Site-activatable targeting of macromolecular alendronate for accelerated fracture healing

ABSTRACT Sustainable social activity is a major goal in an aging society, although this is limited by loss of athleticism, with osteoporosis-related fractures being the most common cause of long-term behavioral restrictions in older people. Therefore, the development of therapeutics that shorten the duration of fracture therapy is essential to improve the quality of life and social activity of older individuals. In this study, we developed a polyethylene glycol-modified alendronate (PEG-ALN) that can efficiently deliver the active ingredient (ALN) to fracture sites. PEG-ALN released ALN in response to an acidic pH and was systemically administered to mice in a fracture model. PEG-ALN exhibited selective accumulation at the fracture site and significantly accelerated bone healing compared to free ALN. This study highlights the utility of a simple polymer modification of ALN as a systemically injectable medicine for patients with bone fractures. GRAPHICAL ABSTRACT IMPACT STATEMENT Bone fractures present a significant risk to the health and quality of life of older individuals in aging societies. However, there are limited options for direct pharmacological treatment of fractures. In this study, we optimized a commonly used but therapeutically challenging drug, alendronate (ALN), to target and improve fracture healing. We modified alendronate with high molecular weight polyethylene glycol (PEG) through an acidic-responsive linker (PEG-ALN) and assessed its effect on osteoblasts and osteoclasts, and thereby on bone healing, both in vitro and in vivo. This PEGylation prevented the rapid clearance of ALN in a mouse fracture model and enabled specific accumulation of PEG-ALN at fracture sites. The accumulated PEG-ALN is cleaved by the acidic environment of these osteoclastic regions to release active ALN directly where required. Consequently, the fracture region is converted to an osteoblast-dominated environment with increased recovery of bone fractures. Thus, our study demonstrated that PEG-ALN provides an effective delivery system for ALN to fracture sites, enabling rapid healing without apparent side effects. We believe that the findings of this study are relevant to the scope of your journal as they address the biomedical use of materials that could enhance debilitating conditions that will become more frequent in a rapidly aging population.


