Effects of risedronate, alendronate, and minodronate alone or in combination with eldecalcitol on bone mineral density, quality, and strength in ovariectomized rats

Combination therapy of active vitamin D3 with some bisphosphonates (BPs) has been reported to be clinically beneficial. However, combination therapy of eldecalcitol (ELD) with BP has to date not been validated as to whether it is beneficial in the clinical setting. Preclinical studies suggested that simultaneous treatment with ELD and some BPs is more effective than monotherapy. However, the relative potency of various BPs, when used in combination with ELD, is completely unknown. In this study, we examined and compared the effects of risedronate (RIS), alendronate (ALN), and minodronate (MIN) alone or in combination with ELD on bone mass, microarchitecture, strength, and material properties in ovariectomized Sprague-Dawley rats aged 13 weeks. RIS, ALN, MIN, and ELD were administered five times weekly for 16 weeks. Micro-computed tomography analysis, compression test, and Fourier transform infrared (FTIR) imaging analysis were performed 16 weeks after treatment initiation. Trabecular and cortical bone mineral density (BMD) in the fourth lumbar vertebra (L4) significantly increased in the RIS + ELD, ALN + ELD, and MIN + ELD groups compared with the vehicle group. Moreover, the bone microarchitecture of L4 in all the BP + ELD groups also significantly improved. On mechanical testing of L4, the maximum load was significantly increased in the RIS + ELD and ALN + ELD groups. FTIR analysis revealed that the mineral-to-collagen ratio of trabecular bone in L3 of all the BP + ELD groups was significantly increased compared with the vehicle group. By contrast, the carbonate-to-phosphate ratio, a parameter of mineral immaturity, was significantly decreased in the RIS + ELD and ALN + ELD groups. BP + ELD improved the BMD and structural properties of the bone to a similar extent. RIS + ELD and ALN + ELD also improved bone strength. Furthermore, treatment with BP + ELD improved the bone material. These results suggest that the combination therapy of BP and ELD is beneficial and warrants further clinical trials.


