−) mice

Bone 44 (2009) 199–207

Contents lists available at ScienceDirect

Bone j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b o n e

Bone density, strength, and formation in adult cathepsin K (−/−) mice B. Pennypacker a,⁎, M. Shea b, Q. Liu b, P. Masarachia a, P. Saftig c, S. Rodan d, G. Rodan †, D. Kimmel a a

Department of Molecular Endocrinology and Bone Biology, Merck Research Laboratories, WP26A-1000 West Point, PA 19486, USA Oregon Health Sciences University, Portland, OR, USA c Christian-Albrechts Universität, Kiel, Germany d Department of Biochemistry, School of Dental Medicine, University of Pennsylvania, Philadelphia, PA, USA b

a r t i c l e

i n f o

Article history: Received 24 January 2008 Revised 27 August 2008 Accepted 29 August 2008 Available online 19 September 2008 Edited by: D. Burr Keywords: Bone quality Density Histomorphometry Strength Formation

a b s t r a c t Cathepsin K (CatK) is a cysteine protease expressed predominantly in osteoclasts, that plays a prominent role in degrading Type I collagen. Growing CatK null mice have osteopetrosis associated with a reduced ability to degrade bone matrix. Bone strength and histomorphometric endpoints in young adult CatK null mice aged more than 10 weeks have not been studied. The purpose of this paper is to describe bone mass, strength, resorption, and formation in young adult CatK null mice. In male and female wild-type (WT), heterozygous, and homozygous CatK null mice (total N = 50) aged 19 weeks, in-life double fluorochrome labeling was performed. Right femurs and lumbar vertebral bodies 1– 3 (LV) were evaluated by dual-energy X-ray absorptiometry (DXA) for bone mineral content (BMC) and bone mineral density (BMD). The trabecular region of the femur and the cortical region of the tibia were evaluated by histomorphometry. The left femur and sixth lumbar vertebral body were tested biomechanically. CatK (−/−) mice show higher BMD at the central and distal femur. Central femur ultimate load was positively influenced by genotype, and was positively correlated with both cortical area and BMC. Lumbar vertebral body ultimate load was also positively correlated to BMC. Genotype did not influence the relationship of ultimate load to BMC in either the central femur or vertebral body. CatK (−/−) mice had less lamellar cortical bone than WT mice. Higher bone volume, trabecular thickness, and trabecular number were observed at the distal femur in CatK (−/−) mice. Smaller marrow cavities were also present at the central femur of CatK (−/−) mice. CatK (−/−) mice exhibited greater trabecular mineralizing surface, associated with normal volume-based formation of trabecular bone. Adult CatK (−/−) mice have higher bone mass in both cortical and cancellous regions than WT mice. Though no direct measures of bone resorption rate were made, the higher cortical bone quantity is associated with a smaller marrow cavity and increased retention of non-lamellar bone, signs of decreased endocortical resorption. The relationship of bone strength to BMC does not differ with genotype, indicating the presence of bone tissue of normal quality in the absence of CatK. © 2008 Elsevier Inc. All rights reserved.

Introduction Cathepsins are lysosomal proteases, most of which belong to the papain-like cysteine protease family, a group of proteolytic enzymes involved in many physiologic and pathologic processes. Eleven distinct human cathepsin sequences (B, C, F, H, K, L, O, S, V, X, and W) exist [1]. Cathepsin K (CatK), the only lysosomal cysteine protease that degrades Type I collagen at neutral and acidic pH [2,3], is predominantly expressed within osteoclasts [4]. Non-specific cysteine protease inhibitors, including leupeptin, E-64, and Z-Phe-Phe-CHN2, inhibit bone resorption in vitro and in vivo [5–10]. The substrate

⁎ Corresponding author. Fax: +1 215 652 4328. E-mail address: [email protected] (B. Pennypacker). † Deceased. 8756-3282/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2008.08.130

preference and cellular distribution of CatK suggest that it plays an important role in the collagen degradation phase of bone resorption. Genetic confirmation for a role of CatK in bone resorption exists in humans in the form of pycnodysostosis, a very rare disease characterized by high bone mineral density (BMD), acroosteolysis of the distal phalanges, short stature, and skull deformities with delayed suture closure. Pycnodysostosis is linked to several loss-of-function mutations in the CatK gene [11–13]. Though osteoclasts have normal ruffled borders and clear zones, their cytoplasmic vacuoles contain collagen fibrils, and the underlying region of demineralized bone is increased [14]. Accumulation of undigested collagen fibrils has also been seen in various connective tissues (periosteum, perichondrium, tendon and synovial membrane), where fibroblasts may express impaired phagocytosis [15]. Thus, the bone phenotype of pycnodysostotic humans suggests that a lack of CatK is associated with functional deficits in bone resorption.

200

B. Pennypacker et al. / Bone 44 (2009) 199–207

Targeted mutation of the CatK gene in mice produces many phenotypic features of osteopetrosis. Long bones and lumbar vertebrae in growing (8–10-week-old) CatK (−/−) mice appear osteopetrotic [16,17]. Trabecular bone volume, thickness, and number are greater in CatK (−/−) than in wild-type (WT) mice [18]. Though 1week-old CatK (−/−) mice have normal intramembranous bone, the marrow cavities of endochondral bones contain both excessive calcified cartilage and trabecular bone at the mid-diaphyses of long bones [17]. Furthermore, large, irregular vacuoles are found in osteoclasts and undigested collagen fibrils are present both intracellularly and extracellularly within the resorption lacunae [17]. As in pycnodysostotic humans, the excessive amount of bone and calcified cartilage tissue and the abnormal appearance of the osteoclasts point to decreased osteoclast activity. One previous study reports high levels of bone formation activity in trabecular regions [19]. Thus, the recorded bone phenotype of CatK (−/−) mice not only suggests that a lack of CatK is associated with functional deficits in bone resorption, but also points to a need to confirm the existence of an unexpectedly high bone formation rate. The bone traits observed during the lifelong absence of CatK in humans and mice suggests that intact CatK is required for bone resorption processes to occur at a normal rate [11,16,17,20]. When bone resorption is slowed throughout the lifespan of mammals, the reduced rate of bone modeling during growth yields an abnormal external shape. Such shape alterations that develop slowly over decades of bone modeling in humans, may offset bone strength increases related to higher BMD [19], creating potential bone quality issues. Pycnodysostotic humans do not exist in sufficient numbers to allow a conclusive study of bone quality. Only one publication on CatK (−/−) mice addresses bone strength in adult CatK null mice [21]. Additional data on adult CatK (−/−) mice can assist in assessing the degree to which altered bone modeling affects bone quality during from-birth loss of CatK function. The purpose of this study is to assess bone quality and formation in young adult CatK (−/−) and WT mice using densitometric, histomorphometric, polarized light, and biomechanical techniques to evaluate trabecular and cortical regions of the femur, tibia, and vertebrae. Materials and methods Animals All procedures were conducted in accordance with National Institutes of Health guidelines and approved by the Merck West Point Institutional Animal Care and Use Committee (IACUC). The method for generating CatK (−/−) mice and determining genotype has been described [16]. Breeding trios of CatK (−/−) mice were maintained at Charles River Laboratories (Wilmington, MA). Fifty mice (22 wild-type (8M, 14F)), 16 heterozygotes (7M, 9F)), and 12 homozygotes (6M, 6F)) from the F2 and later generations were obtained. A fluorochrome label (Calcein; C0875, Sigma; 10 mg/kg subcutaneous) was given on the tenth and third days before necropsy. The mice were sacrificed via CO2 asphyxiation at age 19 weeks. Body weight was recorded. At necropsy, the right femora, right tibiae, and lumbar vertebrae (L1–3) were removed and fixed in 70% ethanol. The left femora and lumbar vertebrae (L4–6) were defleshed, wrapped in saline-soaked gauze, frozen (−20 °C) and shipped to Oregon Health Sciences University (OHSU) for mechanical testing. Densitometry (DXA) Defleshed whole right femora were placed on an acrylic base with posterior side down and scanned using pDXA (Norland; Ft. Atkinson, WI) with a resolution of 0.2 × 0.2 mm, field width of 1.4 cm, and speed of 2 mm/s. Regions of interest (ROI) located at 0–3 mm (distal) and 4– 7 mm (central) from the distal end of the femur were isolated using

