. Author manuscript; available in PMC: 2023 Oct 1.
Abstract
Targeted drug delivery, often referred to as “smart” drug delivery, is a process whereby a therapeutic drug is delivered to specific parts of the body in a manner that increases its concentration at the desired sites relative to others. This approach is poised to revolutionize medicine as exemplified by the recent FDA approval of Cytalux (FDA approves pioneering drug for ovarian cancer surgery - Purdue University News) which is a folate-receptor targeted intraoperative near infrared (NIR) imaging agent that was developed in our laboratories. Fracture-associated morbidities and mortality affect a significant portion of world population. United states, Canada and Europe alone spent $48 billion in treating osteoporosis related fractures although this number doesn’t count the economic burden due to loss in productivity. It is estimated that by 2050 ca 21 million hip fractures would occur globally which will be leading cause of premature death and disability. Despite the need for improvement in the treatment for fracture repair, methods for treating fractures have changed little in recent decades. Systemic delivery of fracture-homing bone anabolics holds great promise as a therapeutic strategy in this regard. Here we report the design of a fracture-targeted peptide comprised of a payload that binds and activates the parathyroid hormone receptor (PTHR1) and is linked to a targeting ligand comprised of 20 D-glutamic acids (D-Glu20) that directs accumulation of the payload specifically at fracture sites. This targeted delivery results in reduction of fracture healing times to <1/2 while creating repaired bones that are >2-fold stronger than saline-treated controls in mice. Moreover, this hydroxyapatite-targeted peptide can be administered without detectable toxicity to healthy tissues or modification of healthy bones in dogs. Additionally, since similar results are obtained upon treatment of osteoporotic and diabetic fractures in mice, and pain resolution is simultaneously accelerated by this approach, we conclude that this fracture-targeted anabolic peptide displays significant potential to revolutionize the treatment of bone fractures.
Keywords: Fracture healing, Anabolic peptide, Targeted delivery, PTHR1 receptor
1. Introduction
Over 18 million bone fractures are reported yearly in the USA alone, imposing enormous physical, emotional and financial burdens on the healthcare system. (1) While some fractures heal without major intervention, most require surgery, often including insertion of screws, rods or plates, or implantation of grafts. (2) In cases where fractures fail to heal completely (nonunion fractures), a decrement in quality of life often ensue. Whenever the recovery process is prolonged, the associated loss of productivity can have significant financial consequences. (3) Moreover, where extended immobilization is required, such as during recovery from a hip fracture, serious co-morbidities can emerge, including: unintentional injuries, diseases of the heart, vascular embolisms, chronic lower respiratory infections, and even death. (3) Taken together, bone fractures are believed to cost >$17 billion annually in the US, (4) not including expenses from loss of productivity.
Despite the need for improvement in the speed and quality of fracture repair, methods for treating fractures have changed little in recent decades. (5) Immobilization as described above has provided the physical stabilization that enables healing to occur, but has done little to augment the biological processes required for replacement of damaged bone (2). Topical application of bone morphogenetic protein 2 (BMP2) to exposed fracture surfaces has been used to enhance the biological processes. However, closure of the wound after application of BMP2 prevents its continuous administration during a healing process. (6)
We hypothesized that the above unmet medical need could be addressed using systemically targeted bone anabolic agents that are designed to allow continuous administration over the duration of the repair process. In exploring this question, we considered multiple targeting ligands that might deliver an attached bone anabolic agent specifically to fracture surfaces following parenteral administration. (5,7) and selected a linear oligopeptide of 20 D-glutamic acids (D-Glu20) that homes to exposed hydoxyapatite which is the primary inorganic constituent of teeth enamel and bone. In this paper we demonstrate that an anabolic peptide designed to bind the PTHR1 receptor (payload) linked the D-Glu20 oligopeptide (targeting ligand) dramatically accelerates bone fracture repair, yielding healed bone that is significantly stronger than undamaged bone in a mice model in less than half of the time required for an untreated stabilized fracture to heal. Moreover, this targeted peptide (referred to as Ab46-D-Glu20 henceforth) can be administered without detectable toxicity to healthy tissues or modification of healthy bones in dogs. Additionally similar results are obtained upon treatment of osteoporotic and diabetic fractures in mice, and pain resolution is simultaneously accelerated by this approach. We conclude that this fracture-targeted anabolic peptide Ab46-D-Glu20 displays significant potential to revolutionize the treatment of bone fractures.
2. Results
2.1. Design, synthesis, and analysis of oligoglutamic acid-targeted drugs
As described in the Introduction, for continuous stimulation of bone fracture healing throughout recovery process, a targeted bone anabolic agent that would concentrate on a fracture surface following systemic administration was required. To confirm that a linear 20-mer of D-Glu (D-Glu20) could deliver a therapeutic cargo to a fracture surface following parenteral injection, we conjugated D-Glu20 to a 99mTc chelating peptide and examined its biodistribution following tail vein injection in Swiss-Webster mice with fractured femurs. (8) As shown in Fig. 1A, the majority of radioactivity was localized to the fracture site, with minor amounts accumulating on growth plates that are known to remain osteogenically active in adult mice. (9)
Fig. 1.
Analysis of the biodistribution and pharmacokinetics of D-Glu20 targeted peptides. A. 99mTc-labeled peptides were injected via tail vein into mice with fractured femurs and its biodistribution was imaged 18 h later. B&C. 125I-labeled abaloparatide and 125I-Ab46-D-Glu20 were injected into mice with fractured femurs via tail vein and their clearance from the bloodstream (panel B) and fracture site (panel C) were quantitated. (Each time point represents data on 5 sacrificed mice; error bars = SEM).
