SEARCH

SEARCH BY CITATION

Keywords:

  • Bone resorption;
  • Cysteine protease;
  • Malignant osteolysis;
  • Surrogate marker

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Background

Cathepsin K (CatK) is a lysosomal protease with collagenolytic activity, and its secretion by osteoclasts is responsible for degrading organic bone matrix. People with pathologic bone resorption have higher circulating CatK concentrations.

Hypothesis

Canine osteosarcoma (OS) cells will possess CatK, and its secretion will be cytokine inducible. Circulating CatK concentrations will be increased in dogs with OS, and will be a surrogate marker of bone resorption.

Animals

Fifty-one dogs with appendicular OS and 18 age- and weight-matched healthy control dogs.

Methods

In a prospective study, expressions of CatK mRNA and protein were investigated in OS cells. The inducible secretion and proteolytic activity of CatK from OS cells was assessed in vitro. Serum CatK concentrations were quantified in normal dogs and dogs with OS and its utility as a bone resorption marker was evaluated in dogs with OS treated with palliative radiation and antiresorptive agents.

Results

Canine OS cells contain preformed CatK within cytoplasmic vesicles. In OS cells, TGFβ1 induced the secretion of CatK, which degraded bone-derived type I collagen in vitro. CatK concentrations were higher in dogs with OS than healthy dogs (11.3 ± 5.2 pmol/L versus 8.1 ± 5.0 pmol/L, = .03). In a subset of dogs with OS, pretreatment CatK concentrations gradually decreased after palliative radiation and antiresorptive treatment, from 9.3 ± 3.2 pmol/L to 5.0 ± 3.1 pmol/L, = .03.

Conclusions and Clinical Importance

Canine OS is associated with pathologic bone resorption, and CatK inhibitors might aid in the management of canine OS-related malignant osteolysis.

Abbreviations
CatK

cathepsin K

NTx

N-telopeptide

OS

osteosarcoma

Cathepsins are lysosomal proteases responsible for a wide range of diverse physiologic and pathophysiologic functions, including tumorigenesis.[1-3] The majority of cathepsins can be categorized as cysteine proteases and participate in key processes including extracellular matrix remodeling of bone.[4] Cathepsin K (CatK), the principal protease secreted by activated osteoclasts, possesses intense collagenolytic activity, cleaving collagens type I and III and subsequently influencing bone resorption.[5, 6] Given CatK's key role in bone resorption, CatK inhibitors are being clinically evaluated for managing disorders associated with dysregulated osteolysis.[7-10]

In addition to its osteoclast-associated activities, CatK has also been identified in several tumor types of skeletal and nonskeletal origin, including osteosarcoma (OS), chondrosarcoma, giant cell tumor, breast and prostate skeletal metastases, and melanoma.[11-16] Specifically for OS, CatK expression was highest in immortalized human OS cell lines with greatest metastatic phenotypes. In spontaneous OS tumor samples, lower CatK expression correlated with improved survival time in patients diagnosed with high-grade and metastatic OS, suggesting a participatory role of CatK in cancer metastases.[16] Although these findings suggest that CatK expression by malignant osteoblasts might contribute to tumor cell dissemination, the participatory role of OS-derived CatK in local tumor progression and malignant bone resorption requires further elucidation.

OS is the most common skeletal tumor affecting dogs, and is characterized by its high metastatic capacity and ability to cause aggressive malignant osteolysis within the bone microenvironment. Pathologic and dysregulated bone resorption induced by OS growth within the marrow cavity, cortical bone, and periosteum results in excruciating pain, and severely diminishes the quality of life of affected dogs. Despite bone-targeted treatment strategies that attenuate the degree and rate of malignant osteolysis secondary to OS progression, pain alleviation and maintenance of bone quality achieved in responsive dogs is not durable.[17-21] As such, novel therapies that could attenuate the extent, rate, and degree of focal malignant osteolysis should be investigated for improving the palliative management of dogs with OS.