manufacturer's software. Bone mineral content (BMC; mg) and bone area (B.Ar; cm2) were determined for each ROI. Bone mineral density (BMD; mg/cm2) was BMC / B.Ar. The defleshed L1–L3 segment was scanned using PIXImus (GELunar Corp.; Madison, WI). The whole segment was isolated using the manufacturer's software. Lumbar vertebral bone mineral content (LVBMC), bone area (LVB.Ar), and bone mineral density (LVBMD) were derived as above. Cancellous bone histomorphometry The distal right femur specimens were dehydrated, and embedded undecalcified in 80% methyl methacrylate (Polysciences, Warrington, PA, USA)/20% dibutyl phthalate). Frontal sections (5 μm) through the central portion of the distal metaphysis were prepared with a Leica Polycut sledge microtome and mounted on glass slides. After deplasticizing with 2-methoxyethylacetate, one set of sections was stained with Masson's trichrome for evaluation of osteoclasts. A second set of sections was left unstained and used for analysis of bone mass and fluorescent labeling of bone formation-related parameters. Coverslips were affixed using Eukitt's mounting media. All slides were given a blind code prior to data collection by one observer. The standard nomenclature for bone histomorphometry was used [22]. A ∼3.5 mm2 ROI located 1.00–2.75 mm proximal to the distal femoral growth cartilage metaphyseal junction (GCMJ) was outlined on the unstained section. The ROI, terminating laterally at each cortex, included only trabecular bone and marrow of the secondary spongiosa. A light/epifluorescent microscope (Optiphot II; Nikon, Japan) equipped with an Optronics DEI-750 CE (Tuttlingen, Germany) video camera interfaced to a computer was used. BIOQUANT Nova Software (Bioquant Image Analysis Corp.; Nashville, TN), was used to collect the raw data. Total tissue area (Tt.Ar), trabecular bone area (B.Ar) and bone perimeter (B.Pm) were measured. Double and single-labeled perimeters (dL. Pm, sL.Pm) were identified. Interlabel width of double-labeled regions (dL.Wi) was measured. From Masson trichrome sections using a similar ROI, osteoclast perimeter (Oc.Pm), osteoclast number (Oc.N), and total bone perimeter (B.Pm) were recorded to assess osteoclast surface (Oc.Pm/B.Pm; %)) and osteoclast number per bone surface (Oc.N/B.Pm; mm− 1). From Tt.Ar and B.Ar, trabecular bone volume (BV/TV; %), trabecular thickness (Tb.Th; μm), and trabecular number (Tb.N; #/mm) were calculated. From B.Pm, dL.Pm, and sL.Pm, both double-labeled surface (dLS.Pm/B.Pm; %) and mineralizing surface (MS/BS) (dLS.Pm + [0.5 ⁎ sL. Pm])/B.Pm; %) were calculated. From dL.Wi and interlabel time period, mineral apposition rate (MAR; μm/d) was calculated. Surface-based (BFR/BS; mm2/mm/yr), bone volume-based (BFR/BV; mm2/mm2/yr), total tissue volume-based bone formation rates (BFR/TV; mm2/mm2/ yr) were calculated. Cortical bone histomorphometry (right central tibia) The right tibia was cross-sectioned at its midpoint with an Exakt Saw 300 CP band system (Exakt Technologies, Inc.; Oklahoma City, OK). The distal end of the proximal piece was marked with indelible ink, then dehydrated and embedded undecalcified in 90% methyl methacrylate/10% dibutyl phthalate. 100 μm sections were prepared using a Leica SP1600 Saw Microtome (Leica Instruments GmbH; Nussloch, Germany). Sections were mounted on glass slides, then coverslipped using Eukitt's mounting media. The imaging system above was used to collect data from the periosteal and endocortical surfaces. Using brightfield (Nikon Optiphot 2), total tibial cortical area (tTi.Co.Ar) and lamellar bone area (Ti.Lm.Co.Ar) were measured. Lamellar bone was identified equally well by polarized and brightfield light (data not shown). Total surface perimeters were measured (PsB. Pm, EcB.Pm). Double and single-labeled perimeters of both surfaces