Next, for selection of the bone-anabolic agent, we explored use of a modified form of residues 1–46 of parathyroid hormone related peptide (PTHrP) containing the substitutions F22E, F23L, H25E, H26K, I22L, A29Aib, E30K, and I31L. (10-12). Conjugation of the nonsignaling COOH-terminus of this modified PTHrP to a linear peptide of 20 D-glutamic acid residues (D-Glu20) allowed for retention of PTHrP signaling activity following its adsorption to exposed hydroxyapatite.
To synthesize Ab46-D-Glu20 conjugate, the PTHrP-like peptide (Ab46) was linked to D-Glu20 using standard solid phase peptide chemistry, yielding a final fusion product (Ab46-D-Glu20; Fig. S1) in 19% overall yield and final purity of 94%, as determined by HPLC and mass spectrometry (Fig. S1). The complete sequence of the peptide is HAVSEHQLLHDKGKSIQDLRRRELLEKLL-AibKLHTAEIRATSEVSPNS-OH. Ab46-D-Glu20 maintained a 27 nM association constant to hydroxyapatite (Fig. S2). Analysis of its signaling activity using a cAMP assay further confirmed that the potencies of both untargeted and targeted peptides (EC50 ~ 0.48 nM and ~ 0.66 nM, respectively) were congruent with values reported for abaloparatide (Abalo) in the literature (13) (Fig. S3).
To obtain an initial indication of an optimal treatment regimen, we first examined the pharmacokinetics of Ab46-D-Glu20 following its subcutaneous injection into fracture-bearing mice. As seen in Fig. 1B, no significant differences were observed between non-targeted 125I-labeled peptide and targeted 125I-Ab46-D-Glu20 in either circulation half-life (4.2 h versus 3.7 h) or cumulative systemic exposure (AUC; 24.4 h and 20 h, respectively), indicating that any resulting systemic toxicity should be similar between the targeted and nontargeted forms. In contrast, a major difference was observed in the residence times of the two peptides at the fracture site, where the half-life of Ab46-D-Glu20 was found to be 67 h, while the half-life of nontargeted peptide was only 8.8 h (Fig. 1C). This difference in half-life on the fracture surface resulted in a 12-fold larger cumulative residence time (AUC) of Ab46-D-Glu20 compared to the untargeted peptide (96 h compared to 8 h, respectively). We also observed that the residence time of Ab46-D-Glu20 in the contralateral femur was 8-fold lower than its residence time in the fractured femur (Fig. 1C), suggesting that concerns regarding stimulation of growth of healthy bones should be minimal. Based on these data, we conclude that although systemic exposure of D-Glu20-targeted and nontargeted peptides are similar, the significantly enhanced accumulation of Ab46-D-Glu20 at the fracture surface should improve repair of broken bones.
2.2. Analyses of fracture repairs in healthy mice
To determine whether the increased retention of Ab46-D-Glu20 at the fracture site translates into improved fracture healing, we conducted a simple comparison of fractured femur healing times using a dose calculated by allometric scaling of the human dose of abaloparatide prescribed for osteoporosis treatment. (11,14) As seen in Fig. 2A-C, scans of the repairing fractures revealed that bone deposition imaged 3 weeks after initiation of treatment with the nontargeted peptide was concentrated on the periphery of the fracture, with minimal density bridging apposing calluses. In contrast, following the same 3 weeks of therapy, bone density in the Ab46-D-Glu20 treated mice distributed more evenly across the fractured area, with the overall size of the callus exceeding that of the non-targeted peptide treated mice. Indeed, bone morphometric analyses confirmed a 1.5-fold increase in the ratio of mineralized bone volume to total bone volume (BV/TV) in the Ab46-D-Glu20 treated mice. More detailed morphometric analyses further revealed that this increase in bone density derived primarily from a reduction in trabecular spacing rather than an increase in trabecular thickness (Table S1).
Fig. 2.
Analysis of the impact of Ab46-D-Glu20 on the healing of fractured femurs containing internal stabilizing rods. A–C. Comparison of microCT scans of representative fractures from mice treated with vehicle (A), abaloparatide (B), and Ab46-D-Glu20 (C) obtained 3 weeks after induction of femur fracture. D and E. Morphometric analysis of the calcified fracture volume (BV; panel D) and BV/TV (panel E). F and G, Mechanical strength analysis of the maximum load (panel F) and work to fracture the repaired femurs (panel G). (n = 10 mice per group; Error bars = SEM). Statistical analyses were made using one way ANOVA (*P < 0.05).
To establish whether the observed increase in bone deposition translates into an improvement in mechanical properties, we next measured the force required to bend the healed femurs until they fractured again. As seen in Fig. 2F, the Ab46-D-Glu20 treated fractures were able to sustain a 2.5-fold greater maximum load than the untargeted peptide-treated femurs. Interestingly, this force to refracture in the Ab46-D-Glu20 treated mice was even found to exceed (by 30%) that required to fracture the contralateral healthy femurs in the same mice (or similar unmodified femurs in saline-treated mice). Furthermore, the work to fracture measured in related mechanical studies was found to average 3.5-fold higher in the Ab46-D-Glu20-treated than untargeted-treated femurs (Fig. 2G). Taken together, these data demonstrate that Ab46-D-Glu20 greatly facilitates the healing process, returning fractured femurs to full mechanical strength much more rapidly than femurs in saline- or Ab46-treated mice. Mice in the study were monitored for their body weight. No body weight abnormalities were observed at any of this study’s doses.
