Pantoprazole, a proton pump inhibitor, does not prevent botulinum toxin induced disuse osteopenia in mice.

OBJECTIVES: Pantoprazole is a proton pump inhibitor that has been shown to inhibit bone resorption. The aim of the study was to investigate whether pantoprazole can prevent development of botulinum toxin (BTX)-induced disuse osteopenia in mice.
METHODS: Forty-eight 16-week-old female C57BL/6J mice were randomized into 4 groups (n=12): Base, Ctrl, BTX, and BTX+Pan. The Base group was euthanized at study start. The BTX and BTX+Pan groups were immobilized by injections with BTX in one hind limb. The BTX+Pan group was injected i.p. daily with 100 mg pantoprazole per kg bodyweight. The mice were euthanized after 3 weeks of treatment. The skeletal status was investigated by DEXA, µCT, mechanical testing, dynamic bone histomorphometry, and RT-qPCR. The bone sites investigated were tibia, femur, L5 vertebra, and humerus.
RESULTS: Injections of BTX induced a pronounced and significant loss of bone density, microstructure, and strength in the immobilized hind limb. Furthermore, the localized intramuscular injections of BTX lead to a slight loss of bone and bone strength at the L5 vertebra and humerus. Treatment with pantoprazole did not have any bone protective or deleterious effects.
CONCLUSION: Pantoprazole was unable to prevent the development of BTX induced disuse osteopenia in skeletally mature female C57BL/6J mice.

Introduction

Currently, most osteoporosis treatments aim at inhibiting osteoclasts thereby reducing bone resorption. Interestingly, this approach resembles the rare bone disease osteopetrosis that results from osteoclastic dysfunction, in which bone resorption is impaired creating a sclerotic and brittle skeleton[1]. Monogenetic mutations have been linked to this disease and in every second osteopetrotic patient a defect is found in the osteoclastic vacuolar H+-ATPase proton pump (V-ATPase)[2]. Consequently, acidification of the osteoclastic resorption pit is insufficient, which subsequently obstructs bone resorption and renewal of the skeleton[3]. Accordingly, by pharmaceutically blocking the V-ATPase in conditions characterized by high bone turnover, e.g. disuse osteopenia and osteoporosis, it might be possible to rescue or prevent the bone loss associated[4].

Proton pump inhibitors (PPIs) are widely used for treatment of stomach related acid disorders, e.g. reflux and ulcers. They work by irreversibly inhibiting the H+/K+-ATPase located in the canalicular membrane of parietal cells in gastric pits in the stomach, thereby reducing gastric acid production[5]. PPIs are pro-drugs, which are accumulated and activated in the acidic milieu that exists in the gastric crypts and in the osteoclastic resorption pits. In vitro studies have suggested that PPIs also have an inhibitory effect on the osteoclastic V-ATPase and thereby are able to reduce bone resorption[6-9]. On the contrary, in vivo studies have been less conclusive as they have shown either no or a slightly negative effect of PPI-treatment on bone integrity[10-13]. In rodents, pantoprazole, a PPI, has been revealed to delay fracture healing and calcium phosphate cement resorption by decreasing both osteoclastic and osteoblastic activity[14,15].

Disuse osteopenia is characterized by a rapid bone loss caused by increased bone resorption and decreased bone formation[16,17]. As suggested, pantoprazole seems to inhibit or reduce osteoclastic activity, and may therefore be able to prevent disuse osteopenia. We and others have extensively investigated the botulinum toxin (BTX) model of disuse osteopenia, in which BTX is injected intramuscularly to paralyze one hind limb in rodents[18-22]. Paralyses is followed by a massive muscle atrophy and bone loss[17,20].

The aim of the present study was to prevent BTX induced disuse osteopenia by daily injections of pantoprazole in female C57BL/6J mice. The skeletal status was investigated with a series of different techniques including Dual Energy X-ray Absorptiometry (DEXA), micro Computed Tomography (μCT), mechanical testing, dynamical bone histomorphometry, and Reverse Transcription quantitative Polymerase Chain Reaction (RT-qPCR).

Materials and methods

Animals

Forty-eight 16-week-old female C57BL/6J mice (Taconic), with a mean body weight (BW) of 22.3±0.8 g, were housed at 20°C with a 12/12 h light/dark cycle. The animals had free access to tap water and standard mice chow (1324, Altromin).