B. Pennypacker et al. / Bone 44 (2009) 199–207

(PsdL.Pm and PssL.Pm, EcdL.Pm and EcsL.Pm) were also recorded. Interlabel widths of double-labeled regions (PsdL.Wi, EcdL.Wi) were measured. Lamellar cortical area (%) was calculated (100XTi.LmCo.Ar/ Ti.Co.Ar). Mineralizing surface, MAR, and surface-based bone formation rate were calculated for both surfaces as above. Cortical microarchitecture (right central femur) The right femur was cross-sectioned with a low speed diamond saw (Isomet, Buehler) at 5 mm and 9 mm proximal to its distal end. This bone segment approximated the central femur region of interest for BMD evaluation. The distal end of the segment was marked with permanent ink and sent to Oregon Health Sciences University (OHSU). At OHSU, the proximal end of the segment was digitally imaged and analyzed using a digital analysis program (Optimas, Inc.; Seattle, WA). Total (TF.Ar), cortical (FCo.Ar), and marrow cavity (FMe.Ar) areas, and cross-sectional moment of inertia (CSMI) were determined. From FCo. Ar and FMe.Ar, cortical thickness (FCt.Th) was calculated, assuming a circular shape. Bone strength Femoral diaphysis The midpoint of the left femur was tested to failure in three-point bending using an electrohydraulic material test system (Instron, Model 4442; Canton, MA). Load and displacement data were collected using system software (Series IX for Windows 95; Canton, MA). The femur was placed with its posterior surface resting on two lower supports located ∼7.2 mm apart, with its midpoint centered between the lower supports. A digital controller was used to displace the actuator from above, first making contact with the femur at its midpoint, and then being progressively lowered at a rate of 0.08 mm/s until failure occurred. After testing, the fracture line was examined to ensure the fracture occurred perpendicular to the longitudinal axis of the bone. Ultimate load (FF.U, N) and stiffness (FS, N/mm) were determined for each femur by software that analyzed the load-displacement curve. The 2% offset method was used to calculate yield load (FF.Y, N). Post-yield deflection (PYD, mm) was calculated as the displacement from the yield point to the breaking point. Failure stress and modulus were calculated using the geometric values [23]. Femoral neck The proximal left femur was embedded in polymethyl methacrylate to the level of the lesser trochanter. A load was applied to the femoral head at a rate of 0.08 mm/s using the Instron 4442, to simulate single-legged stance, until failure occurred. Load and displacement data were recorded; ultimate load (FNF.U, N) and stiffness (FN.S, N/mm) were determined as above. The fracture line was examined to ensure that the fracture occurred perpendicular to the longitudinal axis of the femoral neck. Vertebral body The sixth lumbar vertebral body was dissected free from L5 and the posterior elements were removed. The cartilaginous endplates were removed in a fashion that left parallel planes at the cranial and caudal ends, without removing excess bone. The vertebral body was placed between parallel platens, one stable below and one movable above. A digital controller was used to displace the platen from above, at a rate of 0.08 mm/s until failure occurred. After testing, the fracture line was examined to ensure the fracture occurred perpendicular to the longitudinal axis of the bone. This occurred in three bones whose data were excluded. Load and displacement data were recorded and ultimate load (LVF.U, N) and stiffness (LV.S, N/mm) were determined as above. Post-yield deflection could not be calculated reliably from these load-displacement curves.

201

Statistical methods Each endpoint was analyzed by analysis of variance (ANOVA) [24] using predefined contrasts, namely an overall test for difference between males and females, a test for genotype effect (assuming any difference between (+/−) and WT is intermediate between (−/−) and WT), and a test for whether males and females differ in genotype effect. F-tests used type I sums of squares [24], and the mean square error used only within group differences. A two degree of freedom interaction F-test was also done to assess whether the differences among the six groups were more complex than assumed by the predefined contrasts [25]. When ANOVA within gender was significant for genotype differences, Fisher's Protected Least Significant Differences (PLSD) test was applied to find differences among genotypes. The relationship of FF.U (left femur) to central femur BMC (right), and FCo.Ar, and CSMI, was tested by linear regression and multiple regression. The relationship of LVF.U (LV6) to lumbar vertebral BMC (LV1–3) was tested by linear regression. The following groups were tested in linear regression for femur and vertebral F.U to femur and vertebral BMC: all mice, all males, all females, all wild-types, all heterozygotes, and all (−/−'s). The slopes for genotype and gender groups were compared by generalized linear regression. The assumptions were checked and the following model was fit: F:UfBMC + gender + genotype + ðgender4BMCÞ + ðgenotype4BMCÞ + ðgender4genotype4BMCÞ The hypothesis of no gender difference in slopes was tested by the significance of the term “gender ⁎ BMC”, and the hypothesis of no genotype difference in slopes was tested by the significance of “genotype ⁎ BMC.” Results There were no significant differences in body weight according to either genotype or gender (data not shown). Bone mineral density (BMD) Relative to WT mice, female, but not male, Cat K (−/−) mice had significantly higher distal femur BMD than WT mice (Fig. 1A). Both genders of CatK (−/−) mice had higher central femur BMD than WT mice (Fig. 1B). Heterozygous (+/−) mice of both genders had intermediate BMD at both sites. All trends were similar for BMC (data not shown by groups). In contrast, there were no significant genotype or gender differences in vertebral body BMD (Table 1). Trabecular histomorphometry There was a strong genotype and gender influence on bone volume and trabecular microarchitecture (Table 2). Male and female CatK (−/−) mice had BV/TV nearly two-fold higher in male (−/−), and six-fold higher in female (−/−), than WT mice. BV/TV and Tb.Th of the distal femur were significantly higher in both male and female CatK (−/−) mice than in WT mice. Though Tb.Th was higher in CatK (−/−) mice of both genders, Tb.N was significantly higher only in female CatK (−/−) mice. For each endpoint, female CatK (−/−) mice displayed a more prominent genotype-related difference than did male CatK (−/−) mice. Both MS/BS and MAR were significantly higher in CatK (−/−) mice than in WT mice (Table 3). Surface-based bone formation rate (BFR/ BS) differed significantly with genotype with CatK (−/−) mice being ∼2.5-fold higher than WT mice regardless of gender. There was also a significant gender difference with females showing higher BFR/BS than males (Figs. 2 and 3). Total tissue volume-based bone formation rate (BFR/TV) was significantly higher in CatK (−/−)than in WT mice. However, bone volume-based bone formation rate (BFR/BV) showed

202

B. Pennypacker et al. / Bone 44 (2009) 199–207 Table 2 Trabecular bone volume and microarchitecture (distal femur) Sex

Genotype

Bone volume (BV/TV, %)

Trabecular thickness (Tb.Th, μm)

Trabecular number (Tb.N, #/mm)

M M M F F F

Wild-type +/− (−/−) Wild-type +/− (−/−) Genotype effect Gender effect Interaction

12.4 ± 1.6 16.6 ± 3.0 21.1 ± 1.8⁎ 5.2 ± 0.7 14.2 ± 3.5⁎ 29.4 ± 4.6⁎ b 0.0001 0.3262 0.0143

38.1 ± 3.2 44.8 ± 4.7 53.4 ± 2.5⁎ 26.9 ± 1.5 40.0 ± 5.1⁎ 66.1 ± 5.2⁎ b 0.0001 0.0253 0.0038

3.16 ± 1.2 3.60 ± 3.1 3.95 ± 2.5 1.84 ± 1.0 3.22 ± 1.7⁎ 4.35 ± 1.9⁎ b 0.0001 0.0025 0.0077

Mean ± SEM. ⁎Significantly different from WT of same gender (⁎P b 0.05).

eMS/BS than males. BFR/BS at each surface tended to reflect MS/BS values (data not shown). Cortical microarchitecture For all three area endpoints, (total area (TF.Ar), cortical area (FCo. Ar), and medullary canal area (FMe.Ar)), males had larger values than females (Table 5). Though TF.Ar was not influenced by genotype, FCo. Ar was higher and FMe.Ar was lower in CatK (−/−) mice than in WT mice of both genders. Both CSMI and cortical thickness (Ct.Th) were significantly greater in CatK (−/−) mice than in WT mice of both genders (Table 5). Bone strength Fig. 1. Bone mineral density (BMD) of distal (A) and central (B) femur (Mean ± SEM). There was a significant difference among genotypes with CatK (−/−) mice exhibiting higher BMD than WT mice (p b 0.001)⁎. CatK (−/−) females demonstrated significantly higher BMD than WT females (p b 0.01). The central femur (B) demonstrated only a genotype difference (p b 0.001)+. CatK (−/−) males showed significantly higher central femur BMD than WT males (p b 0.05)⁎. Female CatK (−/−) mice showed significantly higher BMD than WT female mice (p b 0.01).

