| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Ludwig Boltzmann Institute of Osteology, Fourth Medical Department, Hanusch Hospital and UKH-Meidling (N.F.-Z., A.V., P.R., A.N., K.K.), A-1140 Vienna, Austria; Erich Schmid Institute of Materials Sciences, Austrian Academy of Sciences and University of Leoben (A.V., P.F.), A-8700 Leoben, Austria; Institute for Pathology, Hanusch Hospital (A.N.), A-1140 Vienna, Austria; and Departments of Pediatrics and Human Genetics, Mount Sinai School of Medicine (B.D.G.), New York, New York 10029
Address all correspondence and requests for reprints to: Dr. Klaus Klaushofer, Ludwig Boltzmann Institute of Osteology, 4th Medical Department, Hanusch Hospital, Heinrich Collin Strasse 30, A-1140 Wien, Austria. E-mail: klaus.klaushofer{at}univie.ac.at.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The skeletal abnormalities developed by cathepsin K-deficient mice mimic pycnodysostosis, an extremely rare human osteosclerotic disorder. Gelb and co-workers (6) used a positional cloning strategy to identify the cathepsin K gene (CTSK) as the cause of pycnodysostosis. Since then, several mutations causing pycnodysostosis have been characterized, and all result in the loss of protease function (7, 8). Histological analysis revealed that osteoclasts from patients with pycnodysostosis and from cathepsin K-deficient animals manifest a similar morphology; they contain large cytoplasmatic vacuoles with undigested collagen (9), and, due to the lack of proteolytic activity, they form demineralized matrix fringes on the bone surface (3, 4).
Interestingly, bone resorption is defective, but not completely inhibited, in cathepsin K-deficient mice as well as in pycnodysostosis. In fact, it appears that an alternative, nonosteoclastic, biochemical pathway is available for degradation of bone matrix (10). When cathepsin K activity is absent, bone-lining cells that have matrix metalloproteinase activities enter into the resorption lacunae and degrade demineralized collagen left over from osteoclasts (11). It is not clear, however, whether the two pathways are equivalent. The collagen breakdown products are different, and only cathepsin K is competent to solubilize trivalent, cross-linked bone collagen to small peptides (2, 12, 13). This raises the possibility that the distinct pattern of collagen cleavage from alternative proteinases may also alter the microenvironment of bone. Impaired matrix degradation due to defective cathepsin K may further contribute to alterations in osteoblast function and bone formation that would enhance the osteosclerotic phenotype. Indeed, a recent study reported that overexpression of cathepsin K in mice causes high bone turnover osteopenia, emphasizing an important role of cathepsin K in the bone remodeling cycle (14).
Despite the increased bone mass, patients with pycnodysostosis suffer from bone fragility, with a high incidence of pathological fractures, a clinical observation that is not fully understood. Cathepsin K deficiency might lead to problems in trabecular bone at the tissue and material levels, which is optimized to the biomechanical and structural needs through balanced bone turnover. To date, there has been no systematic study of bone structure and quality in pycnodysostosis. We believe that such data would be very valuable, given that cathepsin K antagonists are already being tested in preclinical studies as a possible therapeutic option for patients with metabolic bone diseases such as osteoporosis (15).
In the present study we compared bone tissue from iliac crest biopsies from two unrelated patients with pycnodysostosis, aged 5 and 21 yr, to published normative data. To confirm the clinical diagnosis, CTSK mutations were identified, and one proved to be novel. Keeping in mind that pycnodysostosis is an extremely rare disorder and due to the invasive character of the procedure, bone biopsies are not routinely performed for clinical purposes and would be very difficult to rationalize ethically on a research basis, we did not attempt to increase the number of bone samples. Hence, we concentrated on using a wide range of methods for extracting information on bone structure at microscopic and submicroscopic levels from the two biopsies. We used histology, histomorphometry, quantitative backscattered electron imaging (qBEI), as well as small angle x-ray scattering (SAXS) to obtain information on bone cells, matrix structure, and quality, and we found very similar results in both subjects.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient A was a 5-yr-old boy with short stature and typical facial features. Roentgenograms revealed osteosclerosis, dysostosis, and loss of the mandibular angle. No fractures were observed. He developed obstructive sleep apnea, requiring surgical intervention. His younger sister was also affected with pycnodysostosis.
Patient B was a 20-yr-old female with short stature (145 cm in height), obtuse mandibular angle, dental defects, a history of maxillar and mandibular surgery, and a total of 34 fractures without adequate trauma since early childhood. Later, the patient had a successful pregnancy.