2.3. Analyses of fracture repairs in osteoporotic and diabetic mice
Encouraged by the above observations, we next examined the ability of Ab46-D-Glu20 to treat broken bones in animals with comorbidities known to compromise fracture repair in humans. Patients with osteoporosis are known to have twice the probability of both experiencing a bone fracture and suffering a nonunion fracture during the repair process. (15) For this purpose, we deprived female Swiss-Webster mice of estrogen via ovariectomy to induce osteoporosis (Fig. S4) and then monitored their response to the treatment of Ab46-D-Glu20 during a subsequent repair process. As seen Fig. 3, stabilized femurs in mice treated for four weeks with saline showed minimal bone deposition across the fracture site (panel A), with the residual fragments of pre-existing bone comprising most of the density seen in the CT scans. Analogous scans of mice treated with estrogen beginning at the time of fracture displayed similar deficits in bone deposition (panel B). In contrast, mice treated with Ab46-D-Glu20 exhibited markedly elevated ossification (panel C); this increase in bone formation was confirmed by morphometric analyses of bone volume (panel D) and bone volume/total volume (panel E) (P < 0.05) (Table S2). That this enhanced bone deposition translates into improved mechanical properties was then established by comparing the maximum loads each cohort of repaired femurs could withstand before refracture. As shown in panel F, Ab46-D-Glu20 treated fractures were able to resist rebreaking when exposed to forces ~70% greater than saline controls (P < 0.05). Even daily estrogen supplementation over the course of the healing process was unable to improve bone strength by more than 35% (P < 0.05). Taken together, these data suggest that Ab46-D-Glu20 can enhance the healing of fractured femurs in hypoestrogenic mice, even without estrogen replacement therapy.
Fig. 3.
Evaluation of the impact of Ab46-D-Glu20 on the healing of fractured femurs induced in osteoporotic mice via ovariectomy. A-C. Comparison of microCT scans of representative fractures from mice treated with saline (vehicle, panel A), estrogen (panel B), or Ab46-D-Glu20 (panel C) obtained 4 weeks after induction of femur fracture. D. and E. Morphometric analysis of the calcified fracture volume (BV, panel D) and BV/TV (panel E). F. Comparison of the maximum load before refracture among the different treatment cohorts. (n = 10 mice per group; Error bars = SEM). Statistical analyses were made using one way ANOVA (*P < 0.05; ns, not significant).
Next, we examined the ability of Ab46-D-Glu20 to augment fracture repair in murine models of diabetes mellitus. Patients with diabetes have a 6-fold greater probability of experiencing a bone fracture. (16) Following fracture, they have a 2-fold higher probability of suffering a non-union outcome (17), likely deriving from diabetes-associated lower bone densities (18,19), poorer bone vascularization, and/or peripheral neuropathy (20).
To test the ability of Ab46-D-Glu20 to treat fractures in diabetic mice, diabetes was induced by repeated injections of streptozotocin (STZ), and blood glucose levels were maintained above 250 mg/dL for two months prior to induction of femoral fracture (Fig. S5). Examination of microCT images of the repairing fractures obtained four weeks post-initiation of therapy demonstrated that insulin supplementation significantly improved bone deposition (compare Figs. 4 A and B). However, treatment with Ab46-D-Glu20 induced still greater bone formation, even in the absence of added insulin (panel C). That this enhanced osteogenesis was manifested across all metrics of bone regeneration is shown in panels D and E, where both bone volume and bone density (bone volume/total volume) were observed to increase when comparing saline-treated, to insulin-supplemented, to Ab46-D-Glu20 injected mice (Table S3). More convincingly, because the fundamental lesions in diabetic bones arise primarily from defects in collagen, (16) a more relevant test of bone fracture repair should derive from mechanical measurements of bone properties. Thus, as shown in Fig. 4 panels F and G, Ab46-D-Glu20 again outperformed saline and insulin supplementation by approximately two-fold in both maximum sustainable load and work to refracture, suggesting that the newly deposited bone must contain sufficient healthy collagen to facilitate elastic deformation without rupture. These data thus suggest that Ab46-D-Glu20 can improve fracture repair even under conditions where bone healing is seriously compromised by comorbidities like osteoporosis and diabetes.
Fig. 4.
Evaluation of the impact of Ab46-D-Glu20 on the healing of fractured femurs in mice induced to develop diabetes by injection of streptozotocin and maintenance of blood glucose above 250 mg/dL for two months prior to induction of femur fracture. A–C. Comparison of microCT scans of representative fractures from mice treated with saline (vehicle, panel A), insulin (panel B), and Ab46-D-Glu20 (panel C) obtained 4 weeks after the induction of femur fracture. D. and E. Morphometric analysis of the calcified fracture volume (BV, panel D) and BV/TV (panel E). F. and G. Comparison of the maximum load and work to fracture among the treatment cohorts. (n = 10 mice per group; Error bars = SEM). Statistical analyses were made using one way ANOVA (*P ≤ 0.05, **P ≤ 0.01; ***P ≤ 0.001; ns, not significant).
2.4. Assessment of the functionality of healing bones
Finally, recognizing that the amount of pain experienced during physical activity can impact a patient’s willingness to engage in weight-bearing motion and thereby the healing process, we examined the impact of Ab46-D-Glu20 on commonly employed pain metrics in mice with fractured femurs. For this purpose, the movement patterns of mice in cages were video-recorded and quantitated using a video analysis program. (21) Although none of the treatment cohorts showed enhanced movement immediately after fracture, mice treated with Ab46-D-Glu20 began to travel farther beginning on day 12, eventually moving approximately twice as far per day than mice treated with either saline or ibuprofen (Fig. 5A). Consistent with this difference, the Ab46-D-Glu20 cohort was found to move almost twice as fast as either of the other two cohorts, and this difference also persisted through the end of the study (panel B). (22,23) That the saline and ibuprofen-treated cohorts also desired to move can be further deduced from the data in panel C, where the time spent moving among the three cohorts did not differ, but the distance traveled and speed of motion increased in the Ab46-D-Glu20-treated cohort. (23,24)
Fig. 5.
Evaluation of discomfort experienced by mice with broken femurs treated with saline (●), Ab46-D-Glu20 (■), or ibuprofen (▴). Total distance traveled (panel A), velocity (panel B), and time spent moving (panel C) were videorecorded and analyzed by the program Ethovision XT. (n = 10 mice per group).