no genotype difference, but was influenced by gender, being higher in females than males (Table 3). Though osteoclast number (Oc#/BS) was positively influenced by genotype, osteoclast surface (OcS/BS) was influenced by neither gender nor genotype (Table 3). Cortical histomorphometry Lamellar cortical area (TiLmCo.Ar/Co.Ar) was significantly lower in CatK (−/−) mice than in WT mice of the same gender (Fig. 4 [males], Table 4). The non-lamellar bone was generally buried under lamellar bone, rather than near either surface. Periosteal mineralizing surface (pMS/BS) was significantly greater in female, but not male CatK (−/−) mice (Table 4). Though endocortical mineralizing surface (eMS/BS) showed no genotype effect, females exhibited significantly greater

Table 1 BMD and mechanical properties of lumbar vertebral body Sex

Genotype

Ultimate load (N)

Ultimate stress (N/mm2)

Vertebral body bone mineral density (BMD, mg/cm2)

M M M F F F

Wild-type +/− (−/−) Wild-type +/− (−/−) Genotype effect Gender effect Interaction

42.0 ± 4.8 41.5 ± 7.4 54.5 ± 6.5 45.4 ± 4.4 35.7 ± 3.1 46.3 ± 5.4 0.1601 0.4441 0.5080

19.8 ± 3.0 19.0 ± 2.9 23.5 ± 2.7 25.8 ± 2.7 17.0 ± 1.6 20.8 ± 2.1 0.1933 0.8592 0.2280

51.6 ± 2.0 53.3 ± 1.9 52.5 ± 2.8 51.1 ± 1.7 52.0 ± 3.0 56.1 ± 3.7 0.5474 0.7785 0.6568

The fracture lines for all femurs, and all but three vertebral bodies (excluded from analyses), occurred perpendicular to the long axis of the respective bones. CatK (−/−) mice had greater central femur ultimate load values, but not femoral neck ultimate load values, than WT mice (Table 6). Neither ultimate stress nor post-yield deflection in the central femur showed significant genotype differences. There were no significant genotype or gender differences in lumbar vertebral body ultimate load (Table 1). Though males had greater values for central femur and femoral neck ultimate load than females, females had greater values for central femur ultimate stress than males (Table 6). Bone strength vs. bone mineral content There was an excellent correlation between FF.U and BMC in the central femur (r = 0.83, P b 0.0001; N = 50) for all mice. Central femur BMC and cortical area were highly correlated (r = 0.76, P b 0.0001; N = 50) for all mice. FF.U and FCo.Ar were highly correlated (r = 0.82, P b 0.0001) for all mice. There was also a positive correlation between FF.U and CSMI (r = 0.70, P b 0.0001). When BMC, FCo.Ar, and CSMI were tested by multiple regression as predictors of FF.U, only BMC and FCo.Ar remained independent predictors, together explaining ∼74% of the variance in bone strength in the femur. There was a solid correlation between LVF.U and BMC in the vertebral body (r = 0.53, P b 0.003; N = 44) for all mice. The individual gender/genotype groups, occasionally with only five tested bones/group, had too few datapoints to permit biologicallymeaningful interpretation of the F.U–BMC relationship. When the genotypes were separated, the relationship of FF.U and BMC at the central femur were highly correlated within WT, heterozygote, and (=/=) groups (r = 0.56 to 0.85, P = 0.006 to 0.0001, Table 7). The slopes of the three groups did not differ (P = 0.543) (Fig. 5A). In the vertebral body, the relationships of LVF.U and BMC were correlated in WTs and (−/−'s) (r = 0.64 for both, P = 0.046 to 0.003, Table 7), again with no differences in the slopes among the three groups (P = 0.81) (Fig. 5B). When the genders were separated, the relationship of FF.U and BMC in the central femur was excellent (r = 0.83 and r = 0.89, P b.0001) for males and females, respectively (Table 7). The relationship of LVF.U

B. Pennypacker et al. / Bone 44 (2009) 199–207

203

Table 3 Trabecular bone formation and resorption (distal femur) Sex

Genotype

Mineralizing surface (MS/BS, %)

Mineral apposition rate (MAR, μm/d)

Bone volume-based formation rate (BFR/BV) (mm3/mm2/yr)

Tissue volume-based formation rate (BFR/TV) (mm3/mm2/yr)

Osteoclast #/BS (Oc#/BS, mm− 1)

Osteoclast surface (OcS/BS, %)

M M M F F F

WT +/− (−/−) WT +/− (−/−) Genotype Gender Interaction

13.4 ± 1.2 23.0 ± 3.1⁎ 24.7 ± 2.5⁎ 20.2 ± 1.0 26.4 ± 1.7⁎ 30.4 ± 1.9⁎ b0.0001 0.0051 0.7638

1.66 ± 0.13 1.93 ± 0.13 2.25 ± 0.15⁎ 2.05 ± 0.07 2.56 ± 0.23⁎ 3.36 ± 0.21⁎ b 0.0001 b 0.0001 0.0301

47 ± 8 79 ± 15⁎ 81 ± 15⁎ 121 ± 12 127 ± 9 120 ± 24 0.2197 b 0.0001 0.2212

51 ± 6 130 ± 34⁎ 159 ± 21⁎ 57 ± 7 178 ± 42⁎ 322 ± 37⁎ b0.0001 0.0890 0.0043

0.43 ± 0.32 0.90 ± 0.32 1.50 ± 0.42⁎ 1.00 ± 0.63 1.12 ± 0.61 1.44 ± 0.31⁎ 0.0009 0.2581 0.4113

2.6 ± 1.9 3.4 ± 0.6 4.3 ± 1.3 4.1 ± 2.2 3.5 ± 1.3 4.5 ± 0.3 0.2468 0.2711 0.4189

Mean ± SEM; ⁎Significantly different from WT of same gender (P b 0.05).