Introduction
Nitrogen-containing bisphosphonates (BPs) have been used as firstline drugs to treat osteoporosis for over 30 years (Cosman et al., 2014;Schweser and Crist, 2017). The BPs alendronate (ALN), risedronate (RIS), and minodronate (MIN) have a similar mechanism of anti-resorptive action. They inhibit farnesyl pyrophosphate synthase (FPPS) in the intracellular mevalonate pathway to disrupt protein isoprenylation, thereby causing osteoclast dysfunction and apoptosis (Rogers et al., 2011). The potential for inhibiting the activity of FPPS among these three BPs has been reported as MIN > RIS > ALN (Dunford et al., 2001). Differences in the affinity for hydroxyapatite have also Abbreviations: ALF, alfacalcidol; ALN, alendronate; BMD, bone mineral density; BPs, bisphosphonates; ELD, eldecalcitol; FTIR, Fourier transform infrared; micro-CT, micro-computed tomography; MIN, minodronate; OVX, ovariectomized; RIS, risedronate. been reported (Rogers et al., 2011). Although such differences between these BPs result in distinct anti-resorptive activity in vitro, this does not necessarily predict clinical efficacy in cases of fracture. RIS has previously been reported to display a significantly greater potency for fracture prevention in patients with osteoporosis than ALN in the first year of therapy (Silverman et al., 2007). We also reported that RIS might exhibit better effects on the bone quality compared with ALN in rat models (Yano et al., 2014;Yano et al., 2017). However, the mechanisms of the distinct effects of each BP remain to be fully understood.
Active vitamin D 3 drugs, including calcitriol and alfacalcidol (ALF), have been used as therapeutic agents for osteoporosis (Cianferotti et al., 2015). In addition to enhancing the physiological action of vitamin D, thereby increasing the calcium absorption from the small intestine, these drugs have been suggested to regulate bone metabolism through direct action on the bone cells as well as suppression of parathyroid hormone secretion from the parathyroid glands (Ryan et al., 2013;Urena-Torres and Souberbielle, 2014). Nevertheless, the anti-fracture efficacy of active vitamin D 3 as a monotherapy appears limited (Papadimitropoulos et al., 2002;O'Donnell et al., 2008). Eldecalcitol (ELD), a calcitriol analog developed in Japan, has shown a superior effect to ALF in suppressing bone turnover, increasing the bone mineral density (BMD), and preventing vertebral fracture (Matsumoto and Kubodera, 2007;Matsumoto et al., 2011;Nakamura et al., 2013;Xu et al., 2016), presumably because of its unique pharmacological properties, including higher affinity for vitamin D-binding protein (Okano et al., 1991). ELD has also been reported to increase the bone mass by mini-modeling (Saito et al., 2013) and/or bone formation (Saito et al., 2015).
Several studies have suggested that the addition of active vitamin D 3 to BP therapy is further beneficial (Ringe et al., 2007;Felsenberg et al., 2011;Orimo et al., 2011;Schacht and Ringe, 2011). Because ELD has a unique anti-resorptive effect in addition to its ability to promote calcium absorption similar to that of authentic active vitamin D 3 , its combination with BPs appears to be a promising treatment. Indeed, some combinations of ELD and BP have already been shown to be clinically more effective in suppressing bone turnover and improving the gain in BMD than monotherapy with BP alone (Sakai et al., 2015a;Ebina et al., 2016), although the effect on the fracture incidence has not been reported. An animal study demonstrated that ALN in combination with ELD significantly improved the BMD and bone strength compared with ALN monotherapy in ovariectomized (OVX) rats (Sakai et al., 2012). Another study in rats similarly indicated that a combination of ALN and ELD was superior in increasing the bone mass and improving the quality compared with a combination of ALN and ALF (Sugimoto et al., 2013). These studies also suggested that ALN and ELD improved bone microarchitecture by different mechanisms. Considering the unique characteristics of each BP, it is plausible that different BPs might display distinct effects on bone mass and quality when combined with ELD.
In the present study, we examined the effects of three BPs alone or in combination with ELD in OVX rats. To determine the drug effects on bone quality, we performed Fourier transform infrared (FTIR) spectroscopic imaging in addition to microstructural analysis by microcomputed tomography (micro-CT). FTIR imaging is a powerful tool to assess bone quality in that it can determine the mineral-to-collagen matrix (PO 4 3− /amide I) and carbonate-to-phosphate (CO 3 2− /PO 4 3− ) ratios, crystallinity, mineral maturity, and collagen cross-links and fiber orientation (Paschalis et al., 2011). However, it has never been applied to evaluate the effect of active vitamin D 3 or its analog, except for native vitamin D 3 supplementation (Paschalis et al., 2017). This study is the first direct comparison of the effects of three different BPs in combination with ELD, including quality analysis by FTIR imaging.

Reagents and animals
RIS was synthesized by Takeda Pharmaceutical Company (Tokyo, Japan). ALN, MIN, and ELD were purchased from Wako Pure Chemical Industries (Tokyo, Japan), Chengdu-D-Innovation pharmaceutical (Chengdu, China), and Chugai Pharmaceutical (eldecalcitol capsule, Tokyo, Japan), respectively. Saline and medium-chain triglyceride oil (MCT-oil) were obtained from Otsuka Pharmaceutical (Tokushima, Japan) and Nisshin Oillio Group, Ltd. (Tokyo, Japan), respectively. Isoflurane was purchased from Mylan (Tokyo, Japan). Eighty-one female Sprague-Dawley rats aged 12 weeks were purchased from Charles River Laboratories Japan (Yokohama, Japan). The animals were housed three per cage in a room with a light/dark cycle of 12/12 h and were acclimated to the study conditions for 1 week prior to the start of the experiment. All animals were paired-fed rodent food (CRF-1 formula, Charles River Laboratories Japan). Water from Ajinomoto Pharmaceuticals in-house production was available ad libitum. The experimental protocol was approved by the Animal Care and Use Committee of Ajinomoto Pharmaceuticals Co., Ltd., which was operated in accordance with the Guideline for Proper Conduct of Animal Experiments (Science Council of Japan, 2006).