2.5. Analysis of systemic toxicities
The most obvious advantage of precision targeting arises from selective accumulation of the targeted drug at the diseased site, with little or no uptake by healthy tissues. (5,25) Although the focused accumulation Ab46-D-Glu20 at the fracture site in Fig. 1A would predict reduced off-site toxicity, mouse body weights were still monitored throughout all of the studies and no differences among treatment groups were observed (Figs. S6, S7). To confirm this apparent absence of toxicity in larger animal models, canines with ulnar fractures were treated daily for 17 weeks with Ab46-D-Glu20 doses of 0, 0.5, 1.6, and 5 nmol/kg/day (based on allometric scaling, the 0.5 nmol/kg/day dose was calculated to be comparable to the dose used in all of the murine studies above). As seen in Fig. 6, no significant differences among treatment groups were observed in body weight (panel A), heart rate (panel B), serum calcium (panel C) or serum alkaline phosphatase (panel D); i.e. common metrics changed by toxic levels of PTHrPs. (26) All other blood chemistries and complete blood counts (CBC) were also within normal ranges, and organ weights as well as gross organ pathologies did not differ among cohorts., i. e. no significant histological differences were observed in the lungs, liver, kidneys, heart, spleen, bladder, injection site, or muscles of the three cohorts (panels E-J). Thus, no evidence of toxicity was observed in any of the dogs, even those treated daily for 17 weeks with the highest dose.
Fig. 6.
Analysis of the toxicity of Ab46-D-Glu20 in dogs. Beagles were dosed daily for 17 weeks with 0, 0.5, 1.6, or 5 nmol/kg/day of Ab46-D-Glu20 and then sacrificed for toxicology analyses. Based on allometric scaling, the 0.5 nmol/kg/day dose was calculated to be comparable to both the usual human dose (20 nmol/person/day) as well as the dose used in the murine studies above (3.8 nmol/kg/day). Body weights (A), heart rates (B), serum calcium (C) and serum alkaline phosphatase (D) were monitored regularly. Representative histology images of the kidneys (E & F), liver (G & H) and injection site (I & J) from dogs treated with either saline (left panels) or 5 nmol/kg/d Ab46-D-Glu20 (right panels) reveal no obvious toxicities (Fig. S8). (n = 3 dogs in the saline cohort and 2 dogs each in the other three cohorts).
3. Discussion
We have described a targeted bone anabolic agent that localizes specifically to exposed hydroxyapatite on bone fracture surfaces and thereby improves the fracture repair process both quantitatively and qualitatively. Because of its ability to specifically accumulate on fracture surfaces, Ab46-D-Glu20 can be systemically administered over the duration of a healing process and thereby continuously stimulate fracture repair. Thus, unlike the only FDA-approved fracture-healing drug (Infuse®) which is solely approved for topical application during surgery, (27) Ab46-D-Glu20 can be repeatedly administered systemically during fracture repair while still only activating bone regeneration at the fracture site. The net result observed was a > 50% reduction in fracture repair time with a ~ 200% increase in mechanical strength compared to saline treated controls. The further ability of Ab46-D-Glu20 to treat osteoporotic and diabetic fractures with equal efficacy argues that Ab46-D-Glu20 might even facilitate the healing process in patients most at risk of suffering serious complications. Moreover, because the prolonged inactivity associated with fracture immobilization can trigger serious comorbidities (as often observed during recovery from a hip fracture), (28) the ability of Ab46-D-Glu20 to accelerate a decrease in pain and associated return to normal activity may even have the potential to save lives.
Analysis of the pharmacokinetics of Ab46-D-Glu20 revealed not only its significantly (twelve-fold greater) enhanced accumulation over untargeted peptide at the fracture site, but also its remarkably prolonged retention time (i. e. t1/2–67 h) on the fracture surface. Importantly, although these analyses were performed at an Ab46-D-Glu20 concentration similar to the therapeutic concentrations employed above, more extensive dosing studies suggest that D-Glu20 does not saturate adsorption sites on fracture surfaces until at least 10,000 times this concentration. (5) This residual targeting capacity suggests that considerable flexibility may still exist to optimize either dosing concentration or administration frequency. It will therefore be important to determine whether a different dosing regimen might allow administration of significantly fewer doses without compromising accelerated fracture repair (Fig. S9).
It has been shown previously that PTH or PTHrP ligand analogs can distinguish between two high-affinity PTHR1 conformations, R0 and RG, and that efficient binding to R0 results in prolonged signaling responses in cells and prolonged calcemic responses in animals, whereas selective binding to RG results in more transient responses. Intermittent PTH ligand action is known to favor the bone-formation (anabolic) response, whereas continuous ligand action is known to favors the net bone-resorption/calcemic (catabolic) response. Unlike PTH, PTHRrP analogs such as Abalo has been shown to bind more selectively to the RG conformation of PTHR1 and thus induces more transient signaling responses. (13) This Abalo/PTHrP like response is expected from Ab46-D-Glu20 based on its sequence similarity to PTHrP and Abalo, which explains the bone anabolic activity exhibited by Ab46-D-Glu20 in our experiments.
Considering that D-Glu20 can target an attached cargo specifically to any exposed hydroxyapatite (5), one can hypothesize that D-Glu20 should be capable of delivering a therapeutic or imaging agent to any traumatized tissue where hydroxyapatite is exposed. Although we only examined mice with broken femurs, hydroxyapatite is also prominently exposed in other long-bone fractures, stress fractures, vertebral and rib fractures, etc., as well as during maxillofacial surgeries, spinal fusions, bone grafts, distraction osteogenesis, and prosthetic and dental implantations. (29-31) If Ab46-D-Glu20 can facilitate ossification as efficiently in these other situations as it does in mice with fractured femurs, its contributions to reduction of pain, morbidity and loss of productivity associated with bone traumas could be multifarious.