and BMC in the vertebral body was also solid (r = 0.74 and r = 0.41, P b.0001 and P b.035) for males and females, respectively (Table 7). There was no gender-related difference in the slopes of the regression lines at either the central femur or vertebral body respectively (P = 0.462 and P = 0.345). Discussion The aim was to study an adult animal model that may resemble CatK null or deficient phenotypes in humans, to understand better the influence of from-birth CatK deficiency on bone formation, bone mass, and bone strength. In addition to confirming the increased bone mass of CatK deficient mice in cortical and trabecular regions seen in previous studies, we found that higher bone mass is associated with a proportionate increase in bone strength in the femoral midshaft and vertebral body without differences in ultimate stress or brittleness. There is no genotype-related difference in the relationship of strength to bone mass in either site. In trabecular bone only, CatK null mice had elevated surface-based and total tissue-based bone formation rates with a normal bone volume-based bone formation rate. These data indicate that the skeleton in young adult mice with from-birth lack of CatK function not only has increased mass with normal bone tissue quality, but also displays a normal rate of volume-based bone formation rate in trabecular regions. CatK is an osteoclast-specific enzyme whose gene was discovered in bone using differential display of osteoclast and macrophage cDNA libraries from rabbits [26]. Human, rhesus monkey, rat, and murine CatK enzymes have also been cloned [27–29]. A role for CatK in bone

Fig. 2. Surface-based bone formation rate (BFR/BS) in distal femur (Mean ± SEM). There was a significant difference among genotypes, with CatK (−/−) mice exhibiting increased BFR compared to WT mice (p b 0.001).⁎ Both CatK (−/−) mice and heterozygote bone formation rate were significantly different from WT of the same gender (p b 0.01). Surface-based bone formation rate demonstrated a significant gender difference (p = 0.002) with females exhibiting higher bone formation rate than males.

resorption was first suggested by its sub-localization within osteoclasts to the lysosomes and ruffled border, and in Howship's lacunae [30]. The role of CatK in bone resorption was also suggested by bone changes caused not only by loss-of-function mutations in the CatK enzyme [11,16,17], but also by specific small molecule inhibitors of CatK in animals [31–33]. Our data document the existence of high bone mass in 19-week-old CatK null mice in both trabecular and cortical compartments of long bones, though not in vertebral bodies. Long bones of CatK (−/−) mice have higher trabecular bone volume, smaller medullary cavities, and more retained woven cortical bone, than WT mice. Coupling these CatK-related differences with the lack of significant differences in external dimensions of the bones suggests

Fig. 3. Fluorescent photomicrographs of distal femoral metaphyseal trabecular bone. (A) Trabecular bone of the secondary spongiosa in WT female mouse. Note low trabecular bone volume with the presence of occasional double-labeled surface, reflecting data in Fig. 2. (B) Trabecular bone of the secondary spongiosa in CatK (−/−) female mouse. Note higher trabecular bone volume and greater fraction of surface with double-labeled surface as compared to WT female mouse, again reflecting data in Fig. 2.

204

B. Pennypacker et al. / Bone 44 (2009) 199–207

Fig. 4. Brightfield and polarized light photomicrographs of tibial diaphyseal cortical bone cortex (A, brightfield; B, polarized light) from male WT mouse was comprised of lamellar (L) bone tissue (outlined in red) intermixed with areas of woven (W) bone tissue. CatK (−/−) mice (C and D) showed a higher percentage of cortical bone composed of non-lamellar bone. Lamellar bone was found at most surfaces in CatK (−/−) mice.

that the primary cause of increased bone mass is sub-normal resorptive activity that occurs chronically at surfaces adjacent to marrow cavities. This sub-normal resorptive activity should be viewed in the light of the two to three-fold increase in osteoclast numbers observed at trabecular surfaces, that should be taken as a morphologic indication that infers hypofunction of individual osteoclasts in CatK (−/−) mice. This inhibition of resorption not only leads to elevated bone mass, cortical thickness, and bone strength, but also implies that bone tissue produced by osteoblasts in the absence of CatK is of normal quality. Pycnodysostosis, a very rare human condition in which the skeleton develops, models, and remodels in the absence of CatK, is characterized by high bone mass, short stature, and reports of increased propensity to fracture, that may suggest impaired bone

quality [11,13,15,20,34]. Moreover, young adult CatK (−/−) mice have some phenotypic features of pycnodysostosis, including increased marrow space trabeculation, demineralized bone under osteoclasts, and splenomegaly [17]. However, our findings indicate that high bone mass in young adult CatK (−/−) mice is accompanied by proportionately high bone strength, with no genotype effect on the relationship of ultimate load to bone mineral content. Our data agree with others' that also show increased bone strength in adult CatK null mice [21]. Thus, these biomechanical and bone mass data from young adult mice suggest that bone tissue of normal quality is an inherent feature of bones that develop in the absence of CatK. Our data, though encouraging, cannot resolve the issue of bone quality in the absence of CatK. Excessive bone fragility reported in pycnodysostotic humans may develop only over decades of complete lack of CatK, a much

Table 4 Cortical bone histomorphometry (tibial diaphysis) Sex

Genotype

Periosteal mineralizing surface (pMS/BS, %)

Periosteal mineral apposition rate (pMAR, μm/d)

Endocortical mineralizing surface (eMS/BS, %)

Endocortical mineral apposition rate (eMAR, μm/d)

Lamellar cortical area (LmCo.Ar/Co.Ar, %)

M M M F F F

Wild-type +/− (−/−) Wild-type +/− (−/−) Genotype Gender Interaction

8.6 ± 3.3 29.4 ± 7.2⁎ 12.5 ± 5.8 7.9 ± 1.8 9.8 ± 3.0 20.0 ± 3.1⁎ 0.0285 0.1188 0.4758

0.97 ± 0.11 (3) 1.26 ± 0.11 (5) 1.25 ± 0.25 (2) 0.97 ± 0.06 (2) 1.24 (1) 0.80 ± 0.40 (3) 0.8012 0.2319 0.3067

13.1 ± 1.8 16.4 ± 4.1 13.8 ± 2.8 27.8 ± 2.1 21.2 ± 2.1 28.9 ± 3.6 0.9278 b 0.0001 0.9622

1.13 ± 0.12 (5) 1.33 ± 0.22 (4) 1.65 ± 0.14 (5) 1.44 ± 0.08 (12) 1.45 ± 0.09 (8) 1.52 ± 0.14 (6) 0.0489 0.3640 0.0727

41.5 ± 3.7 31.71 ± 5.48⁎ 26.6 ± 4.1⁎ 47.48 ± 5.35 32.44 ± 4.82⁎ 28.7 ± 2.4⁎ 0.0013 0.2341 0.7667

Mean ± SEM. ⁎Significantly different from WT of same gender (⁎P b 0.05). Numbers in parentheses indicate number of animals with double labels on the periosteal and endocortical surfaces.