Blood was drawn, and transiliac bone biopsies from these two patients were obtained in the standard manner. For patient A, blood was drawn for research purposes with a consent form that had been approved by the Mount Sinai School of Medicine institutional review board. The bone biopsy was performed on a clinical basis, and the existing archived pathological sample was provided to us anonymously. For patient B, blood was drawn, and bone biopsy was performed primarily on clinical basis to exclude metabolic comorbidity such as latent nutritional osteomalacia and to assess the risk for atraumatic fractures based on the patient’s desire to become pregnant. The patient was informed and gave her consent in agreement with the local ethics committee’s regulations.
The histological and histomorphometrical parameters were compared with age-matched normative data for iliac crests (16). Data concerning bone material, such as calcium content, mineral particle thickness, and orientation, were compared with normative data obtained previously from vertebral bone (17).
Determination of CTSK mutations
Genomic DNA was extracted from blood leukocytes using the Puregene genomic DNA isolation kit (Gentra Systems, Minneapolis, MN). Exons 2–8 of the CTSK gene (18) were amplified from genomic DNA of the patients with pycnodysostosis by PCR, isolated, and sequenced by cycle sequencing with an ABI 377 sequencer (PE Applied Biosystems, Foster City, CA). PCR-based restriction fragment length assays were designed for missense alleles and analyzed in 50 normal controls to exclude the possibility of a common polymorphism.
Sample preparation
The undecalcified iliac crest biopsy specimens were fixed in ethanol/formalin (70:30, vol/vol) and dehydrated in a graded series of ethanol before polymethylmethacrylate embedding. Consecutively, 3-µm-thick sections were cut from the tissue blocks with a Jung K microtome (Reichert-Jung, Heidelberg, Germany). The sections were deplastified with 2-methoxyethyl acetate before being stained with either Giemsa or a modified Goldner’s Trichrome method. The residual blocks were prepared with planoparallel surfaces by grinding and polishing and were coated with carbon by vacuum evaporation for the qBEI analysis in the scanning electron microscope (SEM). For the scanning SAXS measurements, 200-µm-thick sections were cut head-on from the blocks using a low speed diamond saw.
After the SAXS measurements were performed, the 200-µm-thick ground sections were glued on 1-mm-thick plastic carriers and then ground to about 100 µm in thickness. Surface staining with Giemsa was performed on these undeplastified ground sections, modified according to Gross and Strunz (19). In this method, mineralized bone stains pink, and mineralized cartilage stains dark purple. For this purpose, the samples were carefully etched with 0.2% formic acid for 30 sec to avoid observable decalcification and pretreated with 20% methanol at 4 C for 1 h. Subsequently, the sections were stained with 20% Giemsa (Merck & Co., Darmstadt, Germany) at 40 C with 30 min. Finally, these surface-stained, undecalcified bone sections were analyzed by light microscopy and, in parallel (identical bone areas) after carbon coating, by SEM using backscattered electrons.
Histology and two-dimensional histomorphometry
Histology and two-dimensional histomorphometric analyses were performed according to Parfitt (20) on the whole area of the bone sections. Additionally, a new parameter to quantify osteoclastic dysfunction, the volume of demineralized bone matrix per total bone volume, was introduced in the histomorphometric analysis. For this purpose the areas of resorption lacunae with demineralized bone matrix and nondigested collagen were measured. A light microscope (Axiophot, Zeiss, Oberkochen, Germany) equipped with a Zeiss AxioCam videocamera was used to obtain digital images of the section. The images were analyzed using standard procedures (NIH Image software versions 1.62 and 1.63, Wayne Rasband, National Institutes of Health, Bethesda, MD) on a Power Macintosh G4.
qBEI
qBEI with a digital SEM (DSM 962, Zeiss) equipped with a four-quadrant semiconductor backscattered electron detector was used to determine mineral density distribution of both cortical and cancellous bone from the biopsies. The accelerating voltage of the electron beam was adjusted to 20 kV, the probe current to 110 pA, and the working distance to 15 mm. The cortical and cancellous bone areas were imaged at x50 nominal magnification, corresponding to a pixel resolution of 4 µm/pixel, using a scan speed of 100 sec/frame, resulting in digital calibrated backscattered electron image of 512 x 512 pixels. The total area analyzed from each sample was approximately 30 mm2. From the digital images, gray level histograms displaying the percentage of bone area occupied by pixels of a certain gray level were derived. The transformation of these data into calcium weight percentage histograms led to a bin width of 0.17 wt % calcium. A technical precision of 0.3% was achieved. Full technical details of the technique and its precision have been reported previously (21, 22). Two parameters were obtained from the bone mineralization density distributions: the typical calcium concentration in the sample represents the peak position in the histogram, and the width of the calcium distribution is a measure of the homogeneity of mineralization within the sample (23).