Given CatK's pivotal involvement in pathologic bone resorption and the anticipated commercial availability of CatK inhibitors, the main purposes of this study were to characterize the expression and functionality of CatK in canine OS. As such, the study objectives of this investigation were multiple, including in vitro experiments (1) to determine if immortalized canine OS cells express CatK; (2) to determine the functional activity of preformed cellular CatK; and (3) to identify bone microenvironmental stimuli capable of inducing CatK secretion. In dogs diagnosed with OS, it was investigated (1) if naturally occurring OS samples express CatK; (2) if circulating CatK concentrations are increased in comparison with normal dogs; and (3) if CatK could serve as a reliable surrogate marker of bone resorption.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Overarching Study Design

In vitro studies were conducted to demonstrate the existence of CatK (mRNA and protein) in canine and human OS cell lines using qualitative polymerase chain reaction, western blot analysis, and immunofluorescent microscopy techniques. After confirmation of CatK expression, a 2nd series of in vitro experiments were performed to determine if CatK stored within OS cells possessed enzymatic activity, and if the secretion of CatK could be elicited by exogenous stimuli commonly associated with the bone microenvironment. Upon the identification of CatK enzymatic activity and active secretion, an in vitro assay was conducted to determine the ability of OS cell-derived CatK to resorb canine trabecular bone. Lastly, in dogs diagnosed with OS, studies were performed to characterize the potential biologic relevance and utility of CatK as a surrogate marker of bone resorption and therapeutic target.

Cell Lines

Five canine (POS, HMPOS, COS31, Abrams, and D17) and 2 human (HOS-MNNG and MG63) OS lines were evaluated for CatK expression. The POS and HMPOS cell lines were provided by James Farese, University of Florida; the COS31 line was provided by Ahmed Shoieb, University of Tennessee; and the Abrams line was provided by Doug Thamm, Colorado State University. All other cell lines were purchased commercially from American Tissue Culture Collection (Manassas, VA). All cell lines were cultured in either RPMI1640 or DMEM, with 10% fetal bovine serum and 1% penicillin/streptomycin. Cell cultures were maintained at 37°C in 5% CO2 and passaged twice weekly.

Reagents and Antibodies

Reagents were purchased to identify exogenous stimuli capable of inducing HMPOS cell secretion of preformed CatK, including human recombinant RANKL (ProSpec-Tany TechnoGene Ltd., East Brunswick, NJ), prostaglandin E2 (Cayman Chemical, Ann Arbor, MI), phorbol myristate acetate (Sigma-Aldrich, St. Louis, MO), human recombinant interleukin-1 (Sigma-Aldrich), human recombinant parathyroid hormone N-terminal peptide (ProSpec-Tany TechnoGene Ltd., East Brunswick, NJ), and human recombinant transforming growth factor β1 (hrTGFβ1, R&D Systems, Minneapolis, MN). Antibodies for western blot analysis and immunofluorescent microscopy studies included rabbit polyclonal anti-human CatK antibody (ab19027, Abcam, Cambridge, MA), mouse monoclonal anti-human Smad2/3 antibody (610842, BD Bioscience, San Jose, CA), rabbit monoclonal anti-human phospho-Smad2 antibody (04-953, Millipore, Billerica, MA), goat anti-rabbit Alexa488 (A-11008, Invitrogen, Carlsbad, CA), and TO-PRO-3 (Invitrogen, Carlsbad, CA).

Qualitative PCR

Total RNA was collected from OS cell lines using a commercial kit,1 and 1 μg of total RNA was reverse transcribed into cDNA (SuperScript Double-Stranded cDNA Synthesis Kit).2 Five microliters of reverse transcribed product was used as a template in a 50-μL polymerase chain reaction containing 100 ng of each oligonucleotide, 2.5 U of AmpliTaqGold, and degenerative forward (5′-CAC TTG GCT GAC ATG ACC AG-3′) and reverse (5′-CCC ACA TAT GGG TAG GCA TC-3′) primers for the detection of canine and human CatK transcripts. Forward (5′-GGA AAT CCC ATC ACC ATC TTC CA-3′) and reverse (5′-CAT CAC GCC ACA GTT TCC CGG AG-3′) primers for GAPDH were included as an internal control. Reactions were performed in a PTC-200 Peltier thermal cycler with the following conditions: 5 minutes at 94°C denaturing step followed by 34 cycles (60 seconds at 94°C, 90 seconds at 55°C, and 60 seconds at 72°C), and concluded by 72°C for 10 minutes. Positive amplification of CatK mRNA was considered when a solitary amplicon of approximately 386 base pairs was identified. Confirmatory qualitative PCR results were derived from 3 independent experiments.