Introduction
A rapid increase in the number of patients suffering from fractures is expected along with the continuous global growth in the aging population [1][2][3][4].According to the International Osteoporosis Foundation, an estimated nine million patients suffer from osteoporosis-related fractures [3].Therefore, it is critical to reduce the number of patients with bone fractures and the treatment time for such fractures to improve the quality of life and social productivity of older individuals.Bone fractures require long-term medical treatment, are the most common injuries in the orthopedic field, and are associated with considerable socioeconomic burden on patients.Bone fractures in an older person can leave them in a weakened state, with them requiring nursing care because of difficulty walking and/or being bedridden due to the marked deterioration of physical functions.Osteoporosis occurs frequently in older individuals and increases the risk of fractures.Although substantial efforts have been made to develop antiosteoporosis agents, the aim of current osteoporosis treatments is to prevent fractures by increasing bone strength.Further fracture prevention through early medical intervention has been actively promoted in patients with osteoporosis; however, the development of direct pharmacotherapy against fractures is currently lagging.Additionally, external fixation using a plaster cast or surgical osteosynthesis is considered the traditional method of fracture treatment [5].However, these therapeutic strategies require longterm treatment, and patients are at risk of serious complications such as fracture malunion.Autogenous bone grafting is already being used to repair massive bone defects in orthopedic and maxillofacial surgeries.However, issues regarding limited bone mass, high invasiveness of the harvested tissue, and patient age and health remain [6][7][8].Therefore, recent studies have been conducted on bone regeneration by regenerative medicine using scaffold materials, bone replacement materials, and bone marrow mesenchymal stem cells (BMSCs) for massive bone defects [9][10][11][12][13][14][15][16][17][18].Although regenerative medicine is an attractive therapeutic technology, it suffers from problems such as implantation-related tissue invasion, safety risks, and high cost, and is therefore difficult to apply to simple fracture treatments as opposed to significant bone defects.
Therapeutic agents, including bisphosphonates, estrogens, and selective estrogen receptor modulators, are being explored for fracture healing [19][20][21][22][23][24][25][26][27].For instance, bisphosphonates, which induce apoptosis of osteoclasts and prevent fragility fractures, promote endochondral ossification by inducing the formation of calli, trabecular bone, and bone minerals.However, delays in maturation and remodeling of the callus have been identified as a side effect in animals [23,26,27].Bisphosphonates have a short blood half-life and low bioavailability, and multiple doses are required to maintain therapeutic efficacy, thereby increasing the risk of delayed callus maturation and remodeling.
Polyethylene glycol (PEG) can used as a pharmaceutical additive (e.g. in chemical compounds, proteins, peptides, and oligonucleotides) and can improve the blood retention and bioavailability of drugs by suppressing rapid renal excretion and reducing antigenicity [28][29][30].The advantages of PEGylation are most likely due to the antifouling properties of PEG.Meanwhile, little attention has been paid to PEGylated drugs for fracture treatment, and development in this field could provide therapeutic benefits to orthopedic patients by altering the administration route, biodistribution, and bioavailability of established drug treatments.
In this study, we synthesized PEGylated alendronate (PEG-ALN) through a pH-responsive linker for the treatment of systemic fractures (Figure 1).Similar to bisphosphonates, alendronate (ALN), a second-generation bisphosphonate analog of pyrophosphate, contains a primary amine group.Based on the structure-activity relationship [31][32][33], this primary amine plays an important role in expressing the therapeutic effect, i.e. the phosphorus-carbonphosphorus (P-C-P) moiety of ALN has a high affinity for the calcium ions of hydroxyapatite present in the bone tissue.In addition, ALN containing a primary amine group exhibited 1000-to 10,000fold higher biological activity than first-generation bisphosphonates.However, its short half-life in and lack of targeting ability need to be improved upon for its use in fracture treatment as an injectable drug.Thus, PEG-ALN was designed to actively accumulate in the bleeding lesion at the fracture site and adsorb onto the fractured bone through systemic injections (Figure 1(a)), providing the accelerated fracture healing (Figure 1(b)).In fact, the accumulated PEG-ALN at the fracture site reduced the number of osteoclasts via the induction of apoptosis in the early stages of fracture healing and promoted the therapeutic effects by replacing the fracture site with an osteoblastdominated environment.PEG-ALN reduced the number of osteoclasts in both the inflammatory and repair phases and enhanced callus formation by osteoblasts.Furthermore, almost all the accumulated PEG-ALN was consumed during the repair phase, and therefore did not affect the maturation and remodeling of the newly formed callus.There are currently no pharmacological treatments for fractures based on biological responses; therefore, this is the first report on the development of a short-term fracture treatment method using a drug delivery system (DDS).
To prepare fluorescent dye-labeled PEG-ALN, N 3 -PEG-NH 2 was prepared from HO-PEG-NH 2 as shown in Supplementary Figure S1(a).Briefly, the amine group on HO-PEG-NH 2 (1.0 g, 0.025 mmol) was protected using (Boc) 2 O (109 mg, 0.5 mmol) in the presence of TEA (25 mg, 0.25 mmol).The reaction mixture was stirred at RT for 24 h and then poured into excess hexane.The precipitate was collected by filtration, and the solid material was washed with benzene and freeze-dried to obtain HO-PEG-Boc.A mixture of methanesulfonyl chloride (MsCl, 30 μL), HO-PEG-Boc (892 mg, 0.22 mmol) and THF (2 mL) was stirred under an inert atmosphere at RT for 24 h.Residual Ms-PEG -Boc was collected and dried in vacuo.Sodium azide (15 mg, 0.24 mmol) was added to the Ms-PEG-Boc/DMF (10 mL) solution under an argon atmosphere and stirred for 3 days at RT.After quenching with DW, the product was extracted with DCM.Re-precipitation was performed using a diethyl ether/DCM system to obtain N 3 -PEG-Boc.The Boc groups were deprotected using a mixture of TFA (5 mL), DCM (5 mL), and DW (25 μL).The reaction mixture was stirred in the dark for 3 h at RT without light.The resulting solution was dialyzed (MWCO = 3,500) against a methanol solution and DW, followed by freeze-drying to afford N 3 -PEG-NH 2 as a white powder (80% yield).N 3 -PEG-CDM and N 3 -PEG-ALN were prepared in almost the same manner as that described above.N 3 -PEG-CDM was identified by 1 H-NMR (98% yield; Supplemental Figure S1(c)).N 3 -PEG-ALN was identified by 31 P-NMR (61% yield); 31 P -NMR peaks: δ (ppm) = 19 (Supplemental Figure S1(d)).The introduction rate of ALN into N 3 -PEG-CDM was calculated in the same manner as described above.Cy5 fluorescent dye-labeled PEG-ALN (Cy5-PEG-ALN) and Cy5-PEG were prepared via a copper-free click chemistry system.Briefly, N 3 -PEG-ALN (46 mg, 1.14 × 10 −3 mmol) or N 3 -PEG-NH 2 (20 mg, 4.91 × 10 −4 mmol) were dissolved in 2 mL of D-PBS and stirred with 93 μL (1.2 mg, 1.2 × 10 −3 mmol for N 3 -PEG-ALN) or 40 μL (0.5 mg, 5.0 × 10 −4 mmol for N 3 -PEG-NH 2 ) of Cy5-DBCO /DMSO solution (13 mg/mL), respectively.The reaction mixtures were stirred overnight at RT, purified using a PD-10 column, and dialyzed (MWCO = 3,500) to obtain the Cy5-labeled polymers.The fluorescence activity of each polymer was confirmed using a Spark Multimode Microplate Reader prior to intravenous injections.
Radio-iodinated polymers were prepared using the Bolton -Hunter method [34] before use in the pharmacokinetic studies.The N-hydroxysuccinimide ester of iodinated p-hydroxyphenylpropionic acid (Bolton -Hunter reagent) was purchased from Parkin Elmer, Inc., MA, U.S.A.. Before radio-iodination, the azide terminus of N 3 -PEG-ALN was converted to an amine group using dithiothreitol (DTT).Briefly, N 3 -PEG-ALN (2 mg, 5.0 × 10 −5 mmol) and DTT (100 eq) were dissolved in NaHCO 3 buffer (pH 8.2, 100 mM).The mixture was then stirred overnight at RT.The reactant was purified by dialysis with NaOH (pH 10) aqueous solution and DW at RT for 18 h.Next, the Bolton -Hunter reagent (5 μL, 0.46 MBq) was added to the NH 2 -PEG-ALN or NH 2 -PEG solutions (100 μL of 10 mg/mL) in PBS (pH 7.5) and agitated at RT for 5 min.The mixture was incubated at RT overnight and purified using a PD-10 column to yield 125 I-labeled polymers.

pH responsivity of PEG-ALN
To confirm pH-dependent ALN-release, PEG-ALN (15 mg) was dissolved in a 0.4 M tri-sodium citrate solution.The pH was adjusted to 3.1 and 4.1, and then the solutions were incubated at 37°C for 48 h.To evaluate ALN-release from PEG-ALN, the reaction mixture was dialyzed against NaOH (pH 10) aqueous solution and DW at RT for 18 h.After freeze-drying, the release of ALN was determined using 31 P-NMR spectroscopy.

