorbed in W2. A small amount of chondrocyte in ltrated around the fracture ends, and the brous bone was sparse. Chondrocyte in ltration and calci ed braided bone mass were increased around the fracture end in DDQL groups, and the healing trend was more signi cant in high dose group. In W4, the fracture ends of the model group were still barely union though a large amount of brous bone tissue appeared around the fracture ends. However, the fracture ends of the low-dose and high-dose groups were all connected, and the brous bone was transformed into lamellar bone, gradually restoring the basic form of bone.

Biomechanical analysis
As shown in Fig 5 and Table 2, no statistically signi cant difference was observed between the low-dose group and the model group in W2(P > 0.05). Statistically signi cant differences were observed between the high-dose group and the model group, and the femur maximum load of the high-dose group was higher than the model group (P < 0.05). Though the femur maximum load of the high-dose group was higher than that of the low-dose group, statistically signi cant difference was not observed (P 0.05). In W4, the femur maximum load of each group was compared, and both the femur maximum load of lowdose and high-dose group were higher than that of model group (P < 0.05). Though the high-dose group was higher than the low-dose group, the difference was not statistically signi cant (P > 0.05).

Serum content test
At each point after modeling, the serum Ca,P and ALP content of mice in sham operation group and control group was compared, and no signi cant difference between the two groups was observed(P > 0.05), which proved sham operation had no impact on the serum Ca, P and ALP contents in mice.
As shown in Fig 6 and Table 3, the serum Ca content of mice in model group, low-dose group and highdose group increased in W1, but there was no signi cant difference (P > 0.05). From W2 to W4, the serum Ca content in model group, low-dose group and high-dose group all showed a diminished trend. In W2, the serum Ca content in low-dose group was signi cantly different from that in sham operation group (P < 0.05), while there was no signi cant difference in other groups (P > 0.05). In W3, the serum Ca content in the high-dose group was signi cantly different from that in the sham group (P < 0.05), but there was no signi cant difference in the other groups (P > 0.05). In W4, the serum Ca content of model group was signi cantly different from that of sham operation group (P < 0.05), but there was no signi cant difference in other groups (P > 0.05).
As shown in Fig 7 and Table 4,the serum P content in model group, low-dose group and high-dose group showed a clear trend of increasing within 28 days after modeling, and reached a peak in W3 after modeling.The serum P content of the high dose group was signi cantly higher than both of the sham group and the model group in W1(P < 0.05), but no statistically signi cant correlation was observed between the other groups (P > 0.05). In W2, compared with the sham operation group, the serum P content of model group, low-dose group and high-dose group was signi cantly different (P < 0.05). The serum P content in low dose group and high dose group was higher than the model group,and there was a signi cant difference (P < 0.05). No signi cant difference in serum P content was found between high dose group and low dose group in W2 (P > 0.05). In W3,ompared with sham operation group, the serum P content of model group, low-dose group and high-dose group was signi cantly different(P < 0.05). The serum P content in low dose group and high dose group was higher than the model group,there was a signi cant difference between each group(P < 0.05). The serum P content of the high dose group was higher thah that of the low-dose group with a signi cant difference(P < 0.05).In W4, the serum P content of model group and high-dose group were higher than the low-dose group,and high dose group was higher than sham group and model group, there was a signi cant difference between each group(P < 0.05).
As shown in Fig 8 and Table 6, The serum ALP content in model group, low-dose group and high-dose group showed an increasing trend within 28 days after modeling, and reached a peak in W3 after modeling. In W1, compared with the sham operation group, the serum ALP content in the high-dose group was higher, and the difference was signi cant (P < 0.05), while the serum ALP content in the other groups had no signi cant difference (P > 0.05). In W2, the serum ALP content of low-dose group and high-dose group was higher than that of sham operation group and model group, and the high-dose group was higher than that of low-dose group, the difference was signi cant (P < 0.05), while the serum ALP content of other groups had no signi cant difference (P > 0.05).In W3, the serum ALP content of low-dose group and high-dose group was higher than that of sham operation group and model group, and the high-dose group was higher than that of low-dose group, the differences were signi cant (P < 0.05), while the serum ALP content of other groups had no signi cant difference (P > 0.05). In W4, the serum ALP content of low dose and high dose was higher than that of sham operation group and model group, with signi cant difference (P < 0.05), while the serum ALP content of other groups had no signi cant difference (P > 0.05).