B. Pennypacker et al. / Bone 44 (2009) 199–207

205

Table 5 Right central femur cortical microarchitecture Sex

Genotype

Total area (mm2, TF.Ar)

Cortical area (mm2, Co.Ar)

Medullary canal area (mm2, Me.Ar)

Moment of inertia (mm4, CSMI)

Cortical thickness (mm, Ct.Th)

M M M F F F

Wild-type +/− (−/−) Wild-type +/− (−/−) Genotype Gender Interaction

2.41 ± 0.09 2.47 ± 0.07 2.37 ± 0.12 1.81 ± 0.06 1.83 ± 0.07 1.99 ± 0.15 0.3965 b 0.0001 0.2857

1.16 ± 0.03 1.34 ± 0.09 1.46 ± 0.13⁎ 0.94 ± 0.03 0.95 ± 0.04 1.17 ± 0.14⁎ 0.0011 b 0.0001 0.5023

1.25 ± 0.06 1.12 ± 0.10 0.91 ± 0.06⁎ 0.87 ± 0.05 0.87 ± 0.06 0.82 ± 0.06 0.0109 b0.0001 0.0305

0.23 ± 0.02 0.26 ± 0.02 0.28 ± 0.04 0.14 ± 0.01 0.14 ± 0.01 0.19 ± 0.08 0.0216 b 0.0001 0.8476

0.246 ± 0.011 0.292 ± 0.067 0.330 ± 0.069⁎ 0.234 ± 0.021 0.236 ± 0.025 0.284 ± 0.065⁎ 0.0008 0.0061 0.3243

Mean ± SEM; ⁎Different from WT of same gender (P b 0.05).

longer period than can be addressed in animal studies. It is possible that biomechanical tests applied in experimental animals do not address strength/quality at bone sites at which fractures occur in pycnodysostotic people. It is also possible that the small number of pycnodysostosis patients reported may be insufficient to truly represent the full spectrum of bone strength findings in humans with from-birth absence of CatK, as persons who lack CatK but have no bone problems, are unlikely to be examined in the current eventdriven reporting system. Our data do suggest that the high bone mass in young adult, small animals with from-birth lack of CatK is associated with appropriately high bone strength and normal quality of bone tissue. A previous report suggests that cortical bone of 10-week-old CatK (−/−) mice contains a smaller proportion of lamellar bone than WT mice [19]. Our findings indicate that this condition persists into young adulthood. In 10-week-old mice, non-lamellar bone extends to the endocortical surface [19], whereas 19-week-old mice generally have non-lamellar bone only deeply buried within the cortex (Fig. 4). Bone tissue formed during young adulthood in CatK (−/−) mice appears to be lamellar. Moreover, despite the lower lamellar content in cortical bone of CatK (−/−) mice, their bones not only had higher BMD and thicker cortices, but were also stronger. It is reasonable to interpret our data to say that the high non-lamellar content of cortical bone in CatK null mice is caused by slow removal of non-lamellar bone normally formed during early skeletal development, rather than abnormal osteoblast function that causes formation of non-lamellar bone in the absence of CatK during adulthood. The strength data further confirm that this amount of non-lamellar bone in adult CatK (−/−) mice, while significantly higher than in WT mice, is not functionally-associated with reduced bone quality. It also suggests an ability of the bone modeling system to mix and match cortical microarchitecture (lamellar content) and bone quantity to maintain appropriate bone strength.

Cortical bone characteristics in 10-week-old CatK null mice include increased brittleness, indicated by decreased post-yield deflection [19]. Post-yield deflection did not differ with either genotype or gender in the current study of 19-week-old CatK (−/−) mice. The

Table 6 Mechanical properties of right femur Sex

Genotype

Central femur ultimate load (N)

Ultimate stress (N/mm2)

Post-yield deflection (mm)

Femoral neck ultimate load (N)

M M M F F F

Wild-type +/− (−/−) Wild-type +/− (−/−) Genotype effect Gender effect Interaction

25.6 ± 1.4 30.9 ± 2.4 36.2 ± 1.7⁎ 24.1 ± 0.8 25.2 ± 1.3 28.2 ± 2.3 b0.0001 0.0002 0.0321

286 ± 19 301 ± 17 329 ± 31 388 ± 14 370 ± 11 356 ± 21 0.9689 b 0.0001 0.0516

0.214 ± 0.145 0.337 ± 0.110 0.276 ± 0.165 0.287 ± 0.122 0.256 ± 0.110 0.322 ± 0.101 0.3412 0.7497 0.3178

32.3 ± 2.0 31.8 ± 2.3 33.2 ± 1.7 24.4 ± 1.2 23.5 ± 2.1 26.0 ± 4.1 0.6698 0.0001 0.9369

Mean ± SEM. Significantly different from WT of same gender (⁎P b 0.05).

Fig. 5. Ultimate load/BMC relationship for central femur and lumbar vertebral body. (A) For the central femur, the three regression lines for the genotype groups are plotted ( WT [r = 0.56; P = 0.006, N = 22]; —————————— +/− [r = 0.85; P = 0.0001, N = 16]; ··············· −/− [r = 0.79; P = 0.002, N = 12]). There are no significant differences among the slopes of the lines for the genotype groups (P = 0.543) (Table 7). (B) For the lumbar vertebral body, the three regression lines for the genotype groups are plotted ( WT [r = 0.64; P = 0.003, N = 20]; —————————— +/− [r = 0.35; P =0.21, N = 14]; ··············· −/− [r = 0.64; P = 0.046, N = 10]). There are no significant differences among the slopes of the lines for the genotype groups (P = 0.81) (Table 7).

–––– ––––

206

B. Pennypacker et al. / Bone 44 (2009) 199–207

Table 7 Regression of ultimate load vs. bone mineral content for femur and vertebral body subgroups Group Central femur Male Female Wild-types Heterozygotes Knockout Vertebral body Male Female Wild-types Heterozygotes Knockout

N

R

P

Slope (±SEM)

Yi (±SEM)