Cell culture
Osteoclasts cells (OSC14 cells), a murine bone marrow-derived osteoclast progenitor cell line, were purchased from Cosmo Bio Co. Ltd. (Tokyo, Japan).Cells were cultured on dentin slices using alpha-modified Eagle medium (α-MEM) containing 10% fetal bovine serum (FBS) (Biosera Inc., Metro Manila, Philippines), receptor activator of NF-κ B ligand (RANKL, 25 ng/ mL), macrophage-colony stimulating factor (M-CSF, 50 mg/mL) and penicillin-streptomycin (Merck KGaA, DA, Germany, 100 units/mL of penicillin and 100 μg/mL of streptomycin), and grown in an atmosphere of 5% CO 2 and 37°C humidified air.KUSA-A1 cells (JCRB1119), a murine mesenchymal stem cell line derived from C3H/He mice, was provided by the Japanese Collection of Research Bioresources Cell Bank through the National Institutes of Biomedical Innovation, Health and Nutrition, Japan.KUSA-A1 cells were maintained in minimum essential medium (MEM) plus 10% FBS, GlutaMAX-I (Thermo Fisher Scientific Inc., Waltham, MA, U.S.A.) and penicillinstreptomycin.The other cell culture conditions were the same as those described above.

Osteoclast differentiation and pit formation assay
OSC14 cells (4 × 10 4 ) were seeded onto a dentin slice plate and cultured for three days.The medium was replaced with fresh osteoclast differentiation medium containing different concentrations of ALN or PEG-ALN three days after seeding (Supplementary Table S1), and continuously cultured for an additional three days.
Tartrate-resistant acid phosphatase (TRAP) staining was performed to assess osteoclast differentiation.TRAP staining was performed according to the manufacturer's instructions (TRAP Staining Kit; Cosmo Bio Co., Ltd., Tokyo, Japan).Briefly, the cells were washed three times with D-PBS before fixation.Absolute ethanol was added to the cells, which were then incubated for 20 min at RT for fixation.The cells were then washed three times with DW.A chromogenic substrate, a mixture of 7-bromo-N-(2-methoxyphenyl)-3-(phosphonooxy)-2-naphthalenecarboxamide and 4-benzamido-5-chloro-2-methylbenzenediazonium chloride (Fast Red Violet LB salt), was added to the cells.Cells were incubated for 60 min at 37°C, and then washed with DW.The stained cells were photographed on a BZ-X710 system using a 20× magnification objective lens.
Bone resorption was evaluated using the pit formation assay.Briefly, von Kossa staining was performed to evaluate bone resorption by the differentiated osteoclasts.OSC14 cells stimulated with different concentrations of ALN or PEG-ALN were washed thrice with DW.Silver nitrate solution (5%) was added to the cells, which were then incubated for 90 min under shaded conditions.After washing with DW, sodium thiosulfate solution (5%) was added and incubated for 2 min at RT.The stained cells were photographed on a BZ-X710 system using a 20× magnification objective lens.

Osteogenic cell differentiation
KUSA-A1 cells (5 × 10 4 /well) were seeded in 24-well plates in maintenance medium and cultured until semi-confluent.The medium was then replaced with an osteogenic differentiation medium comprising RPMI-1640 medium supplemented with Osteoblast-Inducer Reagent (Takara Bio, Inc., Shiga, Japan), 10% FBS, and penicillin-streptomycin.Three days after seeding, this medium was then replaced 3 days after seeding with fresh osteogenic differentiation medium containing different concentrations of ALN or PEG-ALN (Supplementary Table S1).Media were replaced 3 days after ALN stimulation, and the cells were continuously cultured for an additional 6 days.
Alizarin Red S staining was performed to evaluate osteogenic differentiation.Briefly, the cells were washed three times with D-PBS before fixation.PFA (4 wt%) in 0.1 M phosphate buffer solution was added to the cells, which were then incubated for 15 min at RT for fixation.The cells were washed three times with DW and stained with Alizarin Red S solution.The pH was adjusted to 6.3 using an ammonia solution for 3 min, and the cells were then washed with DW three times.Stained cells were photographed on a BZ-X710 system using a 20× magnification objective lens (KEYENCE Co. Ltd., Osaka, Japan).
The calcium concentration was quantified to evaluate the calcification effect of PEG-ALN.KUSA-A1 cells stimulated with different concentrations of ALN or PEG-ALN were washed with DW three times, and the calcification plaques were dissolved in 0.5 N HCl solution.Calcium content was quantified using the methylxylenol blue method (Calcium E-Test Wako, FUJIFILM Wako Chemicals Co. Ltd., Osaka, Japan).