Western Blot Analysis

Cellular proteins from OS cell lines were extracted with a commercial reagent,3 and concentrations were quantified using a standard assay kit (Bicinchoninic Acid Protein Assay, Rockford, IL). For each OS cell line, 50 μg of protein were separated by SDS-PAGE using 10% polyacrylamide gels and transferred onto a nitrocellulose membrane. After 1 hour blocking at room temperature in 5% nonfat dry milk PBS-Tween, the membrane was incubated with a rabbit polyclonal anti-human CatK antibody (1 : 1000) in 5% nonfat dry milk PBS-Tween overnight at 4°C. The membrane was washed 3 times and subsequently incubated for 1 hour at room temperature with a horseradish peroxidase conjugated anti-rabbit secondary antibody (1 : 2000) in 5% nonfat dry milk PBS-Tween and developed using a standard chemiluminence detection kit.4 Positive CatK protein identification was based upon the detection of a band approximately 37 kDa in molecular weight and optimized conditions, and results were derived from 5 independent experiments.

For the detection of Smad2/3 and phospho-Smad2, the HMPOS cell line was grown to 80% confluence, and then serum starved for 12 hours. Cells were incubated with exogenous human recombinant TGFβ1 (2 ng/mL) for 4 hours. Cell lysates (50 μg) were separated by SDS-PAGE using 10% polyacrylamide gels and transferred onto a nitrocellulose membrane. After 1 hour blocking at room temperature in 5% nonfat dry milk PBS-Tween, the membrane was incubated with either a mouse monoclonal anti-human Smad2/3 antibody (1 : 500) or a rabbit monoclonal anti-human phospho-Smad2 antibody (1 : 1000,) in 5% nonfat dry milk PBS-Tween overnight at 4°C. Subsequently membranes were washed and incubated for 1 hour at room temperature with a horseradish peroxidase conjugated anti-mouse secondary antibody (1 : 1000; for Smad2/3) or anti-rabbit secondary antibody (1 : 1000; for phospho-Smad2) in 5% nonfat dry milk PBS-Tween, and developed using a standard chemiluminence detection kit.4 Positive Smad2/3 and phospho-Smad2 protein detection was based upon the identification of 2 bands approximately 55–60 kDa in molecular weight, and optimized conditions and results were derived from 2 independent experiments.

Immunofluorescent Microscopy

The MG63 and HMPOS cell lines were grown for 48 hours in glass chamber slides at a concentration of 105 cells per well. Slides were rinsed briefly with chilled PBS, fixed in 4% paraformaldehyde for 15 minutes, and then washed twice with PBS. Adhered cells were permeabilized with 0.25% Triton X-100 in 1X PBS for 10 minutes at room temperature, and then rinsed 3 times in 1X PBS. Nonspecific protein binding was blocked with 1% BSA/10% FCS/0.3 M glycine for 1 hour at room temperature. Slides were incubated with rabbit polyclonal anti-human CatK antibody (1 : 500) in 1% BSA PBS-Tween overnight at 4°C, then rinsed 3 times in 1X PBS. Slides were incubated with goat anti-rabbit Alexa488 antibody (1 : 1000) in the dark for 90 minutes. Cell nuclei were stained with TO-PRO3 (0.25 μM) for 15 minutes and protected from light. Slides were rinsed with 1X PBS, cover slipped using Fluoro-Gel mounting medium (Electron Microscopy Sciences, Hatfield, PA), and imaged using a multiphoton confocal microscope (Zeiss LSM 710 NLO, Thornwood, NY). Staining for CatK was considered positive when discrete vesicular subcellular localization was identified in both human and canine OS cell lines, and results were derived from 2 independent experiments.

Cathepsin K Activity Assay

Enzymatic activity of preformed CatK propeptide was quantified in the POS, HMPOS, HOS-MNNG, and MG63 OS cell lines. Quadruplicate 106 cell aliquots were collected by centrifugation and lysed in 50 μL of cold cell lysis buffer following kit instructions (Cathepsin K Activity Assay).5 Cell lysates were incubated on ice for 10 minutes followed by centrifugation at 80 × g for 5 minutes. Lysate supernatants were transferred to a 96-well plate followed by the addition of 50 μL of reaction buffer. All samples were preincubated at room temperature for 30 minutes with and without a supplied CatK inhibitor, followed by the addition of a cleavable fluorogenic CatK substrate at 37°C for 1 hour. Fluorescence was read on a FLUOstar Optima plate reader (BMG Labtech, Cary, NC) at 400 nm excitation and 505 nm emission. Documentation of CatK activity was confirmed from 4 independent experiments.

Exogenous Stimuli Induced Secretion of Preformed Cathepsin K Propeptide

HMPOS cells were harvested and seeded in triplicate at a density of 105 cells per mL of complete media, and allowed to adhere in a 24-well plate. After 24 hours, fresh media with 2% fetal calf serum alone (control) or with the addition of RANKL (50 ng/mL), PGE2 (1 μM), PMA (6.2 μg/mL), IL-1 (10 ng/mL), PTH (400 ng/mL), or TGFβ1 (2 ng/mL) was added, and cells were allowed to incubate for an additional 30 hours. Conditioned cell culture supernatants were then collected and the secretion of preformed CatK was quantified in triplicate using a CatK ELISA kit7 following manufacturer's instructions. The induced secretion of preformed CatK from the HMPOS cell line was supported using colorimetric methods from 2 independent experiments.