As noted above, both the probability of suffering a broken bone and the likelihood of experiencing a delayed or incomplete recovery are heightened in patients with osteoporosis and diabetes. Other risk factors that may also predispose a person to an increased probability of fracture and/or prolonged recovery process are age, smoking, genetic abnormalities (osteogenesis imperfecta, sickle cell disease, etc.), and obesity. (32-34) Although the data in Figs. 3 and 4 demonstrate that Ab46-D-Glu20 can enhance fracture healing in animal models of osteoporosis and diabetes, it will be important to evaluate whether the delayed healing associated with these other risk factors can be similarly mitigated by treatment with Ab46-D-Glu20.
To thoroughly understand the impact of Ab46-D-Glu20 on fracture repair in mice, we monitored multiple parameters including bone volume, bone volume/total volume, trabecular thickness, trabecular spacing, maximal sustainable load, work to fracture, and improvement in function and various measurements of pain in most animal studies. After many discussions with orthopedic surgeons, however, we recognized that the best parameters for prediction of the speed of recovery in human patients might be sensation of pain and maximum load. Thus, the sensation of pain might impact the delay after which a patient will resume weight-bearing exercise (i.e., an activity that can enhance fracture repair), and maximum load should determine whether this return to activity will be physically safe. The fact that Ab46-D-Glu20 promotes significant improvements in both parameters argues that it can indeed accelerate the return of a fracture patient to normal activity.
4. Materials and methods
Peptides and estrogen benzoate were dissolved in in sterile PBS and doses were stored in aliquots at −20 °C until injection. Insulin was formulated in sterile PBS. Ibuprofen was formulated at 0.6 g/L in the drinking water. All radioactive usage was performed in accordance with protocols approved by Purdue University’s Radiological and Environmental Management (REM) and the Nuclear Regulatory Commission (NRC). Purdue Authorization # 13PLOW01. All radioactive waste was disposed of in accordance with guidelines set by the NRC.
4.1. Tc Radiolabeling of peptides
For 99mTc labeling, 0.6 mg of EDTA disodium dihydrate dissolved in nitrogen-sparged water (10 mg/mL) was added to 50 mg of sodium gluconate solution (100 mg/mL) in nitrogen-sparged water. To that mixture, a solution of 0.2 mg of tin chloride dihydrate (10 mg/mL) dissolved in nitrogen-sparged 0.2 N HCl was added. Then 4 μmol of 99mTc chelate-containing peptides were added to the solution and the pH was adjusted to 6.8 using NaOH. The solution was flash frozen in liquid nitrogen and lyophilized overnight. The compound was then mixed with 15 mCi of m99Tc (Cardinal Health) and after 15 min of shaking, quantitative chelation was confirmed by analytical radio reversed-phase high-performance liquid chromatography (1260 HPLC; Agilent Flow-RAM radiodectector, Lablogic).
4.2. 125I Labeling of peptides
The peptides had cysteine attached to their N terminus. 125I labeled (S)-2-amino-N-(2-(2,5-dioxo 2,5-dihydro-1H-pyrrol-1-yl) ethyl)-3-(4-hydroxy-3-iodophenyl) propenamide was synthesized and purified via radio-HPLC to a purity of >98%. 250 nmols of the radiotracer were combined with 1 mmol of the peptides at a concentration of 0.5 mM in PBS for 24 h then purified via radio-HPLC to >98%. The resulting radiolabeled test articles were frozen and lyophilized for 24 h.
4.3. Animal experiments: general
All animal experiments were performed in accordance with Purdue University’s PACUC (Purdue Animal Care and Use Committee) approved protocols. All experiments involving rodents were performed under the protocol PACUC 1410001138 and experiments involving dogs were performed under the protocol PACUC 182001694.
4.4. Mice
Female ND-4 Swiss Webster mice were purchased from Envigo, and were housed as social groups of five, and were put on a 12 h light cycle, and were housed in standard specific-pathogen-free conditions. Male CD IGS-1 were also purchased from Charles River Laboratories and were housed as groups of five. These mice were maintained on 2018S diet and RO water, and were on a 12 h light cycle, and were housed in standard specific-pathogen-free conditions. Animals were housed in the Purdue Life Science Animal Housing Facility.
At 3 weeks following fracture induction, mice were euthanized using a CO2 overdose. Both fractured and healthy femurs were excised and fixed in formalin solution along with the kidneys and liver. Mice in which induction of fracture was judged abnormal were not enrolled in the study and mice in which the fracture stabilization pin was observed to move were excluded from subsequent analyses.
Unless specified otherwise, 10 mice per group were used for all the experiments. For pharmacokinetics analysis 125I-Ab46-D-Glu20 and 125I-labeled abaloparatide, five mice were used for each time point. Mice were dosed at 40 nmol/ kg using the radiolabeled peptides.
4.5. Mouse bone fracture
A stabilized femoral fracture was induced under aseptic conditions following confirmed isoflurane anesthesia. Skin around the knee was initially shaved and cleaned with an alcohol pad and then sterilized with betadine solution. A skin incision was made over the knee and the patella was displaced to enable insertion of a 22-gauge locking nail into the intermedulary canal. The incision site was then closed with sutures and the femur was fractured using an Einhorn fracture device (RISystems). Buprenorphine (0.05–0.1 mg/kg) was administered subcutaneously at the time of fracture and dosed every 12 h for 3–7 days post operation to enable unimpeded movement of the mice following emergence from anesthesia.
In the case healthy mice bone fracture model, the mice groups were each dosed subcutaneously with either saline (Gibco), Abalo (3.8 nmol/kg) or Ab46-D-Glu20 (3.8 nmol/kg). Th Mice received approximately 0.1 mL fluid per dose.
4.6. Mechanical force testing
The mechanical properties of the repaired femurs were analyzed using an Instron 4-point-bend-to-failure device (Norwood, MA). Maximum load, stiffness, displacement post-yield, and work-to-fracture data were all quantitated at the end of each study.