Animal experiments
All animal experiments were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of Tokyo Institute of Technology (approval no.D2016013).Female C57BL/6J mice were purchased from Sankyo Labo Service Co. Inc. (Tokyo, Japan) at 8 weeks of age.All animals were anesthetized with pentobarbital (30 mg/ kg) administered intraperitoneally, and xylocaine (66 mg/kg) was injected locally before bone fracture model preparation.The in vivo experiments were performed using a previously reported surgical procedure with modifications [35].Briefly, a closed tibial fracture mouse model was prepared by traumatic three-point bending, and the fracture location was confirmed via X-ray imaging (μFX-1000; FUJIFILM Medical Co. Ltd., Tokyo, Japan).A plaster cast (Castlight-α; Alcare Co. Ltd., Tokyo, Japan) was applied to the fracture leg, and PEG-ALN, PEG and saline were intravenously injected.The tibia-fibula complex containing the fracture site was removed 4 weeks after the operation and fixed in 4 wt% PFA solution in 0.1 M phosphate buffer solution for the assessment of bone regeneration.

Pharmacokinetics of PEG-ALN
The pharmacokinetic performance of PEG-ALN in a mouse fracture model was assayed using Cy5 fluorescent dye-labeled polymers.Briefly, Cy5-PEG-ALN or Cy5-PEG (5 mg/mL, 100 μL in total volume) were injected into the fracture model mouse via the tail vein.Organs (100 mg) were collected at scheduled time points and lysed with radioimmunoprecipitation (RIPA) buffer using a medical bead shocker (Yasui Kikai Co., Ltd., Osaka, Japan).The amount of Cy5-PEG-ALN or Cy5-PEG in each organ was determined using a Spark Multimodal Microplate Reader (Tecan Group Ltd., Zürich, Switzerland).
The accumulation of Cy5-PEG-ALN was evaluated histologically.Three hours after the intravenous injection of Cy5-PEG-ALN (5 mg/mL, 100 μL in total volume), bone tissues were collected and embedded in O.C.T. compound (Sakura Finetek Japan Co. Ltd., Tokyo, Japan).Serial fresh-frozen sections (4-μm thickness) were prepared according to Kawamoto's Film Method [36].Tissue sections were stained with Hoechst 33,342 (Roche Diagnostics GmbH, Mannheim, Germany) and photographed using a BZ-X710 system with a 20× magnification objective lens.

Determination of the optimal concentration of PEG-ALN for fracture treatment
To optimize the concentration of PEG-ALN for bone tissue regeneration, different concentrations of PEG-ALN or ALN were injected into fracture model mice via the tail vein (Supplementary Table S2) or locally at the fracture site.A plaster cast was applied to the fractured leg after PEG-ALN injection.Osteoblasts at the fracture site were collected using cell extraction buffer at 1, 2, 3, and 4 weeks after injection.Briefly, the fractured bone was removed, and soft tissues around the bone were collected.Cell extraction buffer, which contains 13,940 U of collagenase 10,000 U of dispase I and 20% FBS, was added to the collected bone and then incubated at 37°C.After 1.5 h incubation, cells dispersed from the bone were collected.Flow cytometric analysis was performed with a polyclonal antibody against Runt-related transcription factor 2 (Runx2) (Abcam plc., Cambridge, UK) to detect osteoblast cells using the Guava® easyCyte TM Flow Cytometer (Merck Millipore Co. Ltd., Darmstadt, Germany).

Single dose toxicity of PEG-ALN
To evaluate the single dose toxicity, PEG-ALN (3.04 × 10 −5 mmol/mouse) or ALN (2.73 × 10 −9 mmol/ mouse) was administered to C57BL/6 mice (female, nonfractured, 6-week-old) via the tail vein or intramuscularly via a hind limb.Blood samples were collected from the abdominal aorta of each mouse 24-h post-injection, and blood chemistry parameters such as aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, creatine phosphokinase, blood urea nitrogen, creatinine, total protein, and total bilirubin were determined using the automated analysis system Fuji-Dry-Chem 4000 and Fuji-Dry Chem slides (Fuji Photo Film Co. Tokyo, Japan).

Micro-computed tomography (μCT) analysis
BMD was measured, and three-dimensional (3D) images of the regenerated bone in the tibial fracture were visualized using computed tomography scans (CosmoScan FX; Rigaku Co. Tokyo, Japan).Samples were continuously scanned in increments of 10-μm thickness for 2,500 slices with a voxel size of 50 × 50 μm 2 .Two-dimensional images were reconstructed and processed using CosmoScan software to produce 3D images of the regenerated bone.The BMD of each bone fracture was measured at the 50 × 50 × 2,000 μm3 region of interest.The instrument was calibrated automatically.

Histological analysis
Bone tissues were fixed with 4 wt% PFA in 0.1 M phosphate buffer solution overnight at 4°C and decalcified with 0.5 M EDTA solution for 1 week at 4°C.After decalcification, the samples were embedded in paraffin.Thigh specimens were cut into 4-μm thick longitudinal section slices.For histological analysis, sections were stained with H&E and MT.Images were obtained using a microscope (BZ-X710; KEYENCE Co., Osaka, Japan).