In Vitro Bone Resorption Assay

Cancellous bone derived from the proximal humeri of fresh cadaver dogs was harvested and subjected to 3 freeze thaw cycles in liquid nitrogen to eliminate any viable resident stromal cells. Cancellous bone was pulverized in the presence of liquid nitrogen to powder consistency using a SPEX SamplePrep Freezer/Mill (SPEX CertiPrep, Metuchen, NJ). Pulverized bone was mixed with complete RPMI 1640 media to make a 10% bone matrix solution. In a 96-well plate, 250 μL of 10% bone matrix solution was aliquoted per well, bone mineral contents were allowed to settle to the bottom, and residual culture media was aspirated, leaving only a thin film of bone material covering the bottom of each well.

In 24-well plates, 106 HMPOS cells in 1 mL of complete media were allowed to adhere overnight. After adherence, cell culture supernatants were discarded, and HMPOS cells were then incubated in RPMI 1640 and 1% fetal calf serum with or without TGFβ1 (2 ng/mL) for an additional 30 hours. Subsequently, conditioned cell culture supernatants containing secreted preformed CatK propeptide were collected and processed with either 80 μL PBS (nonactivated control) or an equivalent volume of activation buffer (200 mM sodium acetate and 20 mM l-cysteine, at a pH of 4.5 for 10 minutes at 50°C) in accordance with a previously described method.[22] After processing and activation, 250 μL of conditioned cell culture supernatant was then added on top of the prelayered, 96-well, bone matrix plate for 72 hours at 37°C. At the completion of 72 hours of incubation, conditioned cell culture supernatants were collected and assayed for evidence of bone resorptive activities by measuring N-telopeptide (NTx) concentrations using a commercial kit.6 In vitro bone resorptive assays utilized a colorimetric sandwich ELISA for NTx, and results were derived from 2 independent experiments.

Spontaneous Canine OS Histomorphology

Archived tissue of 5 spontaneously arising appendicular OS tumors were histologically reviewed to assess spatial relationships between malignant osteoblasts and areas of resorbing bone matrix. Histologic specimens were evaluated and described by a board certified anatomic pathologist (LBB).

CatK Immunocytochemistry Antibody Validation

The human OS cell line, MG63, was used as a positive control for antibody validation and staining protocol optimization. Briefly, cytospin preparations were incubated in acetone for 10 minutes, allowed to air dry, and loaded onto a DAKO autostainer. To minimize nonspecific peroxidase background staining, all preparations were blocked with 10% hydrogen peroxide for 10 minutes. Cytospin preparations were incubated with a rabbit polyclonal anti-human CatK antibody (1 : 100, 30 minutes), and then incubated with a secondary goat anti-rabbit immunoglobin for 30 minutes. Subsequently, slides were incubated with 3,3′-diaminobenzidine chromagen solution (7 minutes), and then counterstained with Mayer's hematoxylin (5 minutes). Positive CatK propeptide expression was considered when >75% of malignant osteoblasts demonstrated diffuse or punctate cytoplasmic staining. Immunocytochemical preparations were evaluated and reviewed by a board certified clinical pathologist (AMB).

CatK Immunocytochemistry of OS Samples

Fine-needle aspirate cytology of primary OS lesions were collected prospectively in 11 dogs with confirmed OS. Aspirates were allowed to air dry on slides, and processed using the same immunocytochemical staining protocol for the CatK antibody validation. OS samples were considered positive for CatK propeptide when at least 75% of malignant osteoblasts demonstrated diffuse or punctate cytoplasmic staining. When present, reactive osteoclasts collected during bone fine-needle aspirates were used as an internal positive control for CatK propeptide expression, and negative control slides were processed in the absence of primary antibody. At least 1 highly cellular sample (>100 cells per slide) was assessed from each patient, and immunocytochemical preparations were evaluated and described by a board certified clinical pathologist (AMB).