4.7. μCT analysis
A PerkinElmer Quantum GX μCT imager was used to collect CT images and bone data. The bones were scanned while immersed in phosphate-buffered saline to prevent dehydration. ImageJ and BoneJ software were used to analyze the images for bone density, total volume (TV), relative bone volume (BV/TV), trabecular thickness (Tb.Th), and trabecular spacing (Tb.Sp). Locations of interest included the fracture callus at its widest part and both cortical and cancellous bone between the points on the cortical bone at the fracture site.
4.8. Osteoporosis fracture model
Eight-week-old female Swiss-Webster mice underwent bilateral ovariectomies to surgically induce estrogen-deficient osteoporosis. Mice were anesthetized prior to surgery using 2–3% isoflurane and 0.03 mg/kg buprenorphine was administered subcutaneously for postoperative pain relief. A 3 × 2.5-cm area was shaved dorsally from the iliac crest. The area was washed with betadine followed by 70% ethanol and then draped. A dorsal 2 cm midline incision was made above the ovaries and the skin was dissected from the underlying fascia. A 1 cm lateral incision of the midline was then made through the fascia, opening the abdominal cavity. Adipose tissue surrounding the ovary in the abdominal cavity was then gently dissected, and the ovary was isolated and the uterine horns and blood vessels within a 0.5 cm radius were cauterized. The ovary was then removed, and the process was repeated on the contralateral side. The peritoneal cavity was then closed, and the skin was returned and stabilized with sutures. Following recovery from general anesthesia in a clean recovery cage, the mouse was dosed every 12 h with buprenorphine for 3–5 days. Although osteoporosis was generally observed to develop within 4–6 weeks, significant osteoporosis was ensured by waiting 8 weeks post ovariectomy and then confirming osteoporosis by analysis of bone mineral density prior to inducing bone fracture. Insulin was dosed at (50 IU/kg). Mice were dosed at 38 nmol/kg of Abalo or Ab46-D-Glu20 twice/week.
4.9. Diabetic fracture model
Type I diabetes was induced in twelve-week-old male IGS-1 mice using intraperitoneal injections (IP) of streptozotocin (STZ) (40 mg/kg) in a 10 mM citrate buffer daily for 5 days. Mice were 16 weeks old at first injection and were in the 31-35 g weight range. Venous blood was obtained from the tail, and glucose levels were checked using a diabetic blood monitor. Diabetes was confirmed by measuring 3 consecutive days of elevated blood sugar. After 8 weeks of continuous hyperglycemia averaging ~500 mg/dL, the mice underwent the above fracture procedure. Estrogen benzoate, a long-lasting estrogen prodrug was dosed at (30 μg/kg). In case of the peptides, the mice were dosed at 38 nmol/kg of Abalo or Ab46-D-Glu20 twice/week.
4.10. Voluntary movement measurements
For analysis of motility, mice were first habituated to the behavior room for 30 min and then transferred to separate locomotor boxes and their movements were videorecorded and analyzed for 10 min using an EthoVision 3-point directionality tracking instrument. Prior to inducing a bone fracture, voluntary motility data were measured for 2 weeks to obtain baseline motility data. Then, following midshaft femur fracture, were assigned to one of three treatment groups: 1) Ab46-D-Glu20 (3.8 nmol/kg) subcutaneously twice a week; 2) phosphate-buffered saline subcutaneously twice a week; or 3) ibuprofen (0.6 g/L) administered continuously in their drinking water. All groups of mice were treated for 5 weeks post-fracture and their voluntary movements measured once a week.
4.11. Dogs
Beagle dogs were purchased from Marshall Bio Resources. Experiments were performed in accordance with Purdue University’s PACUC (Purdue Animal Care and Use Committee) approved protocols. Dogs were housed individually and were provided with dog bed and were maintained on Envigo 2025 diet and tap water and were on a 12 h light cycle. They were socialized daily, and the temperature and relative humidity of the rooms were controlled (66–67 °F and 30–70% respectively). For analysis of the toxicity of Ab46-D-Glu20 at different doses in dogs, two dogs were used in each cohort with three dogs per saline cohort.
4.12. Ulnar fracture surgeries in dogs
Both forelimbs of the dogs were prepared by the hanging limb technique under general anesthesia and both forelimbs were draped. The distal ulna was approached via a 3 cm caudolateral skin incision. Muscles and periosteum were elevated and retracted with two Hohmann retractors around the ulna. The carpal joint was marked with a 25G hypodermic needle, and a 10 mm transverse osteotomy was made at the distal ulnar diaphysis 5 cm proximal to the carpal joint. After copious flush with sterile saline, the muscle fascia was closed in cruciate pattern with 3–0 PDS. The subcutaneous layer was closed in a simple continuous pattern with 3–0 moncryl. The skin was apposed in a cruciate pattern with 3–0 nylon. For the contra lateral side, a 10 mm transverse osteotomy was created in the same manner. Postoperative radiographs were taken to confirm proper osteotomies.
The incisions were covered with sterile non-adherent dressing, and Modified Robert-Jones bandage were placed and maintained for 5 days after the surgery. The bandages were replaced if it is soiled, wet, slipped, or too tight. Carprofen (2.2 mg/kg) were orally administrated with food twice a day for 5 days. Daily routine physical examination will be performed twice daily for the first 5 days postoperatively followed by once daily through 14 post-operative days and once weekly thereafter. Physical examination included vital signs, pain/lameness, and incision check. The skin sutures were removed in 14 days after the surgery. Qualified veterinary staff provided postoperative care until skin suture removal in 2 weeks after the surgery.
4.13. Safety analysis of multi-week drug treatment in dogs
One-year old canines (beagle dogs) with ulnar fractures were treated daily for 17 straight weeks subcutaneously with Ab46-D-Glu20 doses of 0, 0.5, 1.6, or 5 nmol/kg/day. Heart rates and visual observations of wellbeing were recorded daily. Weights were recorded weekly. Blood panels were taken before the start and at the end of each experiment.