Statistical analysis
All statistical data are expressed as mean ± standard deviation.The data were analyzed by ANOVA to determine the statistical significance between the two mean values.Values with p < 0.05 were considered statistically significant.

Polymer preparation and characterization
Modification of the PEG terminus proceeded with a high yield, as previously reported [37].The endfunctionality of PEG with CDM and ALN installation was confirmed using proton-( 1 H-) and phosphorousnuclear magnetic resonance ( 31 P-NMR) spectroscopy (Figure S1(a-d)).For CDM modification, the quantitative reactions were assessed by comparing the peak areas in 1 H-NMR at 3.6, 2.9 and 2.8 ppm (Figures 1(d) and S1(b,c)).The results of 31 P-NMR revealed successful ALN installation at 90% (19 ppm in Figures 1(e) and S1(d)).

Physical properties of PEG-ALN under acidic condition in vitro
Bone resorption pits have a pH of 4.0-4.8,which is optimal for collagen degradation by collagenases secreted from osteoclasts [38].In this study, we investigated the pH response of PEG-ALN to confirm its ALN release properties under acidic conditions.Figure 2(a) shows the 31 P-NMR spectra of PEG-ALN obtained at different pH conditions.The 19-ppm peak in PEG-ALN decreased after 48 h of incubation at pH 3 and 4, whereas no change was observed at pH 7.These results indicate that ALN was released under acidic conditions for over 48 h.

Antiresorptive effect of PEG-ALN
To evaluate the antiresorptive effect of PEG-ALN against OSC14 cells, TRAP staining was performed to detect osteoclasts after PEG-ALN-treatment (Figure 2(b)).TRAP-positive cells were observed only in the nontreatment (NT) group, with no cells observed in any concentration in the PEG-ALN or ALN-treatment groups.To evaluate bone resorption by osteoclasts treated with PEG-ALN-or ALN, a pit formation assay was performed (Figure 2(c)).No bone resorptive pits were observed in the PEG-ALN-or ALN-treated groups.These results indicated that PEG-ALN can potentially inhibit osteoclast formation and bone resorption.

Influence of PEG-ALN on osteoblasts
This study aimed to use PEG-ALN to decrease the osteoclast population at the fracture site in the early stage of fracture healing and to promote fracture healing by replacing the fracture site with an osteoblast-dominated environment (Figure 1(b)).Treatment with a high concentration of ALN not only suppresses bone resorption by osteoclasts but also calcifies osteoblasts via the induction of apoptosis [39].Therefore, to reveal the relationship between the concentration and efficacy of PEG-ALN on osteoblasts, KUSA-A1-differentiated osteoblast cells were treated with various concentrations of PEG-ALN ranging from 10 −9 to 10 −5 M (Figure 2(d,e)).Comparable effects on calcification were observed in both free ALN-and PEG-ALN-treated groups.Treatment with PEG-ALN and ALN at 10 −9 and 10 −8 M produced similar areas of positive Alizarin Red S staining as in NT group one (Figure 2(d)).However, in the high concentration groups of PEG-ALN and ALNtreatment (10 −7 to 10 −5 M), the positive areas were decreased in a concentration-dependent manner.Notably, no positive areas were observed when cells were stimulated with the maximum concentration (10 −5 M).The quantification of calcium concentration is shown in Figure 2(e), and the amount of deposited calcium was significantly decreased in the 10 −5 M PEG-ALN-and ALNtreated groups, although no significant difference in deposited calcium concentration was observed with PEG-ALN or ALN-treatment at 10 −9 M or 10 −8 M.These results underline the importance of identifying the appropriate amount of drug exposure and that PEG-ALN-treatment between 10 −9 and 10 −8 M did not affect osteoblasts.

Pharmacokinetics of PEG-ALN in vivo
125 I-PEG-ALN and 125 I-PEG were intravenously injected into mice in a bone fracture model to evaluate their pharmacokinetics.The amount of accumulated 125 I-PEG-ALN in the fractured and non-fractured legs 1 h after injection was estimated to be approximately 6.9% and 1.3% ID/g tissue, respectively (Figure 3(a)).An approximate five-fold higher accumulation of 125 I-PEG-ALN was observed in the fractured leg than in the non-fractured leg, and this trend continued for at least 72 h (Figure 3(a)).Histological sections of fractured and non-fractured legs from Cy5-PEG-ALN-injected mice are shown in Figure 3(b).Cy5-PEG-ALN (red color) accumulated around the fractured bone but accumulated less in the non-fractured bone.The blood retention time for 125 I-PEG-ALN was slightly shorter than that of 125 I-PEG (Figure 3(c)).The AUC values for PEG-ALN or PEG up to 72 h were 787.9 and 1278.6 mg/mL*h, respectively, while the blood half-life values of PEG-ALN and PEG were 13.6 and 25.7 h, respectively.We then studied the accumulation and retention behavior of 125 I-labeled PEG-ALN and PEG at the bone of fracture site (Figure 3(d)).The results exhibited that both PEG-ALN and PEG accumulated into the fracture site, and those accumulation amounts were basically comparable at 1 h from administration owing to the same molecular weight of PEG (40k).The following timedependent profile showed a long-time retention with slight increasing trend of PEG-ALN at the fracture site, whereas control PEG was rapidly eliminated from the fracture site (Figure 3(d)).These results indicate the utility and designated association between PEG-ALN and calcium ions of hydroxyapatite on bone tissue which induced by P-C-P moiety of ALN.In the other organs, the amount of 125 I-PEG-ALN decreased in a time-dependent manner (Figure 3(e)).