21 29 22 16 12

0.83 0.89 0.56 0.85 0.79

0.00001 0.00001 0.006 0.0001 0.002

3.63 ± 0.56 1.98 ± 0.20 2.05 ± 0.67 2.42 ± 0.40 2.99 ± 0.74

12.6 ± 1.33 12.6 ± 1.30 12.4 ± 4.1 10.3 ± 3.0 7.57 ± 6.18

17 27 20 14 10

0.74 0.41 0.64 0.35 0.64

0.0007 0.035 0.003 0.22 0.046

3.53 ± 0.83 1.21 ± 0.54 2.72 ± 0.78 1.03 ± 0.80 1.70 ± 0.72

−27.0 ± 17.9 18.45 ± 10.99 − 9.2 ± 15.8 17.4 ± 16.3 15.8 ± 15.5

contrasting findings may be age-related. It is intriguing to consider the possibility that the decrease in non-lamellar tissue that occurs between ages 10 and 19 weeks is associated with a normalization of post-yield deflection. The contrasting finding on brittleness may also be caused by the use of different testing techniques (four-point [19] vs. three-point bending [current]) that may characterize post-yield deflection differently [23]. A difference in strain background (C57BL/ 6J [19] vs. 129SVJ-C57BL/6J [this study]) may also play a role. In a larger sense, it seems relevant to point out that whole bones function in vivo in the pre-yield loading range during essentially all activities, while post-yield characteristics describe behavior after permanent deformation has occurred. One might speculate that while all bone biomechanical properties observed during ex vivo testing are interesting, it is pre-yield properties, such as those describing yield load or stiffness, that best predict a bone's in vivo function, because bones spend virtually all their time in vivo in the pre-yield loading range. It could be that pre-clinical findings of increased ultimate load, yield load, and stiffness are more important for predicting how a treatment or condition will influence fragility fracture risk in humans, than is brittleness variation that is never reflected clinically (e.g., a poor ability to set and heal fractures). Our data show that young adult CatK (−/−) mice have higher cortical BMD and periosteal formation rate than WT mice. Periosteal formation rate is elevated in young adult cynomolgus monkeys treated for 18 months with a small molecule, reversible CatK inhibitor [35]. Cortical bone volume and thickness are greater in OVX mice treated with a small molecule CatK inhibitor [36]. Though increased periosteal formation exists in both CatK null mice and in non-human primates given a specific CatK inhibitor, the higher level of periosteal formation in mice is insufficient to impact external bone size by age 19 weeks. In the current study, the smaller medullary cavities and increased cortical thickness, combined with normal external dimensions, despite elevated periosteal bone formation in CatK (−/−) mice, indicate that their higher cortical bone quantity is due mostly to decreased resorption rate at the endocortical surface. Decreased endocortical resorption in CatK (−/−) mice is consistent with a role for CatK in bone resorption. Like others [16–18], we report that young adult CatK (−/−) mice have higher BMD in trabecular regions than WT mice. The higher trabecular bone quantity implied by distal femur DXA in CatK (−/−) mice is confirmed by histomorphometric data including BV/TV, Tb.Th, and Tb.N. The difference in trabecular bone mass in CatK (−/−) mice appears more marked than the increase in cortical bone mass (75– 500% greater in trabecular bone (Table 2) vs. 20–30% higher in cortical bone (Fig. 1B)). The more profound effect on trabecular bone mass in CatK (−/−) mice is associated with both a higher bone formation rate in trabecular bone than in WT mice, and the generally higher bone formation rate in trabecular bone than cortical bone. One study of 8-week-old mice found that male CatK (−/−) mice had higher total tissue volume-based, lower volume-based, and

similar surface-based bone formation rates in trabecular bone when compared to WT mice [18]. A second study of 10-week-old CatK (−/−) mice of both genders found higher total tissue volume-based formation rate in trabecular regions [19]. Surface-based bone formation rate was also higher in male CatK (−/−) mice and trended higher in female CatK (−/−) mice [19]. Eight–ten weeks is an age at which patterns of fluorochrome labeling associated with rapid bone elongation in trabecular regions causes difficulties, because decreased resorption of bone tissue in the primary and secondary spongiosa induces retention of fluorochrome label that complicates its interpretation. We interpret the bone formation data of these two papers as strongly suggesting that trabecular bone formation rate is elevated in CatK (−/−) mice [18,19]. We evaluated 19-week-old mice of both genders, in which the rapid growth phase has abated [37]. Our data clearly indicate that surface-based and total volume-based trabecular bone formation rate is elevated in CatK (−/−) mice of both genders. Like both previous authors [18,19], we noted a significant increase in mineral apposition rate. Unlike the others, we also saw increased mineralizing surface that led to a three-fold increase in surface-based formation rate in females and a five-fold increase in males. Collectively, the three datasets seem to concur in documenting elevated surface-based and total volume-based bone formation rate, with a normal bone volume-based bone formation rate in CatK (−/−) mice. The elevated levels of surface-based bone formation in trabecular bone of young adult CatK (−/−) mice with no difference in bone volume-based bone formation rate between CatK (−/−) and WT mice, may suggest that trabecular bone tissue itself turns over at a normal rate in the absence of CatK. Essentially, the amount of surface forming appears sufficient to assure that the greater amount of trabecular bone tissue in CatK (−/−) mice turns over at a normal rate. Since there is unresolved debate over whether bone cell activity is organized as modeling or remodeling in trabecular bone of this age rodent [38], one should not directly imply the status of trabecular bone in either humans or large animals with bone remodeling, in conditions of CatK deficiency. In aggregate, these data do suggest that the decreased osteoclast activity and bone resorption that appears to exist in CatK null mice, is accompanied by a bone formation rate appropriate for the volume of trabecular bone. This sort of phenotype has been previously demonstrated in humans with some types of osteopetrosis [39], suggesting that conditions associated with decreased osteoclast activity, but normal numbers, are not necessarily associated with decreased bone formation activity. The increased osteoclast number may indicate that the reduced rate of bone resorption in the absolute absence of CatK may be partially facilitated by non-CatK mediators of collagen dissolution, such as MMP-9, MMP-13, MMP-14, and CatL, as previously suggested by others [18]. In summary, young adult CatK (−/−) mice have higher bone mass and thicker cortices than WT mice in both cortical and trabecular regions. This increase in cortical bone quantity is associated with decreased resorption and increased osteoclast numbers at surfaces adjacent to bone marrow. Cortical bone is stronger in CatK (−/−) mice, commensurate with its increased bone mineral density and thicker cortices. The relationship of bone strength to bone mineral content is not influenced by genotype, indicating that bone tissue in CatK (−/−) mice is of normal quality. The increased trabecular bone quantity is associated with both reduced resorption and a normal volume-based formation rate. References [1] Turk B, Turk D, Turk V. Lysosomal cysteine proteases: more than scavengers. Biochim Biophys Acta 2000;1477:98–111. [2] Bromme D, Okamoto K, Wang B, Biroc S. Human cathepsin O2, a matrix protein-degrading cysteine protease expressed in osteoclasts. J Biol Chem 1996;271:2126–32. [3] Troen BR. The role of cathepsin K in normal bone resorption. Drug News Perspect 2004;17:19–28.