Mineralized cartilage can clearly be discriminated from mineralized lamellar bone tissue due to its higher gray level in the back-scattered electron images that correspond to higher degree of mineralization and, further, by its more homogenous texture (24, 25, 26).
Scanning SAXS
Measurements were performed using an instrument equipped with a 12-kW rotating Cu-generator (CuK-
radiation at 0.154 nm wavelength, operating at 40 kv/40 mA), an evacuated pinhole camera (sample to detector distance, 39 cm), and a two-dimensional, position-sensitive proportional counter with a spatial resolution of 100 µm (Bruckner AXS, Karlsruhe, Germany) as described previously (27, 28). Each sample was mounted on a specimen holder that could be moved automatically with a precision of 2 µm in the plane perpendicular to the beam. The scanning SAXS patterns were corrected for background and analyzed for the mean particle thickness parameter, T, and degree of alignment, 
, as previously described (28). For the evaluation of T, bone was considered a two-phase mixture, consisting of organic matrix as the first phase and mineral as the second phase, strongly differing in electron density from the first. By definition, T = 4 x 
x (1 – )/
, where is the volume fraction of mineral, (1 – ) the volume fraction of organic phase, and is the interface area per unit volume between the two phases.
It can, therefore, be used as a measure of the smallest dimension of needle or plate-like crystals. The alignment of the particles within the x-ray beam interaction volume causes different scanning SAXS patterns, showing a narrow streak for perfect parallel alignment, a spherically shaped pattern for complete random orientation, and an elliptical pattern for all other configurations. Therefore, the shape of the scanning SAXS pattern can be used to determine the degree of alignment () of the mineral particles, with = 0 indicating no predominating orientation, and > 0 revealing the percentage of mineral particles that are not randomly aligned within the area of the x-ray beam.
Statistical analyses
Results from parameters of bone histomorphometry were compared with age-matched normative data published by Glorieux and colleagues (16). The statistical significance of the deviation of an individual measurement from the mean of these control data distribution was defined on basis of z-scores. Significance levels of P = 0.05 and P = 0.01 corresponded to z-scores of ±1.64 and ±2.33, respectively.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient A was homozygous for a C-to-G transversion at nucleotide 1039 of the CTSK gene, predicting the substitution of Arg312 with a Gly (R312G). This mutation has been reported previously, and expression of the R312G mutant protein resulted in unstable precursor and no active cathepsin K enzyme (7).
Patient B had compound heterozygosity for cathepsin K mutations. These defects were a G-to-A transition at nucleotide 340, predicting the substitution of Gly79 with an Arg (G79R), and deletion of the G nucleotide at position 363 (del363G). The G79R was not observed in 100 control chromosomes and closely resembled the previously reported G79E allele (7). Expression of G79E resulted in unstable precursor protein, presumably due to misfolding of the proenzyme. The del363G mutation caused a frameshift, resulting in a novel stop codon at residue 90 that eliminated the mature cathepsin K protein.
Confirmation of osteosclerosis by backscattered electron imaging and evaluation of structural histomorphometric parameters
Evaluation of the backscattered electron images (Fig. 1
) from both biopsies revealed increased cortical thickness as well as dense trabeculation of the bone marrow space. Patient A showed a clear delimitation between trabecular and cortical bone. Single trabeculae were visible, and the bone structure appeared to be homogeneous in thickness. In contrast, patient B displayed a gradual transition between a broad cortical zone and densely connected trabecular bone. In addition, there was a pronounced nonhomogeneity in thickness among the trabeculae, with thicker and thinner structures coexisting close together. The histomorphometric analyses, shown in Table 1, confirmed the increased bone mass. Bones from both patients displayed a significant increase of cortical width, trabecular thickness, and ratio of bone volume to tissue volume compared with age-matched controls. In contrast, trabecular number and separation were within the normal range.
|
|
In both biopsies, large inclusions of highly mineralized cartilage disturbing the homogeneity of the bone matrix were visible by back-scattered electron microscopy and histological surface staining with Giemsa. Cartilage residuals were identified on the undecalcified section by white features in the BEI image (Fig. 2A), which appeared in the light microscopic image as dark purple areas (Fig. 2B).
|
Light microscopic examination of the bone biopsies revealed the typical cellular events leading to osteosclerosis in pycnodysostosis patients. In contrast to normal bone, both biopsies had multinucleated osteoclasts adjacent to lacunae of nondigested collagen on the bone surface (Fig. 3, A–C). In patient A (Fig. 3, A and B) the osteoclasts appeared less flat, and the excavations of demineralized bone matrix appeared larger and deeper than in sample B (Fig. 3C). Demineralized fringes with detached (Fig. 3D, patient A) or without osteoclasts nearby (Fig. 3E, patient A; Fig. 3F, patient B) were observed.