Serum CatK Propeptide Concentrations

Circulating CatK propeptide concentrations were quantified in archived serum samples collected from 40 dogs with confirmed OS and in 18 age- and weight-matched control dogs using a commercial kit.7 Using archived serum samples derived from 10 dogs with OS effectively treated with palliative radiation, doxorubicin, and intravenous pamidronate,[17] serial changes in circulating CatK propeptide were characterized every 28 days for 112 days using an ELISA kit.7 In the same 10 dogs with OS, serial changes in urine NTx excretion were determined using an ELISA kit.6 Serum samples were run in duplicate, and percent coefficient of variation was <20% within individual samples.

Statistical Analysis

Data sets were assessed for normality using the Shapiro-Wilk test. Differences in CatK activity pre- and postincubation with CatK inhibitor were analyzed using a paired Student's t-test. Differences in CatK propeptide secretion induced by various in vitro stimuli were analyzed using 1-ANOVA and posthoc Dunnet's test. The effect of TGFβ1 on CatK propeptide secretion and subsequent in vitro bone resorptive capacity was assessed using 1-ANOVA and posthoc Tukey test. Differences in basal circulating CatK propeptide concentrations in dogs with OS and normal control dogs were analyzed using Student's t-test, and serial changes in circulating CatK propeptide and urine NTx in dogs with OS receiving palliative treatment was analyzed using 1-ANOVA and posthoc Dunnet's test.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Expression of CatK in OS Cell Lines

Transcription of the CatK gene was identified in all OS cell lines (Fig 1A), and its translation into protein, as an inactive propeptide of ~37 kDa, was supported by western blot analysis (Fig 1B). Through the use of confocal fluorescent microscopy, the subcellular localization of CatK propeptide was characterized in the MG63 and HMPOS OS cell lines. In both OS cell lines, CatK propeptide appeared to be stored as punctate granules within the cytosolic compartment, consistent with lysosomal subcellular localization (Fig 1C).

image

Figure 1. Characterizing cathepsin K (CatK) expression in canine (POS, HMPOS, COS31, Abrams, and D17; lanes 1–5, respectively) and human (HOS-MNNG and MG63; lanes 6-7, respectively) OS cell lines by (A) qualitative PCR using degenerative primer sets and (B) western blot analysis. Using fluorescent microscopy (C), preformed CatK propeptide (red) is identified in MG63 (left) and HMPOS (right) cell lines as discrete punctate cytosolic staining consistent with lysosomal vesicle subcellular localization. Cell nuclei are stained blue.

Download figure to PowerPoint

Enzymatic Activity of CatK Propeptide in OS Cell Lines

In 2 canine (POS and HMPOS) and 2 human (HOS-MNNG and MG63) OS cell lines, the proteolytic activity of CatK propeptide was assessed using a fluorometric assay. After incubation with activation buffer that cleaves the pro-signal sequence of CatK, all 4 OS cell lines had measurable CatK enzymatic activity, although the magnitude of proteolytic activities varied. In an order of decreasing CatK activity was MG63 (55,018 ± 4,450 RFU), POS (26,344 ± 5,631 RFU), HMPOS (17,305 ± 1,406 RFU), and HOS-MNNG (12,311 ± 3,649 RFU). Proteolytic activity of the OS cell lines was confirmed to be CatK specific, because the addition of a selective CatK inhibitor significantly reduced cleaved substrate fluorescence compared with uninhibited controls, < .001 for all cell lines (Fig 2A).

image

Figure 2. Demonstration that cathepsin K (CatK) vesicles within canine and human OS cells can be actively secreted and exert proteolytic activities in vitro and ex vivo (A) the enzymatic activity of preformed CatK propeptide is assessed in 4 OS cell lines, in the absence (−) or presence (+) of a specific CatK inhibitor. Complete media with no OS cells serves as a background control. In the HMPOS cell line, (B) the ability of various exogenous stimuli to enhance the secretion of preformed CatK propeptide into cell culture supernatants is evaluated. Phosphorylation of Smad2 (C) supports the notion that human recombinant TGFβ1 is capable of inducing intracellular signaling in the canine HMPOS cell line. (D) In vitro bone resorptive capacity of conditioned media (unactivated and activated) from HMPOS cell line after incubation with TGFβ1 is assessed through the detection of serum N-telopeptide. **< .001 and *< .01.