Optimized concentration of PEG-ALN for induction of osteoblast in vivo
To estimate the optimal amount of PEG-ALN for the induction of bone regeneration, the number of Runx2expressing cells at the fracture site was evaluated over a PEG-ALN concentration range (Figure 4(a)).Runx2 is an essential transcription factor for osteoblast differentiation and has been used as an osteoblast marker.The number of Runx2-expressing cells significantly increased in the 1.5 × 10 −6 M and 3.0 × 10 −5 M PEG-ALN-treated groups compared with that in the NT group and at other concentrations.In particular, the number of Runx2 positive cells decreased significantly at the highest concentration.This indicates the importance of optimal dosage for facilitating osteoblast proliferation and differentiation.Therefore, with the current design of PEG-ALN, we concluded the most appropriate dose was 3.0 × 10 −6 mmol/ mouse (1.2 × 10 −1 mg/mouse) for administration, and we used this for the in vivo investigation.Next, the time-dependent therapeutic effects were evaluated during 4 weeks from administration, and the significantly higher expression of Runx2 was confirmed during initial 2 weeks which can likely contributed to the enhancement of osteoblast proliferation and/or differentiation (Figure 4(b)).Following PEG-ALNtreatment, the number of Runx2-expressing cells significantly increased during the first two weeks.These results indicated that PEG-ALN can provide significant therapeutic effects during the early stages of fracture healing.

Toxicity of PEG-ALN in mice
Single-dose toxicity of PEG-ALN was examined in a non-fracture mouse model.Blood chemistry parameters are summarized in Table 1.The blood chemistry of the groups treated with PEG-ALN, PEG, and ALN was similar to that of the NT mice, indicating that PEG-ALN exhibits no acute toxicity.

Fracture healing effect of PEG-ALN
Typical examples of the μCT radiographs of the bone fracture sites are presented in Figure 4(c).Bone formation was observed at the fracture site in the PEG-ALN treatment group 4 weeks after the onset of treatment (Figure 4(c)), whereas less bone formation was observed in the other groups.The highest bone mineral density (BMD) value was obtained in the PEG-ALN-injected group in contrast to that of the other groups (Figure 4(d)).Histological analyses of the injured tibiae are shown in Figure 5. Tissue sections were stained with hematoxylin and eosin (H&E) and Masson's trichrome (MT).The dotted lines in Figure 5(a-h) indicate the fracture sites.Endochondral ossification was observed in all groups.However, callus formation was observed around the fractured bone in the PEG-ALN-treated group (Figure 5(c,k)).The formation of mature lamellar bone tissue (arrowhead) was observed at the fracture site in the same group and merged with the edge of the fractured bone (Figure 5(g,o)).In contrast, fibrous connective tissue (arrow) was still present in the injured tibiae four weeks after treatment in the PEG, ALN, and NT groups.Altogether, these data indicate that PEG-ALN accumulates around the fracture site to induce bone repair and complete fusion with newly formed callus and collagen tissues.