Experimental design
The animals when aged 13 weeks were assigned to sham surgery (sham, n = 9) or OVX (n = 72) groups by matching their body weight as well as the trabecular BMD of the fourth lumbar vertebra (L4) and the cortical BMD of the left femur by micro-computed tomography (micro-CT, Skyscan1176; Bruker microCT, Kontich, Belgium) using SAS 9.3 (SAS Institute, Cary, USA). The OVX group was further subdivided into the following eight groups: (1) vehicle (saline and MCT-oil), (2) RIS, (3) ALN, (4) MIN, (5) ELD, (6) RIS + ELD, (7) ALN + ELD, and (8) MIN + ELD, with nine animals in each group. RIS (0.7 μg/kg), ALN (1.4 μg/kg), and MIN (0.14 μg/kg) were injected subcutaneously five times weekly for 16 weeks. ELD (0.04 μg/kg) was administered orally five times weekly for 16 weeks. Saline and MCT-oil were administered subcutaneously and orally, respectively, five times weekly for 16 weeks in the vehicle and sham groups. Drug administration was initiated on the day after OVX or sham surgery. The animals were euthanized by exsanguination under anesthesia with isoflurane (2.5-3%) at 16 weeks after the initiation of drug administration and the lumbar vertebrae and femurs were evaluated for the BMD and microarchitecture by micro-CT, bone strength by mechanical testing, and bone quality by FTIR imaging analysis.

Structural analysis by micro-CT
To investigate the effects of BP alone or in combination with ELD treatment on the trabecular and cortical microarchitecture of the L4 and distal femur, in vivo three-dimensional trabecular and cortical analyses were performed by micro-CT. Animals were anesthetized with isoflurane (2.5-3%) before being scanned in a supine position on an animal bed with the left hind leg fixed in a cylindrical plastic holder. The trabecular bone of the femur and the trabecular and cortical bone of the lumbar vertebrae were scanned in 50 slices (0.19 cm total thickness). The cortical bone of the femur was scanned in 130 slices (0.5 cm total thickness). In the trabecular bone, we determined the following histomorphometric parameters on the resultant three-dimensional images: bone volume fraction (BV/TV, %), trabecular number (Tb.N, 1/mm), trabecular thickness (Tb.Th, mm), trabecular separation (Tb.Sp, mm), connectivity density (1/mm 3 ), and BMD (g/cm 3 ). In the cortical bone, we measured the following histomorphometric parameters on the twodimensional images: cortical bone area (Ct.Ar, mm 2 ), cortical thickness (Ct.Th, mm), and BMD (g/cm 3 ) (Dempster et al., 2013). The analysis was performed using CTAn (ver. 1.11.4.2, Bruker).

Mechanical testing
The fourth lumbar vertebrae (L4) were tested for bone strength by a compression test using a previously described method with slight modification (Ohnishi et al., 1997). The left femurs were tested for bone strength using the three-point bending test (Yano et al., 2014). Biomechanical testing was performed using the TK-252/MK (Muromachi Kikai, Tokyo, Japan). For compression of the vertebrae, the L4 was fixed on a holder with an adhesive agent and placed on the testing platform. In three-point bending of the femur, the two support points below the bone were interspaced at 13 mm. The deformation rates of the vertebrae and femurs were 5 and 2 mm/min, respectively. From the load-deformation curve, the stiffness (N/mm) was calculated as previously reported (Turner and Burr, 1993). The maximum load (N) was obtained directly from the load-deformation curve. Stiffness was defined by the slope of the ascending portion of the load-deformation curve between 30% and 70% of the maximum load value.

FTIR imaging analysis
The right femurs and third lumbar vertebrae (L3) were removed and embedded in polymethyl methacrylate (PMMA). Longitudinal sections of 3-μm thickness were prepared using a microtome (Aparicio et al., 2002). FTIR images of the longitudinal sections were acquired using an FTIR imaging system with a mercury-cadmium-telluride linear array detector (Spotlight 400 system, PerkinElmer, Waltham, MA, USA) in transmittance mode with a frequency range of 4000-680 cm − 1 , a resolution of 4 cm − 1 , and a pixel size of 25 × 25 μm. The background spectrum was obtained through a barium fluoride window. Thirty FTIR spectra extracted from each trabecular and cortical bone in the image were obtained with baseline collection and PMMA spectral subtraction using Spectrum 10 software (PerkinElmer) and subsequently used to evaluate the PO 4 3− /amide I and CO 3 2− /PO 4 3− in the femur and L3. The PO 4 3− /amide I was calculated by integrating the area of the phosphate band (PO 4 3− , 1181-906 cm − 1 ) and dividing it by the area of the amide I band (1712-1609 cm − 1 ). The CO 3 2− /PO 4 3− was calculated by integrating the area of the carbonate band (CO 3 2− , 890-851 cm − 1 ) and dividing it by the area of the PO 4 3− band.