Download figure to PowerPoint

Inducible Secretion of Preformed CatK Propeptide

Using the canine HMPOS cell line, hrTGFβ1 was the only stimuli that consistently, although marginally, increased CatK propeptide secretion in comparison to control media. In control media, the basal secretion of CatK propeptide was 2.9 ± 0.6 pmol/L, whereas the addition of hrTGFβ1 increased CatK propeptide concentrations to 8.7 ± 1.9 pmol/L, = .01 (Fig 2B). In contrast, none of the other various inflammatory stimuli (prostaglandin E2 and interleukin-1), bone-associated cytokines receptor activator of nuclear factor kappa-B ligand and parathyroid hormone, and protein kinase C activating agents (phorbol myristate acetate) promoted the secretion of preformed CatK propeptide from their cytosolic storage compartments. The ability of hrTGFβ1 to induce signaling through canine TGFβ receptors (TGFβRI and TGFβRII) was confirmed using western blot analysis of Smad2/3 and phospho-Smad2 (Fig 2C). Incubation of HMPOS cells with hrTGFβ1 (2 ng/mL) for 4 hours resulted in the phosphorylation of Smad2.

In vitro Bone Resorptive Capacity of CatK Propeptide Secretion

Through the use of an in vitro bone resorption assay, the production of NTx from pulverized trabecular bone was quantified after 72-hour incubation with 3 different HMPOS-derived supernatant conditions. Unprocessed HMPOS conditioned media, containing basally secreted CatK propeptide, resulted in relatively low concentrations of NTx production (67.1 ± 6.3 nM/BCE). When HMPOS conditioned media was subjected to activating conditions, basally secreted CatK propeptide was enzymatically activated, resulting in a dramatic increase in NTx production (410.7 ± 27.3 nM/BCE), < .001. Furthermore, media derived from HMPOS cells grown in the presence of rhTGFβ1 and subsequently exposed to activating conditions resulted in the greatest production of NTx (518.2 ± 51.6 nM/BCE) (Fig 2D).

Spontaneous Canine OS Histomorphology and CatK Propeptide Expression

Prospectively, in 11 dogs with confirmed OS, fine-needle aspirate cytology of the primary OS lesion was collected and evaluated for CatK propeptide expression. Nine of 11 OS samples demonstrated strong and uniform staining of malignant osteoblasts for CatK propeptide (Fig 3A,B). Activated osteoclasts derived from the bone tumor microenvironment were also strongly positive for CatK propeptide (Fig 3C). Five archived OS samples had clusters of malignant osteoblasts, readily identified as invading into discrete islands of bone, resulting in a “scalloped” resorptive pattern (Fig 3D,E).

image

Figure 3. Cytology from a spontaneously arising canine osteosarcoma (OS) tumor sample (A) demonstrates that malignant osteoblasts uniformly staining for preformed cathepsin K (CatK) propeptide (50X magnification). High magnification (100X) cytology of a small cluster of malignant osteoblasts (B) demonstrating vesicle subcellular localization of preformed CatK propeptide. High magnification (100X) cytology of an activated osteoclast (C) with intense and diffuse cytosolic staining for preformed CatK propeptide. Two representative histologic canine OS samples (D and E) showing close spatial relationships between malignant osteoblasts and fragments of resorbing bone. Note that malignant osteoblasts appear to invade (D, black arrowheads) into or form scalloped indentations (E, black arrowheads) into bone. Mature osteocytes within normal bone can be readily identified (white arrowheads). Hematoxylin and eosin stain 100X magnification.

Download figure to PowerPoint

Serum CatK Propeptide Concentrations in Dogs with OS

Dogs with OS (11.3 ± 5.2 pmol/L; range 3.6–25.3 pmol/L) had significantly higher concentrations of CatK propeptide in comparison with healthy control dogs (8.1 ± 5.0 pmol/L; range 1.3–16.2 pmol/L; = .03) (Fig 4A). In a subset of dogs with OS (n = 10), before initiating palliative treatment, baseline CatK propeptide concentrations on Day 0 were 9.3 ± 3.2 pmol/L, and concentrations gradually decreased with the continuation of treatment. After several consecutive treatments with palliative treatment, circulating concentrations of CatK propeptide achieved significant reduction on Day 84 (5.6 ± 2.5, = .01) and Day 112 (5.0 ± 3.1, = .03) in comparison to pretreatment values (Fig 4B). Although circulating CatK propeptide concentrations gradually decreased as a function of treatment duration and clinical pain alleviation, the changes in CatK propeptide were not dynamic or rapid in onset in comparison with urine NTx excretion in the same 10 dogs with OS (Fig 4C).

image

Figure 4. Baseline (A) circulating concentrations of cathepsin K (CatK) propeptide in dogs with osteosarcoma (OS) (n = 40) and normal control dogs (n = 18) from archived serum samples are significantly different. Circulating CatK propeptide concentrations (B) in 10 dogs with OS gradually decrease after effective treatment with standardized palliative treatment. CatK concentrations at Days 84 and 112 are significantly reduced in comparison with pretreatment values. In the same 10 dogs with OS achieving excellent clinical response from palliative treatment, (C) urine N-telopeptide excretion demonstrates a much more rapid and dynamic change upon the onset of treatment.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