Discussion
Delayed remodeling with late fracture healing is commonly observed in orthopedic surgery when using anti-osteoporosis agents such as bisphosphonates [19,20,22,23,27].The second-generation bisphosphonate ALN is currently used as an osteoporosis drug that induces osteoclast apoptosis.However, almost all administered ALN is rapidly excreted, and its bioavailability only reaches 0.6%-1.85%[40] because of the short half-life of ALN in the blood − 0.65 h.Residual ALN diffuses to bones throughout the body and causes unexpected side effects.These issues limit the therapeutic efficacy of ALN.Consequently, the development of alternative therapeutic agents for osteoporosis-related bone fractures is urgently required in aging societies to provide safe and efficient therapeutic options and ensure that older individuals can engage in sustainable social activities.
Herein, we synthesized a macromolecule containing ALN by the simple conjugation of ALN and 40k-PEG (Figure 1 and S1) with a pH-responsive linker.Our design has the advantage of an increased molecular weight for prolonged blood retention, which enables the delivery of ALN to injured-sites through fracture-related hemorrhage.This strategy is based on the PEGylation of pharmaceuticals [41][42][43].Generally, PEGylation provides new properties, including water solubility and antifouling effects from biological inactivation (i.e.opsonization), which can reduce immunogenicity compared to their unPEGylated counterparts.The molecular weight and diameter of PEG are fundamental factors that control the performance of PEGylated pharmaceuticals.Yamaoka et al. revealed the relationship between the molecular weight of PEG and its elimination from the blood, and demonstrated the threshold for renal excretion of PEG at a molecular weight of 30k [44,45].ALN is a low-molecular-weight drug (325.1 g/mol) that easily penetrates the renal glomeruli and normal blood vessels and is consequently quickly eliminated by renal excretion and diffusion.We observed that the blood retention time of PEG-ALN was slightly lower than that of free 40k-PEG because of the high accumulation of PEG-ALN at the fracture site (Figure 3(c,d)).These results indicate that PEGylation of ALN improved blood retention and fracture accumulation, which could be due to the increased molecular weight and hydrodynamic diameter of the PEG strand.The PEG-ALN synthesized here was pH-responsive (Figure 2(a)), enabling the amine terminus of the released ALN to play an important role in its pharmacological activity [46].Adhered-ALNs on the bone surface detach in an acidic environment because of the low pH conditions caused by acids secreted from osteoclasts that form in the resorption fossa (Howship's lacuna).PEG-ALN inhibited osteoclast formation and bone resorption, which did not occur in the absence of amine residues, highlighting the importance of PEG-ALN cleavage (Figure 2(b,c)).In addition, the in vitro activity of PEG-ALN showed that free ALN and PEG-ALN had comparable osteoclastogenic effects (Figure 2(d,e)), indicating the successful release of intact ALN from PEG-ALN.
Among practical DDS, many studies are intended for cancer therapy, but their applications for bone regeneration have been limited to sustained-release methods, such as gelatin-based hydrogels, polymerceramic composite materials, and gelatin-coated bioactive glass scaffold implantation [11,12,15,17,[47][48][49][50][51][52].Therefore, an injectable DDS for simple fracture treatment has not yet been reported.The enhanced permeability and retention (EPR) effect is one of the main driving forces behind the accumulation of systemically injectable nanomedicines [53][54][55], and this effect is observed in leaky angiogenic vasculature.Similarly, leakage occurs at the fracture site due to damage to the blood vessels, bone marrow, and surrounding tissues.We hypothesized that therapeutic agents with appropriate physical characteristics can be delivered to the fracture site by utilizing this leakage in a manner similar to the EPR effect.PEG-ALN showed high accumulation at the fracture sites and was retained for a long period (Figure 3(a)).For drug retention in the tumor, incomplete drainage of accumulated drugs occurs because of immature lymphatic tissue [53][54][55].However, as fracture sites have undergone trauma, their biological conditions differ from those found in tumors, and the accumulated drugs can be quickly excreted from the fracture site via the mature lymphoid tissue.In this study, PEG-ALN was retained at the fracture site for over 48 h, whereas PEG was quickly disappeared (Figure 3(d)).Because both PEG-ALN and PEG equipped comparable molecular weight, this PEG-ALN related long-time retention was not likely due to the size of injected therapeutic agents.Therefore, it is reasonable to conclude that the association between ALN group and bone tissue, most likely the interaction of the P-C-P moiety and the calcium ions of hydroxyapatite, worked as a key of this difference of retention profile [32].Moreover, free ALN is generally known to rapidly eliminate from body after their administration.The utility of PEG-ALN was also observed by their pharmacokinetic parameters.As calculated from the biodistribution study, the half-life and AUC 0-6 value of PEG-ALN are 13.5 h and 140.0 mg/mL*h respectively.Meanwhile, the half-life of free ALN is approximately 0.5-1.0h and their AUC value is approximately 1000 ng/mL*h [56].
ALN concentration is a fundamental factor in ensuring therapeutic outcomes against osteoblasts as well as osteoclasts [57][58][59][60][61]. Low concentrations of ALN promote osteoblast proliferation and bone differentiation, whereas high concentrations inhibit bone-formation.Stimulation of osteoblasts with 10 −8 M ALN promotes the expression of osteogenic genes and cell proliferation, whereas concentrations > 10 −4 M inhibit cell proliferation [59].Studies with BMSCs showed that calcification is facilitated at 10 −9 to 10 −7 M ALN [58].These reports suggest the existence of an optimal ALN concentration (10 −8 M) for the induction of osteoblast differentiation, and possible cytotoxicity and growth inhibitory effects with higher drug  Generally, class I tibial fractures in humans exhibit up to 10.7 ml/kg of bleeding [62].Thus, a maximum of approximately 270 μL of bleeding occurred in the case of the mice used in this study (C57BL/6J; 25 g).Similarly, this indicate that approximately 270 mg of blood clots are formed at the fracture site.While, based on the pharmacokinetic result using 125 I-PEG-ALNs, approximately 6.9 ± 0.4% ID/g tissue of PEG-ALN was accumulated at the fracture site after 1 h from administration (Figure 3(a)).Based on these results, it can be calculated that the maximum accumulation of PEG-ALN at the fracture site (blood clots) was approximately 10 −8 M, when 1.5 × 10 −6 M of PEG-ALN was injected.Although a reduction in osteoclast formation is the dominant factor that accelerates therapeutic the effect in the initial stage of fracture/regeneration, it is technically difficult to collect osteoclasts from small animals.However, our in vitro analyses (Figure 2(b-e)) suggested a high sensitivity and wide concentration range of PEG-ALN for medicinal expression against osteoclasts/osteoblasts.Thus, PEG-ALN accumulated at the fracture site at an optimal concentration for the induction of osteoblast proliferation and differentiation in vivo.Complete callus formation was only observed in the PEG-ALN group (Figure 4(c)).Histological analysis revealed the presence of hypertrophic chondrocytes in all treatment groups, indicating the progression of endochondral ossification.In the PEG-ALN-treated group, the periphery of the newly formed callus (Figure 5) exhibited strong evidence of fracture healing, including the presence of fibroblast-like cells (Figure 5(k)), bone tissue with an abundance of collagen fibers (Figure 5(o)), and continuity between the fracture ends (Figure 5(c,g)).
A further direction of this study will be to provide additional evidence regarding therapeutic effects using large animals, i.e. rats, rabbits, and/or dogs, because current study using mice fracture models include some limitations such as detection of quantification analysis of osteoclast and BMD quantification.Those advanced experiments are ongoing in our project.
In summary, we achieved accelerated fracture healing by actively targeting macromolecular ALN (PEG-ALN) to the fracture site at an optimal concentration for bone regeneration.Systemic administration of PEG-ALN decreases the number of osteoclasts in the early phase of fracture healing, accelerates callus formation, and shows trend toward increase in BMD.PEG-ALN inhibits bone resorption at the inflammatory phase and induces bone formation; therefore, it has the potential to be a versatile treatment for intractable fractures.