Statistical analysis
All data are expressed as mean ± standard deviation (SD). According to the results of the Shapiro-Wilk test for normal distribution, group differences were analyzed by t-test or Wilcoxon test and analysis of variance with post-hoc Dunnett's test, Tukey test, Steel or Steel-Dwass test on JMP (ver. 12.0.1; SAS Institute, Tokyo, Japan). P values of <0.05 were considered statistically significant.

BMD, bone microarchitecture, and mechanical testing of the fourth lumbar vertebrae
Trabecular BMD and bone microarchitecture in the L4 vertebrae were significantly lower in the vehicle-treated OVX rats than in the sham group (Fig. 1A). All the drug groups displayed a significantly increased trabecular BMD compared with the vehicle group. ALN + ELD and MIN + ELD significantly increased the trabecular BMD compared with that of each monotherapy (Fig. 1A). All combinations of BP + ELD and almost all the monotherapies consistently and significantly improved the trabecular bone microarchitecture compared with that in the vehicle   group (Table 1). In particular, the Tb.N in the ALN + ELD and MIN + ELD groups was higher compared with each BP alone, and all combinations with BP were better than ELD alone. All combinations significantly increased the cortical BMD in L4 compared with that in the vehicle group, whereas the BMD changes by each monotherapy were non-significant (Fig. 1B). Additionally, RIS + ELD and MIN + ELD significantly increased the cortical BMD compared with that of each monotherapy (Fig. 1B). The Ct.Ar was significantly decreased in the vehicle-treated OVX rats compared with that in the sham group. All combination therapies significantly increased the Ct.Ar and Ct.Th compared with that in the vehicle group and in each monotherapy group (Table 1).
The compression test for L4 bone strength revealed that the maximum load in the vehicle-treated OVX rats was significantly lower than in the sham group ( Fig. 2A). Stiffness was also slightly decreased in the vehicle group without statistical significance (Fig. 2B). RIS + ELD and ALN + ELD significantly increased the maximum load and stiffness compared with that in the vehicle group. MIN + ELD also increased these parameters (+44% and + 17%, respectively); however, the effect did not reach statistical significance ( Fig. 2A and B). None of the BPs significantly improved the mechanical parameters when used alone. Therefore, RIS and ALN improved mechanical properties of the vertebral body when used in combination with ELD.

BMD and bone microarchitecture of the distal femur
In the analysis of femurs, the vehicle group displayed significantly reduced trabecular BMD and deteriorated bone microarchitecture of the distal femur compared with the sham group ( Fig. 1C and Table 2). Combination therapy significantly increased the trabecular BMD of the distal femur compared with that in the vehicle group and each monotherapy group (Fig. 1C). Additionally, combination therapy significantly improved the BV/TV, Tb.Th, Tb.N, and Tb.Sp compared with that in the vehicle group (Table 2). Moreover, combination therapy significantly improved the BV/TV, Tb.Th, and Tb.Sp compared with that of each monotherapy, clearly demonstrating the beneficial effects of cotreatment with BP and ELD. None of the drug-treated groups displayed significant changes in the BMD (Fig. 1D), bone microarchitecture (Table 2), or strength (data not shown) in the cortical bone of the distal femur at 16 weeks.