One of the major findings in this study was the demonstration that both canine and human OS cell lines possess preformed CatK propeptide within defined cytosolic compartments, and upon release and activation, CatK exerted proteolytic activity. Given that CatK's natural substrate is type I collagen derived from organic bone matrix, and the findings from this investigation suggest that malignant osteoblasts have the potential to directly participate in pathologic bone resorption. Another discovery from this investigation was the identification of TGFβ1 as a cytokine trigger that promotes the secretion of CatK propeptide from canine OS cells. Because TGFβ1 is predominantly stored within bone matrix, its liberation as a consequence of OS-associated bone destruction may be conducive to sustaining malignant osteolysis by promoting CatK secretion by OS cells. Lastly, this research study provides preliminary evidence that circulating CatK propeptide is increased in dogs diagnosed with focal malignant osteolysis, and CatK may serve as a surrogate marker of pathologic bone destruction.

From this study, it was an interesting, but not a completely unexpected finding that malignant osteoblasts possessed preformed CatK. Prior research reports have demonstrated that nontransformed human osteoblasts derived from pathologic bone diseases as well as malignant osteoblasts derived from naturally occurring human OS express CatK protein.[16, 23] Although normal osteoblasts are classically responsible for new bone formation through the deposition of osteoid matrix, limited evidence exists which supports their direct participation in the final stages of bone resorption.[24, 25] In nontransformed osteoblasts, the exact proteolytic mechanisms utilized for bone resorption remain to be thoroughly characterized; however, matrix metalloproteinases and CatK are probably involved.[23-25] Given the severity of skeletal pathology associated with OS, it is possible that malignant osteoblasts may also participate directly in bone resorption, but perhaps in a dysregulated manner.

The process of bone resorption is complex, and requires a series of coordinated cellular events. Despite the identification of preformed CatK within malignant osteoblasts, possession of requisite enzymatic machinery such as CatK does not absolutely obligate OS cells to directly participate in pathologic osteolysis. However, at least 2 potential operative mechanisms for OS cells to directly participate in bone resorption may be proposed. First, given existing research that demonstrates the capacity of normal osteoblasts to phagocytize bone,[26, 27] OS cells might directly resorb organic bone matrix by endocytosis of minute bone fragments as lysosomal vesicles. Subsequently, acidic phagolysosomes formed through the fusion of preformed CatK propeptide and bone fragment containing vesicles would activate CatK and result in type I collagen degradation. Alternatively and similar to osteoclast-mediate bone resorption, malignant OS cells may co-secrete protons and preformed CatK propeptide within close proximity to the bone surface, resulting in the extracellular degradation of collagen type I. Coupled with evidence that normal osteoblasts can secrete protons,[24] and the findings from this study which demonstrates the secretion of CatK propeptide by malignant OS cells, this mechanism of OS-induced bone resorption remains plausible. Regardless of the exact mechanisms, observations from this study demonstrate that a close spatial relationship exists between OS cells and resorbing bone fragments, and that OS cells contain one of the pivotal enzymatic proteins involved in bone degradation.

Transforming growth factor β1 is a pleiotropic cytokine that is predominantly stored within bone matrix.[28, 29] During skeletal metastases in people, carcinoma cells subvert and potentiate osteoclastic activities, liberating TGFβ1 from bone. The subsequent release of TGFβ1 promotes the growth and survival of carcinoma cells, and results in the establishment of a “vicious cycle.”[30, 31] In the current study, TGFβ1 potentiated the secretion of CatK propeptide from canine OS cells, which upon activation, was able to degrade type I collagen in vitro. Based upon these study findings, it is plausible that an analogous “vicious cycle” may be established between malignant OS cells and progressive bone resorption, whereby OS cells sustain and promote their capacity for further malignant osteolysis as a consequence of TGFβ1 liberation. If such a paracrine loop exists, TGFβ inhibition may attenuate OS-related bone destruction.

In dogs with OS, a close spatial relationship between malignant OS cells and fragments of resorbing bone was identified histologically. In addition to the observed spatial proximity of OS cells and areas of bone resorption, the majority of spontaneous canine OS samples (9/11) were composed of malignant osteoblasts possessing CatK propeptide. Based upon these collective in vivo observations, it is possible that canine OS cells could actively participate in malignant bone resorption, and skeletal pathology associated with canine OS may be the combined resorptive capacities of not only activated osteoclasts but also malignant osteoblasts.