|
, D–F). According to Everts and colleagues (11), these are metabolically highly active cells with matrix metalloproteinase activity, which degrade nondigested collagen left over from osteoclasts. Static histomorphometric parameters of bone formation and bone resorption revealed decreased bone formation
Comparison of the parameters of the pycnodysostotic bone to age-matched normative data (16) are shown in Table 2. Both subjects were found to have a significant decrease in parameters of bone formation (osteoid surface/bone surface, osteoid volume/bone volume, and osteoblast surface/bone surface), except for osteoid thickness, which was increased in the child and within the normal range in the adult. In particular there was a lack of cuboidal osteoblasts in patient B. In contrast, osteoclast parameters were increased in the child and were within the normal range in the adult.
|
The degree of matrix mineralization was decreased in the child, but normal in the adult
Evaluation of the qBEI histograms revealed a decrease in typical calcium values for sample A, whereas the degree of tissue mineralization was within normal range for the adult sample compared with age-matched controls (Fig. 4A). The values for the peak width of the calcium distribution were not altered (Fig. 4B).
|
The mineral thickness T, as measured by SAXS, was smaller in patient A (3.56 nm) than in patient B (3.83 nm) in agreement with the fact the mean crystal increases thickness with age (28). In both cases, however, the size was significantly larger compared with controls (17), indicating an older, less remodeled bone material (Fig. 5). The scanning SAXS measurements (Fig. 6) revealed a high variability in orientation of the mineral particles, reflecting a disordered architecture of the collagen fibrils in both specimens (Fig. 6, A and B). This contrasted sharply with the large degree of order that exists in normal bone as reported previously (Fig. 6C) (17, 28).
|
|
The pattern of lamellar organization of the trabecular architecture appeared highly disordered in both patients. As seen in polarized light microscopy (Fig. 7), the orientation of the collagen fibrils did not follow the direction of the trabeculae, but showed a very irregular pattern. In contrast, the fibril orientation follows the trabecular orientation closely in normal human bone tissue (28, 29).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Both individuals displayed severe osteosclerosis, a marked decrease in histomorphometric parameters of bone formation, and an increase in mean crystal thickness, suggesting low bone turnover. The architecture of the trabeculae as well as the lamellar arrangement of the collagen fibrils were highly disturbed, reflecting a lack of adaptation to mechanical demands contributing to the increased bone fragility.
As modeling and remodeling activities are high in the growing skeleton and change with age, we took advantage of recently published age-matched reference data and used them as a control group for our patients with pycnodysostosis. Normative histomorphometric data were available from iliac crest biopsies (16), but bone mineralization data were only available from vertebral spongiosa (17). Although this is a potential limitation of our study, it has been shown that the mineralization density distribution for trabecular bone is constant, independent of skeletal site or differences in mechanical loading conditions for a given individual (23).
The structural parameters measured with light microscopy and backscattered electron imaging showed that the degree of osteosclerosis was similar in our two patients. Evaluation of histomorphometric parameters revealed that bone formation was markedly decreased in both, whereas osteoclast parameters were rather high for patient A or mostly within the normal range for patient B. In addition, there were alterations of osteoclast morphology: the osteoclasts of patient B appeared flatter and the excavated areas shallower than those in patient A. A diminished osteoclast activity in patient B was also reflected by the decreased percentage of demineralized fringes on the total bone surface as well as per resorption lacunae.
However, it has to be emphasized that in pycnodysostosis osteoclasts are dysfunctional and resorb only the mineral, not the matrix. Therefore, histomorphometric evaluations do not really reflect bone resorption. The emerging pictures is, indeed, low bone formation in the child in the presence of a high number of dysfunctional osteoclasts and low bone formation in the presence of a decreased number of dysfunctional osteoclasts in the adult. Further information arises from our SAXS data. At the nanometer scale, the size of mineral particles embedded in the collagen matrix is known to increase with age (17) and with reduced bone turnover activity, reflecting different stages of bone maturity (30). Both patients with pycnodysostosis had markedly enlarged mineral crystals, indicating much older and less remodeled tissue than expected for the chronological age. From these data it can be assumed that both subjects had a marked slowdown of bone turnover.