At the age of 15 weeks, one week prior to study start, the animals were randomized according to their BW into 4 groups (n=12): Base, Ctrl, BTX, and BTX+Pan. At study start, the mice in the BTX groups were injected i.m. with 20 IU BTX (Botox, Allergan) per kg BW, distributed equally into the quadriceps muscle and calf muscles of the right hind limb. The Ctrl group was injected i.m. with saline in the same regimen as the BTX injections. The BTX+Pan group was injected i.p. with 100 mg pantoprazole (Pantoloc, Takada Pharma) per kg BW daily. The dosage was chosen based on the study by Histing et al., showing that pantoprazole inhibits fracture healing in mice[14]. The Ctrl and BTX groups were injected i.p. with saline. The mice were injected i.p. with alizarin (20 mg/kg), calcein (20 mg/kg), and tetracycline (20 mg/kg) 12, 8, and 4 days before euthanasia, respectively.

The treatment lasted three weeks after which the mice were euthanized by an overdose of anesthesia (IsoFlo Vet, Orion Pharma Animal Health) and removal of the heart. Each mouse received the last i.p. injection 6 h prior to euthanasia. One mouse in the BTX+Pan group was euthanized prematurely due to an ileus-like condition. The Base group was euthanized at study start to serve as baseline.

At euthanasia, the right tibiae were quickly isolated and carefully cleaned from soft tissue. The distal part was snap frozen in liquid nitrogen and stored at -80°C, while the proximal part was immersion-fixed in 0.1 M sodium phosphate buffered formaldehyde (4% formaldehyde, pH 7.0) for 48 h and then stored in 70% ethanol. The rectus femoris muscles were isolated and the wet weight determined. The right femora, right humeri, and L5 vertebrae were isolated, carefully cleaned from soft tissue, and stored in Ringer’s solution at -20°C. The length of the femora and humeri were measured with a digital sliding caliper.

Dual energy X-ray absorptiometry (DEXA)

The femora and humeri were placed in a DEXA scanner (Sabre XL, Norland Stratec) and scanned with an isotropic pixel size of 0.1 mm. Bone mineral content (BMC) and areal bone mineral density (aBMD) were determined for the whole femur and humerus. Quality assurance was performed by scans of the two solid-state phantoms provided with the scanner.

Micro computed tomography (μCT)

The femoral mid-diaphysis, distal femur, L5 vertebra, humeral mid-diaphysis, and proximal humerus were scanned in a desktop μCT scanner (Scanco μCT 35, Scanco Medical AG). The femoral and humeral mid-diaphyses were scanned with an isotropic voxel size of 7 μm, X-ray voltage of 55 kVp and current of 145 μA, and an integration time of 300 ms. The distal femur and proximal humerus were scanned with an isotropic voxel size of 3.5 μm, X-ray voltage of 55 kVp and current of 145 μA, and an integration time of 800 ms. The L5 vertebra was scanned with an isotropic voxel size of 6 μm, X-ray voltage of 70 kVp and current of 114 μA, and an integration time of 800 ms. Volumes of interest (VOI) were analyzed with the software provided with μCT scanner (version 6.5). The threshold used for segmentation of the 3D data was automatically determined as the minimum point between the marrow and the bone peak in the attenuation histogram using IPL (version 5.11).

The femoral diaphysis was analyzed by placing an 819-μm-high VOI in the middle of the scanned region (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=0.8, support=1) and segmented with a fixed threshold filter (556.4 mg HA/cm3). The humeral diaphysis was analyzed by placing a 420-μm-high VOI including cortical bone only between the deltoid tuberosity and epicondyles (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=0.8, support=1) and segmented with a fixed threshold filter (550.2 mg HA/cm3). The cortical bone analysis included: Bone area, tissue area, marrow area, polar moment of inertia, and bone material density.

Figure 1

Volumes of interest (VOI) analyzed by μCT. The size of the femur, humerus, and vertebra are proportional to each other. The blue area represents the VOI at the diaphyses, the red area represents the VOI at the metaphyses, the green area represents the VOI at the epiphyses, and the pink area represents the VOI at the corpus vertebra.

The distal femoral metaphysis was analyzed by placing a 1-mm-high VOI 0.2 mm proximal to the growth plate including trabecular bone only (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=1, support=2) and segmented with a fixed threshold filter (517.3 mg HA/cm3). The distal femoral epiphysis was analyzed by placing a VOI distal to the growth plate until the intercondylar fossa including trabecular bone only (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=1, support=2) and segmented with a fixed threshold filter (525.5 mg HA/cm3).

The L5 vertebral body was analyzed by placing a VOI between the primary spongiosa including trabecular bone only (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=1.3, support=2) and segmented with a fixed threshold filter (531.7 mg HA/cm3).