4.14. Statistical analysis
Earlier work with mice has established that, for one-tailed comparisons, seven animals/group are needed for an α ≤ 0.05, and a power of 0.80 result. As this number is speculative and as surgical complications do occur, 10 mice/group were employed in all studies. Statistical differences were determined using ANOVA, and where appropriate, Dunnett’s post hoc analyses were performed.
Supplementary Material
supplementary
Acknowledgments
We greatly appreciate the Purdue University Imaging Core, Pain Core, and Vet School. We thank Gabby Setiva, Sarah Patshnick, and Lina Trigg for their contributions in the laboratory and data collection.
Funding
This research was supported in part by gifted funds from Purdue University, Indiana CTSI, and National Institute of Dental & Craniofacial Research of the National Institutes of Health under Award Number R44DE028713. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Data and materials availability
All data is available in the main text or the supplementary materials.
Declaration of Competing Interest
P.S.L and S.A.L own stock in Quince Tx. S.A.L, C.M.C, M.T. and J.J.N are employees of Quince Tx.
References
[1].Fracture Trends, BMUS: The Burden of Musculoskeletal Diseases in the United States, available at, https://www.boneandjointburden.org/2014-report/via23/fracture-trends. [2].Karataglis D, Stavridis SI, Petsatodis G, Papadopoulos P, Christodoulou A, New trends in fixation of proximal humeral fractures: a review, Injury 42 (2011) 330–338, 10.1016/j.injury.2010.10.016. [DOI] [PubMed] [Google Scholar] [3].Agarwal A, Nonunions: Diagnosis, Evaluation and Management, Springer, 2017. [Google Scholar] [4].Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A, Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025, J. Bone Miner. Res 22 (2007) 465–475, 10.1359/jbmr.061113. [DOI] [PubMed] [Google Scholar] [5].Nielsen JJ, Low SA, Ramseier NT, Hadap RV, Young NA, Wang M, Low PS, Analysis of the bone fracture targeting properties of osteotropic ligands, J. Control. Release (2020), 10.1016/j.jconrel.2020.09.047. [DOI] [PubMed] [Google Scholar] [6].James AW, LaChaud G, Shen J, Asatrian G, Nguyen V, Zhang X, Ting K, Soo C, A review of the clinical side effects of bone morphogenetic protein-2, Tissue Eng Part B Rev. 22 (2016) 284–297, 10.1089/ten.TEB.2015.0357. [DOI] [PMC free article] [PubMed] [Google Scholar] [7].Nielsen JJ, Low SA, Bone-targeting systems to systemically deliver therapeutics to bone fractures for accelerated healing, Curr. Osteoporos Rep 18 (2020) 449–459, 10.1007/s11914-020-00604-4. [DOI] [PMC free article] [PubMed] [Google Scholar] [8].Leamon CP, Parker MA, Vlahov IR, Xu L-C, Reddy JA, Vetzel M, Douglas N, Synthesis and biological evaluation of EC20: a new folate-derived, 99m Tc-based radiopharmaceutical, Bioconjug. Chem 13 (2002) 1200–1210, 10.1021/bc050122m. [DOI] [PubMed] [Google Scholar] [9].Börjesson AE, Windahl SH, Karimian E, Eriksson EE, Lagerquist MK, Engdahl C, Antal MC, Krust A, Chambon P, Sävendahl L, Ohlsson C, The role of estrogen receptor-α and its activation function-1 for growth plate closure in female mice, Am. J. Physiol 302 (2012) E1381–E1389, 10.1152/ajpendo.00646.2011. [DOI] [PMC free article] [PubMed] [Google Scholar] [10].Dong JZ, et al. , Highly potent analogs of human parathyroid hormone and human parathyroid hormone-related protein, in: Lebl M, Houghten RA (Eds.), Peptides: The Wave of the Future. American Peptide Symposia 7, Springer, Dordrecht, 2001, 10.1007/978-94-010-0464-0_311. [DOI] [Google Scholar] [11].Chew CK, Clarke BL, Abaloparatide: recombinant human PTHrP (1-34) anabolic therapy for osteoporosis, Maturitas 97 (2017) 53–60, 10.1016/j.maturitas.2016.12.003. [DOI] [PubMed] [Google Scholar] [12].Shi Z, Zhou H, Pan B, Lu L, Liu J, Kang Y, Yao X, Feng S, Effectiveness of teriparatide on fracture healing: a systematic review and meta-analysis, PLoS One 11 (2016), e0168691, 10.1371/journal.pone.0168691. [DOI] [PMC free article] [PubMed] [Google Scholar] [13].Hattersley G, Dean T, Corbin BA, Bahar H, Gardella TJ, Binding selectivity of abaloparatide for PTH-type-1-receptor conformations and effects on downstream signaling, Endocrinology 157 (2015) 141–149, 10.1210/en.2015-1726. [DOI] [PMC free article] [PubMed] [Google Scholar] [14].Guidance for Industry: Estimating the Maximum Safe Starting Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteers, Center for Drug Evaluation and Research (CDER), 2005. http://purl.access.gpo.gov/GPO/LPS119812. [Google Scholar] [15].Cheung WH, Miclau T, Chow SK-H, Yang FF, Alt V, Fracture healing in osteoporotic bone, Injury 47 (2016) S21–S26, 10.1016/j.injury.2007.02.014. [DOI] [PubMed] [Google Scholar] [16].Henderson S, Ibe I, Cahill S, Chung Y-H, Lee FY, Bone quality and fracture-healing in type-1 and