Similar to people diagnosed with pathologic bone resorptive conditions,[32-35] dogs with focal malignant osteolysis had significantly higher concentrations of circulating CatK propeptide in comparison with healthy control animals, suggesting that CatK could be a surrogate marker of bone resorption in companion animals. To further support the potential role of CatK as a biomarker of osteolysis, dogs with OS effectively treated with palliative therapies demonstrated gradual reductions in circulating CatK propeptide concentrations as a function of time and clinical disease control. The observed CatK reductions in dogs with OS treated effectively with palliative treatment strategies is congruent with reductions in serum CatK documented in people treated for dysregulated bone resorptive disorders with aminobisphosphonates.[33, 34] Although the findings from this study suggest the potential usefulness of CatK as a resorptive marker in companion animals, the sensitivity and dynamic range of CatK appeared inferior in comparison to urine NTx, a more conventional surrogate marker of bone resorption previously evaluated in dogs with OS.[36]

Focal malignant osteolysis associated with canine OS remains a clinical problem, because progressive bone destruction results in intense pain and poor quality of life for affected dogs. As such, the identification of adjunctive treatment strategies capable for slowing or reversing the process of pathologic bone destruction is a high clinical priority for pet owners and veterinarians alike. Based upon the results of this study, it would appear that CatK may be a druggable target for improving the palliative management of canine OS. Given CatK's pivotal role in bone resorption, irrespective of its cellular origin (osteoclast or malignant osteoblast), it is reasonable to believe that CatK inhibitors may be a justified adjuvant treatment option for dogs with OS. Already in human phase III clinical trials, odanacatib, a selective CatK inhibitor, has demonstrated biologic effectiveness for managing pathologic bone disorders, including osteoporosis and cancer-induced bone resorption.[7, 8] Given the favorable results with odanacatib in people, the current study demonstrating the putative role of CatK in canine OS is clinically significant and provides a rational basis to investigate similar agents in dogs with OS.

Although many novel findings were generated from this investigation, there are several limitations that should be acknowledged. First and foremost, the experimental design utilized in this investigation does not permit any definitive conclusions to be drawn regarding the direct role of OS cells in the process of bone resorption. Rather, the data and observations garnered only provide plausible evidence for a participatory role of OS-derived CatK in the process of malignant bone destruction. Second, the in vitro methods used to activate CatK propeptide were artificial, and do not recapitulate the operative mechanisms employed by osteoclasts or osteoblasts for activating CatK in vivo. In addition, the experiments identifying TGFβ1 as a cytokine capable of inducing CatK secretion by canine OS cells were contrived, and were not likely representative of the conditions encountered within the bone microenvironment of a living animal. Similarly, many of the evaluated cytokines used in the CatK release studies were human recombinant proteins (RANKL, IL-1, PTH), and the failure to elicit any biologic consequence could have been because of differences in cross-species protein homology, rather than a true absence of effect. Third, for the in vitro bone resorption assay, the increased production of NTx cannot be definitively associated with CatK propeptide activation, because it remains possible that other OS-associated collagenolytic proteases, such as MMP-1, -8, and -13, may have contributed to type I collagen degradation. Finally, although CatK concentrations are increased in dogs with OS, it was not possible to delineate the cellular source of CatK, and the contribution of malignant osteoblast-derived CatK in focal bone destruction in dogs with OS could not be assessed.

Despite these limitations, this investigation provides new information regarding the potential role of CatK in dogs with OS. Of greatest clinical interest is the documented CatK increases in dogs with OS, which may serve as a druggable target. Given the anticipated commercial availability of CatK inhibitors, like odanacatib, future clinical trials evaluating the biodistribution, pharmacokinetics/pharmacodynamics, and ultimately clinical efficacy of CatK inhibitors would be warranted for companion animals suffering from pathologic bone resorptive disorders.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The authors thank the technicians and veterinary oncology residents of the University of Illinois Cancer Care Clinic for their contributions in this study.

Footnotes
  1. 1

    RNeasy mini kit, Qiagen, Valencia, CA

  2. 2

    Invitrogen, Carlsbad, CA

  3. 3

    M-PER, Pierce, Rockford, IL

  4. 4

    ECL Kit, GE Healthcare Life Sciences, Piscataway, NJ

  5. 5

    BioVision Research Products, Mountain View, CA

  6. 6

    Cathepsin K ELISA, Biomedica Gruppe, Vienna, Austria

  7. 7

    Osteomark, Wampole Laboratories, Princeton, NJ

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References