The proximal humeral metaphysis was analyzed by placing a 1-mm-high VOI 0.2 mm distal to the growth plate including trabecular bone (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=1.3, support=2) and segmented with a fixed threshold filter (525.5 mg HA/cm3). The proximal humeral epiphysis was analyzed by placing a VOI in the humeral head including the trabecular bone only (Figure 1). The 3D data sets were low-pass filtered using a Gaussian filter (σ=1.3, support=2) and segmented with a fixed threshold filter (535.8 mg HA/cm3). The trabecular bone analysis included: trabecular bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular separation (Tb.Sp), connectivity density (CD), structure model index (SMI), and bone material density.

The μCT analyzes were performed according to the current guidelines[23]. Quality assurance was performed by weekly (density) and monthly (geometry) scans of the solid-state calibration phantom provided with the scanner. 3D visualization and imaging was performed using Amira (version 5.6.0, FEI).

Mechanical testing

The fracture strength of the femoral and humeral diaphysis was determined with a three point bending test. After the μCT scans, the femora and humeri were placed on a custom-made testing jig, supporting the proximal and distal part of the shaft. The distance between the two supporting rods was 7.1 mm. Load was applied at the midpoint of the femora and humeri in an anterior-posterior direction at a constant deflection rate of 2 mm/min until fracture using a materials testing machine (5566, Instron). Subsequently, the proximal part of the femora was placed in another custom made fixation jig for femoral neck testing supporting the shaft under the neck. Load was applied to the top of the femoral head with a constant deflection rate of 2 mm/min until fracture of the femoral neck.

After the μCT scans, the L5 vertebrae were prepared for mechanical testing. The vertebral bodies were isolated by carefully removing the posterior arch and the cranial and caudal intervertebral discs. The height of the vertebral body was measured with a digital micrometer screw gauge and then placed in the materials testing machine. Load was applied in a cranial-caudal direction with a constant deflection rate of 2 mm/min until fracture.

Dynamic bone histomorphometry

After mechanical testing, the distal part of the femora were immersion-fixed in 0.1 M sodium phosphate buffered formaldehyde (4% formaldehyde, pH 7.0) for 48 h, dehydrated in ethanol, embedded undecalcified in methylmetacrylate, and 7-μm-thick sections were cut using a hard tissue microtome (RM 2065, Leica). The sections were mounted unstained on microscope slides and placed in a microscope (Eclipse 80i, Nikon) equipped for fluorescence microscopy. The stereology system newCAST (Version 6.4.1.2240, Visiopharm) was then used to count intersections with fluorochrome labels. Briefly, a region of interest (ROI), excluding cortical bone, was placed at the distal femoral epiphysis. In addition, a 1000-μm-high ROI, excluding cortical bone and primary spongiosa, was placed 200 μm proximal to growth plate covering the distal femoral metaphysis. At a magnification of ×200, a line grid was superimposed onto fields of view that covered 100% of the ROIs. The line grid was used to count intersections between the fluorochrome labels, while the distance between two fluorochrome labels was determined using the ruler tool in newCAST. Mineralizing surfaces (MS/BS), mineral apposition rate (MAR), and bone formation rate (BFR/BS) were determined.

The proximal part of the tibiae were decalcified using formic acid, dehydrated, embedded in paraffin, and cut into 2-μm-thick sections using a microtome (RM 2165, Leica). The sections was immunohistochemically stained for cathepsin K (AP08851PU-N, Acris) as described earlier[24]. The sections were scanned in a digital slide scanner (NanoZoomer-XR, Hamamatsu) and analyzed using newCAST. A 1000-μm-high ROI, excluding cortical bone and primary spongiosa, was placed 200 μm distal to the growth plate. At a magnification of ×600, a line grid was superimposed onto fields of view that covered 100% of the ROI. The line grid was used to count intersections with osteoclasts and bone surfaces, and the percentage of osteoclast covered bone surfaces (Oc.S/BS) was calculated.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Six distal tibiae from each group were used for RT-qPCR. The distal tibiae were grinded in 1.5 ml microtubes with a micropestle (VWR), while kept frozen on dry ice in a special metal alloy microtube-rack (CoolRack M30 PF, BioCision) to minimize RNase activity. Lysis buffer from a PureLink RNA Mini kit (Ambion) was added and the pulverized bone was homogenized with rotor/stator homogenizer (VDI 12, VWR). RNA was isolated and purified with the PureLink RNA Mini kit. RNA quality was checked by running 1 μg total RNA on an agarose gel. 1 μg total RNA was transcribed into cDNA using qScript cDNA SuperMix (Quanta Biosciences). cDNA was diluted 1:5 with DEPC-treated water and 2 μl diluted cDNA was used in each reaction. PerfeCTa qPCR FastMix II (Quanta Biosciences) and TaqMan Gene Expression Assays (LifeTechnologies) was used with the following amplification protocol; hot start 95°C for 30 s followed by 40 cycles of amplification and quantification at 95°C for 3 s and 60°C for 30 s on a LightCycler 480 (Version 1.5, Roche). The studied genes were Bglap (osteocalcin) (Mm03413826_mH), Ctsk (cathepsin K) (Mm00484039_m1), Nfatc1 (nuclear factor of activated T-cells, cytoplasmic 1) (Mm00479445_m1), Rankl (receptor activator of nuclear κ-B ligand) (Mm00441906_m1), Opg (osteoprotegerin) (Mm01205928_m1), and Atp6v0d2 (V-type proton ATPase subunit d2) (Mm01222963_m1). Each sample was run in triplicates, and normalized to the reference genes B2m (beta-2 microglobulin) (Mm00437762_m1) and Gapdh (glyceraldehyde-3-phosphate dehydrogenase) (Mm99999915_g1).