Discussion
Fracture healing is a complicated dynamic process involving histology, biology, endocrinology and biomechanics, which has always been a key issue in the field of orthopedics [[i]]. About 10% of fractures each year end up as nonunion or delayed union, requiring additional surgery including bone grafting, internal fixation and external fixation [[ii]]. Callus formation and bone resorption depend on the integrity of blood supply and soft tissue, which is general factors for nonunion of fractures [[iii]]. The study has found that the use of rigid plates and reamed nails can affect the vascularization of bone fragments, resulting in high infection rates and delayed union [[iv]]. Although autologous bone transplantation is still the gold standard of bone regeneration, studies found that autologous bone transplantation may give rise to nerve injury, wound infection, postoperative persistent pain and secondary fracture and other adverse reactions [[v]]. Safer treatment with cost-effective and simple way of administration are thereby still required.
Our HE pathological staining and Micro-CT 3D imaging analyses demonstrated the healing process of femur fracture in mice. Micro-CT, as the gold standard for the evaluation of bone trabecular structure, can measure the structure and mineralization of fracture callus in a noninvasive, quantitative and three-dimensional manner[[vi] -[vii]] . As a simple and lowcost method, HE pathological staining has been widely used in bone morphological assessment. In W2, the callus formation in the high-dose and low-dose groups was significantly better than that in the model group, and the callus content in the high-dose group was significantly higher than that in the low-dose group. In W4, compared with the model group, the callus in the low-dose group was more and showed a slight tendency of calcification, while the fracture end in the high-dose group was completely filled with calcified new bone and associated with the surrounding bone cortex. BVF at the fractured end of mice analyzed by micro-CT detection, refers to the ratio of the volume of mineralized callus to the total callus in the area of interest, reflecting the content of callus tissue at the fractured end, which is relevant to the stiffness and strength of callus [[viii]]. The BVF of low dose group and high dose group were both higher than that of the model group in W2 and W4, which indicated that DDQL tablet could improve bone strength by increasing the content of callus. Moreover, the callus content in the low dose group was significantly lower than that in the high-dose group, indicating that DDQL may have a dose-dependent effect on the degree of callus formation. It verified that diedaqili tablet can effectively promote the formation of callus around the fracture end in mice and shorten the time of fracture healing.
The recovery of bone strength and stiffness is the basic characteristic of fracture healing. Biomechanical testing can be used to evaluate the recovery of bone strength and stiffness. Currently, the main detection methods of biomechanics include three-point bending test, four-point bending test, torsion test and tensile test, etc., and the widely used three-point bending test was selected in this study [[ix]]. According to the Stat theory of mechanics, the bone geometry is suitable to ensure that the maximum tissue strain from the universal load is kept within a certain range, and when this strain capacity is exceeded, the bone tissue will fracture [[x]]. The Maximum Load (Max Load) tested by the three-point bending test refers to the maximum bending load borne by the whole bone before it is destroyed, which indicates the strength of the healed bone and can directly reflect the biomechanical properties. The study observed that the Max Load of femur in both dose group was better than that in the model group in W2 and W4, while the improvement of low dose group in W2 was not obvious.The biomechanical results were consistent with the morphological findings, indicating that DDQL tablet could not only improve the bone morphological structure of mice femur significantly, but also enhance bone strength and stiffness in fracture healing.
Bone strength can reflect the integrity of bone quality, determined by both bone microstructure and bone mass. Calcium salt deposition is one of the key steps in fracture repair. Bone tissue repair can only be completed by the calcification of bone matrix, and the strength of new bone tissue is closely related to the degree of matrix calcification [[xi]]. The mineralization and mechanical properties of bone depend largely on the deposition levels of Ca and P. Ca and P exist in the blood in the form of ions, which are interactive systems [[xii]]. The normal operation of osteogenesis and bone dissolution is an important link to maintain the stable content of calcium and phosphorus in the blood [[xiii]]. The experimental results of this study showed that in W1 after fracture modeling, serum Ca content of mice showed an increasing trend, which may be caused by the release of a large number of Ca ions into the blood due to the dissolution of bone fragments by osteoclasts. From W1 to W4 after modeling, serum P content increased and peaked in W3, and the serum P content in both high-dose and low-dose groups was higher than that in the model group, indicating that osteogenesis was active and phosphate release increased. From W2 to W4, serum Ca decreased significantly after fracture modeling, suggesting that the deposition of calcium and phosphate in serum increased, which was in the state of osteogenic calcification. These results were consistent with previous studies, however, experimental results showed that DDQL tablets at low and high doses did not significantly change serum Ca in fracture mice.
The activity of ALP synthesized by the osteoblasts is the typical markers of bone turnover, which catalyzes mineralization and bone formation making osteogenesis possible[[xiv] -[xv]] . As the osteogenic differentiation of cells progresses, its activity gradually increases [[xvi]].The increase of ALP in each model group within 28 days after the modeling could indicate intensified bone tissue remodeling, which reached the peak in the third week, same as the tendency of serum P content. And the content of ALP in high dose group was obviously higher than that in low dose group and model group, demonstrated DDQL tablet can improve the content of ALP, increase osteogenic activity, promote phosphate release, and increase calcium deposition to promote fracture healing of femur in mice, with a dosedependent effect [[xvii]].
In conclusion, this study demonstrated that DDQL tablet has an effect of improving bone formation through promoting osteogenic capability, calcium and phosphorus metabolism. We believe that DDQL tablet is a safe, effective and economical oral drug for fracture healing in the long term. Tables   Table 1 Comparison of BV/TV among each group at each point. The data of three groups was analyzed by One Way ANOVA.* represents the difference was statistically signi cant compared with the model group. △ represents the difference was statistically signi cant compared with the low-dose group.   Comparison of mice femur among each group in W2 and W4 after fracture via Micro-CT 3d imaging Comparison of maximum load among each group in W2 and W4 after fracture Figure 6 Page 18/19

Declarations
Comparison of serum Ca content among each group at each point after fracture Comparison of serum P content among each group at each point Figure 8