Figure 1 .
Figure 1.Schematic illustration of fracture healing by PEG-ALN.(a) Blood vessels are damaged and bleed at the fracture site.Administered PEG-ALN leaks out with blood and accumulates at the fracture site.Accumulated PEG-ALN is anchored to the fracture bone via the P-C-P moiety of ALN.PEG-ALN is cleaved to PEG and ALN in Howship's lacuna formed by osteoclasts, which is an acidic environment, to affect osteoclasts and osteoblasts.(b) PEG-ALN accumulated at the fracture site reduces the number of osteoclasts via induction of apoptosis at the early stage of fracture healing and promotes fracture healing by replacing the fracture site with an osteoblast-dominated environment.Therefore, the number of osteoclasts is reduced during the inflammatory and repair phases, and callus formation is induced by osteoblasts in the repair phase.Synthesis and NMR spectrum of PEG-ALN.(c) Synthetic routes of PEG-ALN.(d) 1 H NMR (400 MHz, CDCl 3 ) spectrum of PEG-CDM.(e) 31 P-NMR (400 MHz, D 2 O) spectrum of ALN, PEG-CDM, and PEG-ALN.

Figure 2 .
Figure 2. Physicochemical and pharmacological properties of macromolecular ALN.(a) Physical properties of PEG-ALN under acidic conditions in vitro. 31P-NMR (400 MHz, D 2 O) spectrum of PEG-ALN in different pH conditions.(b) Antiresorptive effect of PEG-ALN against osteoclasts.TRAP-stained images of osteoclasts after treatment with different concentrations of ALN (upper panels) or PEG-ALN (lower panels).(c) Von Kossa-staining images of osteoclasts after treatment with different concentrations of ALN (upper panels) or PEG-ALN (lower panels); the scale bar indicates 900 μm.(d) Concentration effect of PEG-ALN against osteoblasts.Alizarin Red S-stained images of osteoblasts after treatment with different doses of ALN (upper panels) or PEG-ALN (lower panels); the scale bar indicates 900 μm.(e) Quantitative result of calcium deposition after treatment with different concentrations of ALN or PEG-ALN.Each value represents the mean ± S.D. of three separate experiments.**p < 0.01: significantly different from the value of 10 −5 M of PEG-ALN-or ALN-treated group.

Figure 3 .
Figure 3. Pharmacokinetics of PEG-ALN or PEG in mice. 125I-labeled PEG-ALN or 125 I-labeled PEG were intravenously injected following fracture, and their tissue distribution and blood concentrations were investigated.(a) Accumulation at fracture/nonfracture area.(b) Histological evaluation of Cy5-labeled PEG-ALN accumulation.Upper photograph shows fracture limb, and lower shows non-fracture limb.Left side shows H&E stains, and right side shows fluorescent images; red indicates Cy5-labeled PEG-ALN, and blue indicates nucleus.The scale bar indicates 900 μm (H&E stains) and 300 μm (fluorescent images).(c) Blood concentration of PEG-ALN (•) or PEG (△).(d) Accumulation ratio of PEG-ALN (•) or PEG (△) at the site of fracture.Tissue distribution of PEG-ALN at the major organs such as (e) heart, (f) lung, (g) liver, (h) kidney, and (i) spleen.Each value represents the mean ± S.D. of five separate experiments.**p < 0.01: significantly different from the value of PEG.

Figure 4 .
Figure 4. Concentration and the duration time of PEG-ALN for induction of fracture healing, and radio photographic analysis of fracture healing effect.(a) Ratio of Runx2-expressing cells at the fracture site 1 week after administration of different concentrations of PEG-ALN.(b) Ratio of Runx2-expressing cells at different timepoints after PEG-ALN administration.*p < 0.05, **p < 0.01: significant difference from the NT group value.(c) μCT radio photography images of fractured bone 4 weeks after treatment.Treatment groups are NT, PEG, PEG-ALN, and ALN from the left.(d) BMD values at the fracture area 4 weeks after treatment.Treatment groups are NT, PEG, PEG-ALN and ALN from the left.Each value represents the mean ± S.D. of five separate experiments.

Figure 5 .
Figure 5. Histological sections of fracture site 4 weeks after treatment.The treatment groups are NT (a, e, i and m), PEG (b, f, j and n), PEG-ALN (c, g, k and o) and ALN (d, h, l and P); (a -d and i -l) H&e and (e -h and m -p) MT stains.High magnification image of the area enclosed by a square (i -p); the scale bar indicates 900 μm (a -h) and 60 μm (i -p).Both black and white arrows represent fibrous connective tissues.

Table 1 .
Blood chemistry parameters after a single dose of PEG-ALN.