Statistics

Data are given as mean ± SD. Differences between the groups were analyzed by a parametric one way ANOVA, whenever normal distribution requirements were met or otherwise by a non-parametric Kruskal-Wallis test. When significant variance was found, multiple comparisons between the groups were performed with the Student-Newman-Keuls test. Results were defined as statistically significant, if the two-tailed p<0.05. The statistical analyzes and graph drawing was performed using SigmaPlot (Version 13.0, Systat Software).

Results

The BTX-injected mice rapidly lost BW, which stayed below that of the Base and Ctrl groups until euthanasia (Figure 2). Furthermore, the BTX injections resulted in substantial atrophy of the rectus femoris muscles compared to the Base and Ctrl groups (Table 1). Neither BTX nor pantoprazole affected the length or height of the femora, humeri, and L5 vertebrae (Table 1).

Figure 2

Temporal development in body weight. a denotes a significant difference from Base and b donates a significant difference from Ctrl. Mean±SD.

Table 1

Weight of the rectus femoris muscles, length of the femora and humeri, and height of the corpus of the L5 vertebrae.

	Rectus femoris muscle mass (mg)	Femur length (mm)	L5 corpus height (mm)	Humerus length (mm)
Base	67.5< ± 5.2	15.32 ± 0.32	2.67 ± 0.16	11.80 ± 0.13
Ctrl	61.8 ± 6.5 [a]	15.48 ± 0.15	2.81 ± 0.14	11.90 ± 0.09
BTX	33.9 ± 5.1 [a,b]	15.37 ± 0.38	2.75 ± 0.13	11.84 ± 0.14
BTX+Pan	31.9 ± 4.7 [a,b]	15.54 ± 0.26	2.75 ± 0.11	11.87 ± 0.14

denotes a significant difference from Base and

denotes a significant difference from Ctrl. Mean ± SD.

Pantoprazole was not able to counteract the BTX-induced loss of femoral BMC and aBMD compared to the Base and Ctrl groups (Table 2). Unexpectedly, the mice injected with pantoprazole had a lower humeral aBMD than both the Base, Ctrl, and BTX groups (Table 2).

Table 2

DEXA of the whole femur and humerus.

	BMC (mg)	aBMD (mg/cm2)
Femur
Base	15.0 ± 1.2	44.3 ± 1.7
Ctrl	16.3 ± 0.9 [a]	47.3 ± 1.5 [a]
BTX	11.3 ± 0.7 [a,b]	35.7 ± 1.4 [a,b]
BTX+Pan	11.3 ± 0.7 [a,b]	35.1 ± 1.4 [a,b]
Humerus
Base	8.6 ± 1.1	43.1 ± 3.1
Ctrl	8.6 ± 1.2	40.8 ± 3.3
BTX	8.5 ± 1.1	42.0 ± 3.1
BTX+Pan	7.8 ± 1.1	37.9 ± 2.5 [a,b,c]

Bone mineral content (BMC). Areal bone mineral density (aBMD).

denotes a significant difference from Base,

denotes a significant difference from Ctrl, and

denotes a significant difference from BTX. Mean ± SD.

The BTX-injections resulted in a lower cortical bone area at the femoral mid-diaphysis compared to both the Base and Ctrl groups, which was caused by endosteal resorption as the marrow area of the BTX-injected groups was larger than that of the Ctrl group (Figure 3 and Table 3). The lower bone area was followed by a lower polar moment of inertia in the BTX-injected groups compared to both the Base and Ctrl groups (Table 3). Moreover, the marrow area at the humeral mid-diaphysis expanded due to the BTX-injections when compared to both the Base and Ctrl groups (Figure 3 and Table 3). Pantoprazole did not prevent the loss of cortical bone.

Figure 3

Transverse sections of the femoral and humeral mid-diaphysis acquired from the μCT scans. The samples shown are those closest to the mean values of the respective groups.

Table 3

μCT of the femoral and humeral mid-diaphyses.