FTIR imaging analysis
Figs. 3 and 4 show FTIR images and bone quality-related parameters of the L3 and femurs, respectively. In L3, the vehicle group displayed a significantly decreased PO 4 3− /amide I in both trabecular and cortical bone compared with the sham group ( Fig. 3B and D). All the drugtreated groups showed a significantly increased PO 4 3− /amide I in the trabecular bone compared with the vehicle group (Fig. 3B). Moreover, RIS + ELD and MIN + ELD significantly increased the PO 4 3− /amide I compared with that of each monotherapy. The cortical PO 4 3− /amide I of L3 was significantly increased in the ALN + ELD and MIN + ELD groups compared with that in the vehicle-treated OVX rats (Fig. 3D). By contrast, the CO 3 2− /PO 4 3− in the L3 trabecular bone tended to increase in the vehicle-treated OVX rats compared with that in the sham group and significantly decreased on treatment with RIS or ALN alone or in combination with ELD (Fig. 3C). The CO 3 2− /PO 4 3− of the trabecular bone in the MIN + ELD group was significantly higher compared with the RIS + ELD and ALN + ELD groups (Fig. 3C). Similarly, in the cortical bone in L3 (Fig. 3E), the CO 3 2− /PO 4 3− was significantly increased in the vehicle-treated OVX rats compared with that in the sham group, and all drugs caused a significant decrease in the CO 3 2− /PO 4 3− .
In both trabecular and cortical bone of the femur, the PO 4 3− /amide I was significantly decreased in the vehicle-treated OVX rats compared with that in the sham group ( Fig. 4B and D). Almost all the drugs significantly increased the PO 4 3− /amide I compared with that in the vehicle group ( Fig. 4B and D). Treatment with ALN + ELD and MIN + ELD significantly increased the PO 4 3− /amide I compared with that with ELD in the trabecular bone (Fig. 4B). In the cortical bone of the femur, almost all the drugs significantly increased the PO 4 3− /amide I; however, the benefit of combination therapy was not evident (Fig. 4D). Regarding the CO 3 2− /PO 4 3− in the femur, OVX did not cause any significant changes in either the trabecular or cortical bone ( Fig. 4C and E). BP alone and in combination with ELD displayed little or no effect on the CO 3 2− /PO 4 3− , except MIN + ELD, which significantly decreased the ratio compared with that in the vehicle group ( Fig. 4C and E).