B. Pennypacker et al. / Bone 44 (2009) 199–207 [4] Bromme D, Okamoto K. Human cathepsin O2, a novel cysteine protease highly expressed in osteoclastomas and ovary molecular cloning, sequencing and tissue distribution. Biol Chem Hoppe-Seyler 1995;376:379–84. [5] Delaisse JM, Eeckhout Y, Vaes G. In vivo and in vitro evidence for the involvement of cysteine proteases in bone resorption. Biochem Biophys Res Commun 1984;125:441–7. [6] Everts V, Delaisse JM, Korper W, Beersen W. Cysteine proteases and matrix metalloproteinases play distinct roles in the subosteoclastic resorption zone. J Bone Miner Res 1998;13:1420–30. [7] Hill PA, Buttle DJ, Jones SJ, Boyde A, Murata M, Reynolds JJ, et al. Inhibition of bone resorption by selective inactivators of cysteine proteinases. J Cell Biochem 1994;56:118–30. [8] Van Noorden CFJ, Everts V. Selective inhibition of cysteine proteinases by Z-PheAlaCH2D suppress digestion of collagen by fibroblasts and osteoclasts. Biochem Biophys Res Commun 1991;178:178–84. [9] Drake F, Dodds R, James I, Connor J, Debouck C, Richardson S, et al. Cathepsin K, but cathepsins B, L, or S, is abundantly expressed in human osteoclasts. J Biol Chem 1996;271:12511–6. [10] Votta B, Levy M, Badger A, Bradbeer J, Dodds R, James I, et al. Peptide aldehyde inhibitors of cathepsin K inhibit bone resorption both in vitro and in vivo. J Bone Miner Res 1997;12:1396–406. [11] Gelb B, Shi G, Chapman H, Desnick R. Pycnodysostosis, a lysosomal disease caused by cathepsin K deficiency. Science 1996;273:1236–8. [12] Johnson M, Polymeropoulos M, Vos H, Ortiz de Luna R, Francomano C. A nonsense mutation in the cathepsin K gene observed in a family with pycnodysostosis. Genome Res 1996;6:1050–5. [13] Ho N, Punturieri A, Wilkin D, Szabo J, Johnson M, Whaley J, et al. Mutations of CTSK result in pycnodysostosis via a reduction in cathepsin K protein. J Bone Miner Res 1999;14:1649–53. [14] Everts V, Aronson D, Beertsen W. Phagocytosis of bone collagen by osteoclasts in two cases of pycnodysostosis. Calcif Tissue Int 1985;37:25–31. [15] Everts V, Hou WS, Rialland X, Tigchelaar W, Saftig P, Bromme D, et al. Cathepsin K deficiency in pycnodysostosis results in accumulation of non-digested phagocytosed collagen in fibroblasts. Calcif Tissue Int 2003;73:380–6. [16] Saftig P, Hunziger E, Wehmeyer O, Jones S, Boyde A, Rommerskirch W, et al. Impaired osteoclastic bone resorption leads to osteopetrosis in cathepsin K deficient mice. Proc Natl Acad Sci 1998;95:13453–8. [17] Gowen M, Lazner F, Dodds R, Kapadia R, Feild J, Tavaria M, et al. Cathepsin, K knockout mice develop osteopetrosis due to a deficit in matrix degradation but not demineralization. J Bone Miner Res 1999;14:1654–63. [18] Kiviranta RMJ, Alatalo S, NicAmhlaoibh R, Risteli J, Laitala-Leinonen T, Vuorio E. Impaired bone resorption in cathepsin K-deficient mice is partially compensated for by enhanced osteoclastogenesis and increased expression of other proteases via an increased RANKL/OPG ratio. Bone 2005;36:159–72. [19] Li CY, Jepsen K, Majeska RJ, Zhang J, Gelb BD, Schaffler MB. Mice lacking cathepsin K maintain bone remodeling but develop bone fragility despite high bone mass. J Bone Miner Res 2006;21(6):865–75. [20] Fratzl-Zelman N, Valenta A, Roschger P, Nader A, Gelb B, Fratzl P, et al. Decrease bone turnover and deterioration of bone structure in two cases of pycnodysostosis. J Clin Endocrinol Metab 2004;89(4):1538–47.

207

[21] Hoffman S, Shen V, Liang X, Capriotti C, Stroup G, Kumar S. Knockout of Cathepsin K in adult mice does not result in bone fragility. J Bone Miner Res 2007;22(S1):S378. [22] Parfitt M, Glorieux F, Kanis J, Malluche H, Meunier P, Ott S, et al. Bone histomorphometry: standardization of nomenclature, symbols, and units. J Bone Miner Res 1987;2:595–610. [23] Turner CH, Burr DB. Basic biomechanical measurements of bone: a tutorial. Bone 1993;14:595–608. [24] Snedecor GWaC. W.G. Statistical Methods. 8th ed. Ames, Iowa: Iowa: State University Press; 1989. [25] SAS Institute, I.. SAS/STAT User's Guide. Cary, NC; 1999. p. 1467–636. [26] Tezuka K, Tezuka Y, Maejima A, Sato T, Nemoto K, Kamioka H, et al. Molecular cloning of a possible cysteine proteinase predominantly expressed in osteoclasts. J Biol Chem 1994;269:1106–9. [27] Inaoka T, Bilbe G, Ishibashi O, Tezuka K, Kumegawa M, Kokubo T. Molecular cloning of human cDNA for cathepsin K: novel cysteine protease predominantly expressed in bone. Biochem Biophys Res Commun 1995;206:89–96. [28] Rantakokko J, Aro HT, Savontaus M, Vuorio E. Mouse cathepsin K: cDNA cloning and predominant expression of the gene in osteoclasts, and in some hypertrophying chondrocytes during mouse development. FEBS Lett 1996;393:307–13. [29] Guay J, Riendeau D, Mancini JA. Cloning and expression of rhesus monkey cathepsin K. Bone 1999;25(2):205–9. [30] Yamaza T, Goto T, Kamiya T, Kobayashi Y, Sakai H, Tanaka T. Study of immunoelectronmicroscopic localization of cathepsin K in osteoclasts and other bone cells in the mouse femur. Bone 1998;23:499–509. [31] Lark MW, Stroup GB, James IE, Dodds RA, Hwang SM, Blake SM, et al. A potent small molecule, nonpeptide inhibitor of cathepsin K (SB 331750) prevents bone matrix resorption in the ovariectomized rat. Bone 2002;30(5):746–53. [32] Stroup GB, Lark MW, Veber DF, Bhattacharyya A, Blake S, Dare LC, et al. Potent and selective inhibition of human cathepsin K leads to inhibition of bone resorption in vivo in a nonhuman primate. J Bone Miner Res 2001;16(10):1739–46. [33] Li CS, Deschenes D, Desmarais S, Falgueryet JP, Gauthier JY, Kimmel DB, et al. Identification of a potent and selective non-basic cathepsin K inhibitor. Bioorg Med Chem Lett 2006;16:1985–9. [34] Motyckova GaDEF. Pycnodysostosis: role and regulation of cathepsin K in osteoclast function and human disease. Curr Mol Med 2002;2:407–21. [35] Jerome CP, Missbach M, Gamse R. AAE581, a novel cathepsin K inhibitor, protects against ovariectomy-induced bone loss in non-human primates, in part by stimulation of periosteal bone formation. J Bone Miner Res 2005;20(Suppl. 1): S46. [36] Xiang A, Kanematsu M, Kumar S, Yamashita D, Kaise T, Kikkawa H, et al. Changes in micro-CT 3D bone parameters reflect effects of a potent cathepsin K inhibitor (SB-553484) on bone resorption and cortical bone formation in ovariectomized mice. Bone 2007;40:1231–7. [37] Silbermann MaTK. Age-related changes in cellular population of growth plate of normal mouse. Acta Anat 1977;97(4):459–68. [38] Erben RG. Trabecular and endocortical bone surfaces in the rat: modeling or remodeling? Anat Rec 1996;245:39–46. [39] Bollerslev J, Steiniche T, Melsen F, Mosekilde L. Structural and histomorphometric studies of iliac crest trabecular and cortical bone in autosomal dominant osteopetrosis: a study of two radiological types. Bone 1989;10:19–24.