	Bone area (mm2)	Marrow area (mm2)	Tissue area (mm2)	Polar moment of inertia (mm2)	Bone material density (mg/cm2)
Femoral mid-diaphysis
Base	0.74 ± 0.03	1.02 ± 0.05	1.76 ± 0.08	0.35 ± 0.03	1126 ± 10
Ctrl	0.78 ± 0.04 [a]	1.00 ± 0.07	1.78 ± 0.11	0.37 ± 0.04	1141 ± 5 [a]
BTX	0.64 ± 0.03 [a,b]	1.07 ± 0.06 [b]	1.72 ± 0.05	0.30 ± 0.02 [a,b]	1130 ± 9 [b]
BTX+Pan	0.64 ± 0.04 [a,b]	1.07 ± 0.06 [b]	1.71 ± 0.08	0.30 ± 0.03 [a,b]	1130 ± 7 [b]
Humeral mid-diaphysis
Base	0.50 ± 0.02	0.41 ± 0.02	0.91 ± 0.04	0.10 ± 0.01	1141 ± 14
Ctrl	0.51 ± 0.02	0.42 ± 0.02	0.93 ± 0.04	0.11 ± 0.01	1152 ± 10
BTX	0.49 ± 0.03	0.44 ± 0.02 [a,b]	0.93 ± 0.04	0.11 ± 0.01	1150 ± 10
BTX+Pan	0.48 ± 0.02	0.44 ± 0.01 [a,b]	0.93 ± 0.02	0.10 ± 0.01	1152 ± 11

denotes a significant difference from Base and

denotes a significant difference from Ctrl. Mean ± SD.

At both the distal femoral metaphysis and epiphysis, the BTX-injections resulted in a bone loss indicated by lower BV/TV, Tb.Th, Tb.N, and bone material density and higher Tb.Sp compared to the Base and Ctrl groups (Figure 4, Figure 5, and Table 4). At the distal femoral epiphysis, the BTX-induced bone loss was accompanied by a higher SMI, indicating a shift towards more rod-like trabeculae, compared to the Base and Ctrl groups (Figure 4 and Table 4). At the L5 vertebral body, BTX resulted in a lower BV/TV and Tb.Th, and a slightly higher SMI, compared to the Ctrl group (Figure 4, Figure 5, and Table 4). At the proximal humeral metaphysis and epiphysis, BTX had a minor deleterious effect on the trabecular bone compared to the Base or Ctrl group (Figure 4, Figure 5, and Table 4). Only at the vertebra L5 did pantoprazole seem to have a preventable effect on the loss of BV/TV and Tb.Th (Figure 5).

Figure 4

3D images reconstructed from the μCT scans. The size of the different bone sites are not proportional to each other. The samples shown are those closest to the mean values of the respective groups.

Figure 5

Trabecular bone volume fraction (BV/TV) and trabecular thickness (Tb.Th) of the analyzed VOI from the μCT scans that contains trabecular bone. a denotes a significant difference from Base, b donates a significant difference from Ctrl, and c denotes a significant difference from BTX. Mean ± SD.

Table 4

µCT of the distal femoral metaphysis and epiphysis, L5 vertebra, and proximal humeral metaphysis and epiphysis.