type-2 diabetes mellitus, J. Bone Joint Surg. Am 101 (2019) 1399–1410, 10.2106/JBJS.18.01297. [DOI] [PubMed] [Google Scholar] [17].Nyman JS, Even JL, Jo CH, Herbert EG, Murry MR, Cockrell GE, Wahl EC, Bunn RC, Lumpkin CK, Fowlkes JL, Thrailkill KM, Increasing duration of type 1 diabetes perturbs the strength-structure relationship and increases brittleness of bone, Bone 48 (2011) 733–740, 10.1016/j.bone.2010.12.016. [DOI] [PMC free article] [PubMed] [Google Scholar] [18].Zhang X, Fei Y, Zhang M, Wei D, Li M, Ding W, Yang J, Reversal of osteoporotic changes of mineral composition in femurs of diabetic rats by insulin, Biol. Trace Elem. Res 121 (2008) 233–242, 10.1007/s12011-007-8043-1. [DOI] [PubMed] [Google Scholar] [19].Bunck MC, Eliasson B, Cornér A, Heine RJ, Shaginian RM, Taskinen MR, Yki-Järvinen H, Smith U, Diamant M, Exenatide treatment did not affect bone mineral density despite body weight reduction in patients with type 2 diabetes, Diabetes Obes. Metab 13 (2011) 374–377, 10.1111/j.1463-1326.2010.01355.x. [DOI] [PubMed] [Google Scholar] [20].Alblowi J, Kayal RA, Siqueria M, McKenzie E, Krothapalli N, McLean J, Conn J, Nikolajczyk B, Einhorn TA, Gerstenfeld L, Graves DT, High levels of tumor necrosis factor-α contribute to accelerated loss of cartilage in diabetic fracture healing, Am. J. Pathol 175 (2009) 1574–1585, 10.2353/ajpath.2009.090148. [DOI] [PMC free article] [PubMed] [Google Scholar] [21].Noldus LP, Spink AJ, Tegelenbosch RA, EthoVision: a versatile video tracking system for automation of behavioral experiments, Behav. Res. Methods Instrum. Comput 33 (2001) 398–414, 10.3758/bf03195394. [DOI] [PubMed] [Google Scholar] [22].Cobos EJ, Ghasemlou N, Araldi D, Segal D, Duong K, Woolf CJ, Inflammation-induced decrease in voluntary wheel running in mice: a non-reflexive test for evaluating inflammatory pain and analgesia, Pain 153 (2012) 876–884, 10.1016/j.pain.2012.01.016. [DOI] [PMC free article] [PubMed] [Google Scholar] [23].Nozaki C, Bourdonnec BL, Reiss D, Windh RT, Little PJ, Dolle RE, Kieffer BL, Gavériaux-Ruff C, δ-Opioid mechanisms for ADL5747 and ADL5859 effects in mice: analgesia, locomotion, and receptor internalization, J. Pharmacol. Exp. Ther 342 (2012) 799–807, 10.1124/jpet.111.188987. [DOI] [PMC free article] [PubMed] [Google Scholar] [24].Choi S, Na HS, Kim J, Lee J, Lee S, Kim D, Park J, Chen C-C, Campbell KP, Shin H-S, Attenuated pain responses in mice lacking CaV3.2 T-type channels, Genes Brain Behav. 6 (2007) 425–431, 10.1111/j.1601-183X.2006.00268.x. [DOI] [PubMed] [Google Scholar] [25].Low SA, Galliford CV, Yang J, Low PS, Kopeček J, Biodistribution of fracture-targeted GSK3β inhibitor-loaded micelles for improved fracture healing, Biomacromolecules 16 (2015) 3145–3153, 10.1021/acs.biomac.5b00777. [DOI] [PMC free article] [PubMed] [Google Scholar] [26].Tella SH, Kommalapati A, Correa R, Profile of abaloparatide and its potential in the treatment of postmenopausal osteoporosis, Cureus 9 (2017), 10.7759/cureus.1300. [DOI] [PMC free article] [PubMed] [Google Scholar] [27].Epstein NE, Complications due to the use of BMP/INFUSE in spine surgery: the evidence continues to mount, Surg. Neurol. Int 4 (2013) S343–S352, 10.4103/2152-7806.114813. [DOI] [PMC free article] [PubMed] [Google Scholar] [28].Bentler SE, Liu L, Obrizan M, Cook EA, Wright KB, Geweke JF, Chrischilles EA, Pavlik CE, Wallace RB, Ohsfeldt RI, Jones MP, Rosenthal GE, Wolinsky FD, The aftermath of hip fracture: discharge placement, functional status change, and mortality, Am. J. Epidemiol 170 (2009) 1290–1299, 10.1093/aje/kwp266. [DOI] [PMC free article] [PubMed] [Google Scholar] [29].Moran DS, Evans RK, Hadad E, Imaging of lower extremity stress fracture injuries, Sports Med. 38 (2008) 345–356, 10.2165/00007256-200838040-00005. [DOI] [PubMed] [Google Scholar] [30].Brenner AI, Koshy J, Morey J, Lin C, DiPoce J, The bone scan, Semin. Nucl. Med 42 (2012) 11–26, 10.1053/j.semnuclmed.2011.07.005. [DOI] [PubMed] [Google Scholar] [31].Ferreira RL, de Almeida SM, Bóscolo FN, Santos AO, Camargo EE, Bone scintigraphy as an adjunct for the diagnosis of oral diseases, J. Dent. Educ 66 (2002) 1381–1387, 10.1002/j.0022-0337.2002.66.12.tb03612.x. [DOI] [PubMed] [Google Scholar] [32].Lynch JR, Taitsman LA, Barei DP, Nork SE, Femoral nonunion: risk factors and treatment options, J. Am. Acad. Orthop. Surg 16 (2008) 88–97, 10.5435/00124635-200802000-00006. [DOI] [PubMed] [Google Scholar] [33].Moberg L, Lunds universitet, fractures in postmenopausal women. Background factors and fracture risk assessment, Lund University, Faculty of Medicine, Lund, 2016. [Google Scholar] [34].Wehren LE, Magaziner J, Hip fracture: risk factors and outcomes, Curr. Osteoporos. Rep 1 (2003) 78–85, 10.1007/s11914-003-0013-8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
supplementary