The degradation of the matrix is a crucial step for promoting new bone apposition. It is widely accepted that bone formation and bone resorption are linked processes that maintain structural and mechanical properties of bone. Dysfunction of one step would induce an imbalance in bone turnover. Several mouse models have been developed in which ablation of genes responsible for osteoclastic development (e.g. c-Scr and c-fos, OPGL) resulted in an arrest in osteoclast differentiation or a loss of function (31, 32, 33). In animal models with enhanced bone formation (e.g. decreased c-Src expression and fra-1 overexpression), osteosclerosis worsens with time despite arrested or reduced bone resorption (34, 35). In contrast to cathepsin K deficiency, genetically induced osteoclastic abnormalities and their resulting bone phenotypes have not been described in humans. Moreover, it is remarkable that cathepsin K-deficient mice have milder osteosclerosis than patients with pycnodysostosis. This may be due to different collagen breakdown pathways in nonosteoclastic cells, for example, in mice and humans, as intracellular collagen accumulates in fibroblasts of pycnodysostotic patients, but not in cathepsin K-deficient mice (36). In addition, the results of this study suggest that the age of the patients and the mechanical load conditions might play important roles. In fact, none of the cathepsin K-deficient animals studied was older than 10 months (3, 4), a time period that may be too short to evaluate the long-term outcome of disturbed bone turnover. The situation may also differ between growing children and mice, where bone turnover is directed for modeling of the skeleton and increased bone mass, and adults, where maintenance of a constant amount of bone is the primary tool. Our results suggest that in pycnodysostosis the bone volume increases toward some natural limit due to dysfunctional osteoclastic activity. When a critical amount of bone mass is formed, osteoclastic and osteoblastic activities are slowed. In accordance with a previous study in which overexpression of cathepsin K in mice caused marked increases in the number of osteoblasts and in the rate of bone turnover (14), our findings underline the potential importance of cathepsin K in bone remodeling. Modulation of bone turnover might be related to the specific and unique collagen cleavage pattern of cathepsin K (2).
The abnormal bone turnover profoundly affected the bone matrix of both patients. The homogeneity of the matrix appeared severely disturbed by the inclusion of large residuals of mineralized cartilage. The most relevant finding for the mechanical properties of bone, however, was the highly disordered lamellar architecture. In normal bone and according to Wolff’s law (37), the architecture of the trabeculae follows the principal stress directions. In pycnodysostotic bone, we found a highly disordered organization, with lamellae that were neither smooth nor oriented parallel to the main axis of the trabeculae. The chaotic lamellar organization of the trabeculae might be a key element to understanding the bone fragility observed in pycnodysostotis. Consistently, the degree of alignment of mineral particles, which also reflects the orientation of collagen fibrils, appeared to be dramatically reduced.
Taken together, our results strongly suggest that the increased susceptibility to fractures observed in pycnodysostosis results from a deterioration of bone architecture that cannot properly adapt to mechanical loads, leading to a highly disordered lamellar structure. Cathepsin K seems to play a pivotal role in modulating bone turnover, and long-term deficiency seems to have severe repercussions on bone strength. This fact should be taken into account when cathepsin K antagonists are considered as therapeutic agents to treat metabolic bone diseases such as osteoporosis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address for P.F.: Department of Biomaterials, Max Planck Institute of Colloids and Interfaces, D-14424 Potsdam, Germany.
Abbreviations: qBEI, Quantitative backscattered electron imaging; SAXS, small angle x-ray scattering; SEM, scanning electron microscope.
Received June 19, 2003.
Accepted January 6, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-line intensities. Scanning Microsc 9:75–88[Medline]
This article has been cited by other articles:
 |
|
 |
|
  S. B. Rodan and L. T. Duong Cathepsin K - A New Molecular Target for Osteoporosis IBMS BoneKEy, January 1, 2008; 5(1): 16 - 24. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Chavassieux, E. Seeman, and P. D. Delmas Insights into Material and Structural Basis of Bone Fragility from Diseases Associated with Fractures: How Determinants of the Biomechanical Properties of Bone Are Compromised by Disease Endocr. Rev., April 1, 2007; 28(2): 151 - 164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Chen, S. Yang, Y. Abe, M. Li, Y. Wang, J. Shao, E. Li, and Y.-P. Li Novel pycnodysostosis mouse model uncovers cathepsin K function as a potential regulator of osteoclast apoptosis and senescence Hum. Mol. Genet., February 15, 2007; 16(4): 410 - 423. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Karsdal, K. Henriksen, M. G. Sorensen, J. Gram, S. Schaller, M. H. Dziegiel, A.-M. Heegaard, P. Christophersen, T. J. Martin, C. Christiansen, et al. Acidification of the Osteoclastic Resorption Compartment Provides Insight into the Coupling of Bone Formation to Bone Resorption Am. J. Pathol., February 1, 2005; 166(2): 467 - 476. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