	BV/TV (%)	Tb.Th (µm)	Tb.N (1/mm)	Tb.Sp (µm)	CD (1/mm3)	SMI (-)	Bone material density (mg/cm3)
Distal femoral metaphysis
Base	5.7 ± 1.1	33.3 ± 4.7	3.6 ± 0.2	276 ± 13	208 ± 38	2.3 ± 0.2	866 ± 15
Ctrl	6.0 ± 1.1	37.5 ± 3.3 [a]	3.4 ± 0.2 [a]	294 ± 14 [a]	157 ± 31 [a]	2.2 ± 0.2	882 ± 14 [a]
BTX	3.1 ± 0.5 [a,b]	24.2 ± 2.3 [a,b]	3.2 ± 0.1 [a,b]	313 ± 13 [a,b]	134 ± 24 [a]	2.3 ± 0.1	833 ± 11 [a,b]
BTX+Pan	3.0 ± 0.6 [a,b]	23.3 ± 2.1 [a,b]	3.2 ± 0.1 [a,b]	310 ± 15 [a,b]	145 ± 26 [a]	2.3 ± 0.1	825 ± 19 [a,b]
Distal femoral epiphysis
Base	21.6 ± 1.7	50.0 ± 1.8	5.2 ± 0.4	212 ± 14	293 ± 48	0.3 ± 0.1	949 ± 8
Ctrl	24.3 ± 1.4 [a]	59.3 ± 4.0 [a]	5.0 ± 0.4	223 ± 17	212 ± 62 [a]	0.1 ± 0.1 [a]	968 ± 14 [a]
BTX	10.6 ± 1.4 [a,b]	33.3 ± 2.7 [a,b]	4.2 ± 0.3 [a,b]	245 ± 19 [a,b]	300 ± 50 [b]	0.9 ± 0.1 [a,b]	903 ± 14 [a,b]
BTX+Pan	10.2 ± 2.5 [a,b]	32.5 ± 4.8 [a,b]	4.0 ± 0.2 [a,b]	257 ± 17 [a,b]	321 ± 61 [b]	1.0 ± 0.2 [a,b]	893 ± 16 [a,b]
L5 Vertebra
Base	13.9 ± 1.3	37.8 ± 2.3	3.8 ± 0.2	263 ± 17	267 ± 50	1.0 ± 0.1	924 ± 9
Ctrl	16.4 ± 1.4 [a]	43.6 ± 2.7 [a]	3.7 ± 0.2	265 ± 20	258 ± 50	0.9 ± 0.1 [a]	933 ± 8 [a]
BTX	12.7 ± 1.5 [b]	36.9 ± 2.4 [b]	3.5 ± 0.3	284 ± 25	253 ± 38	1.1 ± 0.2 [b]	928 ± 7
BTX+Pan	14.1 ± 1.4 [b]	39.8 ± 2.1 [b,c]	3.6 ± 0.3	279 ± 28	272 ± 51	1.1 ± 0.1 [b]	928 ± 5
Proximal humeral metaphysis
Base	4.9 ± 0.8	35.8 ± 2.4	3.4 ± 0.2	292 ± 20	102 ± 26	2.4 ± 0.2	872 ± 9
Ctrl	4.5 ± 0.7	33.9 ± 2.9	3.2 ± 0.2 [a]	317 ± 19 [a]	83 ± 23	2.4 ± 0.1	872 ± 13
BTX	3.8 ± 0.9 [a]	32.6 ± 2.6	3.1 ± 0.2 [a]	323 ± 21 [a]	77 ± 20 [a]	2.5 ± 0.2	858 ± 17
BTX+Pan	4.2 ± 0.9	33.5 ± 5.9	3.1 ± 0.2 [a]	326 ± 20 [a]	84 ± 18	2.4 ± 0.2	862 ± 21
Proximal humeral epiphysis
Base	24.9 ± 1.6	43.0 ± 1.0	6.3 ± 0.4	161 ± 11	438 ± 37	0.1 ± 0.1	919 ± 7
Ctrl	23.8 ± 2.3	44.2 ± 2.5	5.9 ± 0.5	173 ± 14 [a]	364 ± 47 [a]	0.0 ± 0.2	923 ± 14
BTX	22.7 ± 1.2 [a]	41.5 ± 1.7 [b]	5.8 ± 0.4 [a]	175 ± 11 [a]	410 ± 63	0.1 ± 0.1	915 ± 14
BTX+Pan	22.1 ± 2.8 [a]	41.3 ± 1.3 [b]	5.7 ± 0.5 [a]	179 ± 16 [a]	375 ± 61 [a]	0.1 ± 0.2	913 ± 11

Trabecular bone volume fraction (BV/TV). Trabecular thickness (Tb.Th). Trabecular number (Tb.N). Trabecular spacing (Tb.Sp). Connectivity density (CD). Structure model index (SMI).

denotes a significant difference from Base,

denotes a significant difference from Ctrl, and

denotes a significant difference from BTX. Mean±SD.

The BTX-injections resulted in a lower fracture strength and maximum stiffness at both the femoral mid-diaphysis and neck compared to the Base and Ctrl groups (Figure 6). At the L5 vertebra, the BTX group had lower fracture strength than both the Base and Ctrl groups, whereas the BTX+Pan group had a fracture strength that was lower than the Ctrl group only (Figure 6). At the humeral mid-diaphysis, the BTX-injected mice had lower fracture strength than the Ctrl group (Figure 6).

Figure 6

Fracture strength and maximum stiffness obtained by mechanical testing. a denotes a significant difference from Base and b donates a significant difference from Ctrl. Mean ± SD.

During day 13-17 of the study, BTX resulted in higher BFR/BS at both the distal femoral metaphysis and epiphysis compared to the Ctrl group (Table 5). At the distal femoral metaphysis, the higher BFR/BS was driven by a higher MAR, whereas at the distal femoral epiphysis the higher BFR/BS was a combination of both a higher MS/BS and MAR (Table 5). In general, during day 9-13 of the study, there was no or only minor indications of higher osteoblastic activity in the BTX-injected mice compared to the Ctrl group (Table 5). Treatment with pantoprazole did not have any effect on either MS/BS, MAR, or BFR/BS.