Discussion
In the present study, we investigated the effect of three different BPs, alone and in combination with ELD, by direct comparison in OVX rats. The results indicated that ELD co-treatment with each of the three BPs tested displayed consistent beneficial effects on bone mass and microarchitecture on micro-CT as well as on material properties evaluated by FTIR, and consequently, on mechanical strength.
Regarding the volumetric BMD, BPs and ELD displayed apparent additive or synergistic effects, particularly in L4. The increase in the trabecular and cortical BMD of the lumbar spine appears to be directly reflected in the improved mechanical properties, as demonstrated by the compression test. The advantage of combination therapy with BP and ELD over BP monotherapy is consistent with previous reports based on dual energy X-ray absorptiometry-measured areal BMD using ALN (Sakai et al., 2012;Sugimoto et al., 2013) or ibandronate (Sakai et al., 2015b). In the femur, we observed a prominent positive effect of combination therapy on the trabecular BMD in contrast to little if any effect on the cortical bone. We also observed clear effects on the trabecular structural parameters, including the Tb.Th and Tb.Sp. The results suggested that the previously reported increase in femoral areal BMD (Sakai et al., 2012;Sugimoto et al., 2013;Sakai et al., 2015a) by combination therapy with BP and ELD was mostly attributed to the effect on the trabecular bone.
The current study was the first to evaluate the effect of ELD alone or in combination with BPs on bone quality by FTIR analysis. FTIR imaging analysis of L3 revealed that all the BPs and their combination therapy with ELD significantly improved the PO 4 3− /amide I in the trabecular bone, which had been reduced by ovariectomy. Moreover, the improvement in the PO 4 3− /amide I was greater with the combination therapy with ELD compared with that of each BP alone. These results indicated that the addition of ELD to BPs further improved the PO 4 3− /a (Aparicio et al., 2002) mide I in the trabecular bone of the lumbar vertebrae, which most likely additionally contributed to better mechanical strength. It is notable that changes in the PO 4 3− /amide I by combination therapy appeared to be parallel to those in the volumetric BMD; however, the additive effect of ELD was relatively modest. It thus appears likely that co-treatment with ELD induced an increase in the BMD as well as improved microarchitecture, ultimately resulting in greater physical strength of the bone tissue without hypermineralization. Conversely, an increased PO 4 3− /amide I also suggested that, at least in the context of combination therapy with BPs, the contribution of mini-modeling-based new bone formation to an increase in the BMD was minor. All the ELD and BP combinations reversed the ovariectomy-induced increase in the CO 3 2− /PO 4 3− in both the trabecular and the cortical bone of L3. Our observations are consistent with previous studies (Hara et al., 2007;Gadeleta et al., 2000) that have shown that ovariectomy increased the CO 3 2− /PO 4 3− in the femurs of OVX rats and lumbar vertebrae of OVX monkeys. The increased CO 3 2− /PO 4 3− in OVX animals is considered to  Data are presented as mean ± SD. n = 9. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared with the sham group. a p < 0.05, b p < 0.01, and c p < 0.001 compared with the vehicle group. † p < 0.05, † † p < 0.01, and † † † p < 0.001 compared with RIS. ‡ p < 0.05, ‡ ‡ p < 0.01, ‡ ‡ ‡ p < 0.001, and ‡ ‡ ‡ ‡ p < 0.0001 compared with ALN. # p < 0.05, ## p < 0.01, ### p < 0.001, and #### p < 0.0001 compared with MIN. $ p < 0.05, $$ p < 0.01, $$$ p < 0.001, and $$$$ p < 0.0001 compared with ELD. ¶ p < 0.05 compared with ALN + ELD. Abbreviations: ALN, alendronate; BP, bisphosphonate; Conn D, connectivity density; ELD, eldecalcitol; MIN, minodronate; OVX, ovariectomized; RIS, risedronate.
result from a reduction of mineral maturity. Our results suggested that the mineral in OVX rats was poorly crystallized apatite, and treatment with all the BPs or their combinations with ELD improved the mineral maturity, increasing the content of stoichiometric apatite in both trabecular and cortical bone in L3. Moreover, the effect of RIS + ELD and ALN + ELD on the CO 3 2− /PO 4 3− in trabecular bone of L3 correlated with the bone strength. Therefore, bone strength was strongly reflected by mineral maturity in the present study. In the femur, no reduction of the CO 3 2− /PO 4 3− by combination therapy was detected, except in rats treated with MIN + ELD. The reason for this result is currently unknown, although the strongest anti-resorptive activity of MIN among the three BPs investigated in the current study (Dunford et al., 2001) might be in some way associated with the large increase in the PO 4 3− /amide I and decrease in the CO 3 2− /PO 4 3− . The significance of this observation remains to be determined. Although crystallinity and collagen cross-links were evaluated in the present study, all BPs and their combination with ELD did not induce significant changes compared with the vehicle group (data not shown). A previous study (Durchschlag et al., 2006) showed that RIS significantly decreased the crystallinity and collagen cross-links near bone-forming sites in trabecular bone of iliac crest biopsies in a time-dependent manner. However, RIS did not affect these parameters near the resorbing sites. This may be because the sites where we analyzed these parameters were randomly chosen from the whole bone of L3 in this study. We speculate that the effect of ELD and BPs on these parameters is significant only when the forming and resorbing sites are analyzed separately after a long-term treatment.
There are limitations to the present study. First, the dosage of each of the three BPs was determined empirically based on the literature and our experience, so that it was proportional to the clinical dosage. Second, because BPs and ELD were administered immediately after OVX, the observed effect was more representative of prevention rather than treatment, and may thus not be applicable to established postmenopausal osteoporosis. The relevance of our findings to the clinical setting should thus be carefully evaluated. Finally, the study was not designed to fully elucidate the mechanism of drug action and did not provide information about bone metabolism using biochemical markers or dynamic histomorphometry. Nonetheless, the strength of the current study is the fact that three BPs were tested at the same time, allowing direct comparison for the first time in the context of co-treatment with ELD. This study is also novel in that the effect of combination therapy with BP and ELD on the material properties of the bone was evaluated by FTIR imaging.

Conclusion
Combinations of ELD and three nitrogen-containing BPs, namely RIS, ALN, and MIN, significantly improved the bone mass and microarchitecture in OVX rats, consequently leading to increased mechanical strength. The improvement with the combination therapy was significantly superior to that with each BP or ELD alone. FTIR imaging analysis excluded hypermineralization and compromised mineral maturity. Therefore, combination therapy with ELD and BP can be validated as a potential therapeutic approach for the treatment of osteoporosis in the clinical setting.

Declaration of competing interest
The authors declare that they have no conflict of interest in this study.
HK has received research funding from EA pharma (formerly known as Ajinomoto Pharmaceutical).
DI has received honoraria and research funding from Astellas Pharma, Amgen, EA Pharma, Chugai Pharmaceutical, Daiichi Sankyo, Taisyo Pharmaceutical, and Teijin Pharma.