Table 5

Dynamic bone histomorphometry of the distal femoral metaphysis and epiphysis.

	Day 9-13			Day 13-17
	MS/BS (%)	MAR (µm/d)	BFR/BS (µm3/µm2/d)	MS/BS (%)	MAR (µm/d)	BFR/BS (µm3/µm2/d)
Distal femoral metaphysis
Base	-	-	-	-	-	-
Ctrl	37 ± 6	1.8 ± 0.2	0.7 ± 0.1	35 ± 7	2.0 ± 0.3	0.7 ± 0.2
BTX	28 ± 6 [b]	2.1 ± 0.7	0.6 ± 0.2	38 ± 8	2.5 ± 0.4 [b]	1.0 ± 0.3 [b]
BTX+Pan	29 ± 7 [b]	2.2 ± 0.5	0.6 ± 0.2	37 ± 3	2.6 ± 0.4 [b]	1.0 ± 0.2 [b]
Distal femoral epiphysis
Base	-	-	-	-	-	-
Ctrl	56 ± 7	1.1 ± 0.1	0.6 ± 0.1	45 ± 12	1.2 ± 0.2	0.5 ± 0.2
BTX	51 ± 13	1.4 ± 0.2 [b]	0.7 ± 0.2	55 ± 7 [b]	1.6 ± 0.2 [b]	0.9 ± 0.1 [b]
BTX+Pan	51 ± 9	1.6 ± 0.2 [b]	0.8 ± 0.2 [b]	53 ± 7 [b]	1.7 ± 0.2 [b]	0.9 ± 0.1 [b]

Mineralizing surfaces (MS/BS). Mineral apposition rate (MAR). Bone formation rate (BFR/BS).

denotes a significant difference from Ctrl. Mean ± SD.

At euthanasia, neither BTX nor treatment with pantoprazole affected the Oc.S/BS (Figure 7).

Figure 7

Osteoclast covered bone surfaces (Oc.S/BS) at the proximal tibial metaphysis. Mean ± SD.

At euthanasia, Bglap, which is the mRNA template that osteocalcin is translated from, was higher expressed in the distal tibia of the BTX-injected hind limb, indicating higher osteoblastic activity, compared to the Ctrl group (Figure 8). Ctsk and Atp6v0d2, which are translated to cathepsin K and the V-type proton ATPase subunit d2, respectively, were both higher expressed in the BTX-injected mice, indicating higher osteoclastic activity, compared to both the Base and Ctrl groups (Figure 8). The ratio of Rankl/Opg, which is an indicator of osteoclastic differentiation from the hematopoietic stem cell lineage, was higher in the BTX-injected groups than in the Ctrl group (Figure 8). Nfatc1, which is an indicator of osteoclastic differentiation, was not affected by BTX (Data not shown). Pantoprazole did not have any effect on osteoblastic or osteoclastic related gene expression.

Figure 8

Relative gene expression obtained from the distal tibia. a denotes a significant difference from Base and b donates a significant difference from Ctrl. Mean ± SD.

Discussion

The present study did not demonstrate any bone specific effects of treatment with pantoprazole on either BTX induced disuse osteopenia or at the remaining intact skeleton in skeletally mature C57BL/6J female mice.

Pantoprazole does not prevent disuse osteopenia

The localized injections of BTX induced a profound and significant bone loss in the affected hind limb. In pre-clinical studies, pantoprazole has been shown to affect both osteoblasts and osteoclasts, which led to the hypothesis that pantoprazole might have a bone protective effect[9,25]. However, microstructural, mechanical, histological, and molecular analyses did not reveal any bone specific effects, although pantoprazole was given in a dose that is 100-200 times higher than therapeutic human doses.

PPIs irreversibly inhibit the gastric H+/K+-ATPase by covalently binding cysteine residues with disulfide bonds and thereby reducing the capacity for proton pumping[5]. Accordingly, PPIs might potentially inhibit the V-ATPase and thereby disrupt bone resorption. Furthermore, the different PPIs react with different cysteine residues, and this might be important for a possible osteoclast inhibiting effect[5]. However, in vitro experiments have shown an osteoclast inhibitory effect of both omeprazole[6-8,26], esomeprazole[8], lansoprazole[8], and pantoprazole[9].

In the present study, we chose to treat disuse osteopenia with pantoprazole based upon the study by Histing et al., in which daily pantoprazole injections impaired bone healing in a murine mid-femoral fracture model[14]. Most in vivo studies have been performed on rats, in which treatment with PPIs in young animals resulted in lower BMC[27-29], whereas PPIs seemed to have either no or a bone protective effect in intact and gastrectomized older animals[11,30]. Moreover, PPIs have been associated with reduced expression of osteoclast specific mRNA, osteoclast numbers, and resorption of calcium cements[13,15,31]. However, the present study was not able to detect any of these osteoclast specific effects of PPI treatment on microstructural properties, gene expression, or Oc.S/BS in either the intact or the disuse osteopenic mouse skeleton.

Although the hypothesis of the present study was that the PPI pantoprazole would have an osteoprotective effect, current literature suggests an increased incidence of osteoporotic fractures following both short and long-term PPI treatment in a clinical setting[32,33]. The biological explanation has been proposed to be a decrease in calcium and magnesium uptake due to higher gastric pH induced by PPI treatment[34,35], increase in the incidence of falls leading to fragility fractures[36], and a direct effect on bone cells[37]. However, the studies are conflicting and a persuasive biological explanation remains to be found[32,33,38,39]. Furthermore, most of the clinical studies are observational and confounding due to e.g. polypharmacy and multi-morbidity might explain the increased fracture risk in PPI users rather than the drug itself.

The systemic skeletal effects of localized botulinum toxin injections are widespread

In the present study, we investigated both the disuse osteopenic hind limb, and the L5 vertebral body and humerus located distant from the BTX-injected limb, in order to elucidate any bone specific effects of the pantoprazole treatment. Although pantoprazole did not have any bone specific effects, we found that the BTX induced disuse affected the supposedly intact skeleton at bone sites distant from the paralyzed muscles. Both the L5 vertebral body and humerus had deteriorated microstructural and mechanical properties because of the BTX injections.

Systemic effects of the localized intramuscular injections of BTX have been reported several times and seems to be a consistent finding[19,21,22,40-44]. The effects range from small differences in BW between BTX injected and control mice to substantial muscle and bone wasting of the non-injected contralateral hind limb compared to control mice. The present study is the first to report widespread skeletal systemic effects in the spine and humerus. However, the bone loss in the immobilized hind limb by far outweighs any systemic effects.

Tail suspension (TS) and sciatic neurectomy (SN) are also widely used to study disuse osteopenia in mice. TS remove the ground reaction force on the hind limbs, but not the muscle forces, by elevating the rearmost part of the mouse body. SN paralyzes not only the muscles innervated by the sciatic nerve, but also disrupts the autonomic and sensory nervous system associated. A systemic skeletal effect of unloading is also seen with TS and SN induced disuse osteopenia. TS induces lower bone mass and bone strength of the vertebra and humerus compared to control mice[45,46], whereas only the vertebral bone loss has been investigated in SN[47]. It is not surprising that hind limb immobilization causes lower bone mass and bone strength in the lumbar vertebrae with BTX, TS, and SN induced disuse osteopenia, as loads and muscle forces exerted on the spine are altered. However, it is surprising that the humerus, which must bear more loads, especially with TS, is negatively affected by disuse. A possible explanation for this might be systemically uptake of botulinum toxin, reduced food intake, decreased ambulation, increased levels of glucocorticoids and other catabolic stress hormones, and “spill-over” of bone acting molecules from the immobilized hind limb. Recently, the muscle-bone interaction has attracted much attention and molecules secreted by skeletal muscles, i.e. myokines, are recognized as important factors in regulating bone homeostasis and turnover[48]. In the present study, BTX induced a massive muscle loss and myokines secreted during atrophy might influence bone health at distant skeletal sites and explain the widespread systemic effect.

Limitations and future perspectives

The present study did not find any bone specific effects of pantoprazole, which might be related to the short study duration or that pantoprazole does not have a direct inhibiting effect on the osteoclastic V-ATPase. The duration of the present study was chosen based on earlier investigations of BTX-induced osteopenia, which showed that the BTX-induced bone loss levels out after three weeks[17,40]. Long-term treatment in intact, fully-grown mice might elucidate bone specific effects and help explain the clinically observed PPI treatment related increased fracture risk. Furthermore, the substantial and rapid bone loss induced by BTX might overshadow any osteoclast inhibiting effects of pantoprazole. However, as no substantial effect was observed at the skeletal sites distant from the immobilized hind limb, it seems unlikely that pantoprazole directly influences bone cell function in mice. We interpret the minor, but significant, bone-protecting effect of pantoprazole seen on the trabecular thickness of the L5 vertebral body as a statistical type-1 error.

The widespread systemic effects on the non-immobilized skeleton are interesting and the biological mechanism remains to be found. The explanation might just be reduced food intake and reduced BW, but hormonal factors, e.g. myokines, cannot be ruled out and insights might provide new drug targets.

Conclusion

The PPI pantoprazole was unable to prevent the development of BTX induced disuse osteopenia in skeletally mature female C57BL/6J mice.