Article Text

Concise report
Decreased levels of nucleotide pyrophosphatase phosphodiesterase 1 are associated with cartilage calcification in osteoarthritis and trigger osteoarthritic changes in mice
  1. J Bertrand1,
  2. Y Nitschke2,
  3. M Fuerst3,
  4. S Hermann4,
  5. M Schäfers4,5,
  6. J Sherwood1,
  7. G Nalesso1,
  8. W Ruether6,
  9. F Rutsch2,
  10. F Dell'Accio1,
  11. T Pap7
  1. 1Department of Experimental Medicine and Rheumatology, Barts and The London, Queen Mary School of Medicine and Dentistry, William Harvey Research Institute, London, UK
  2. 2Department of General Pediatrics, Münster University Children's Hospital, Münster, Germany
  3. 3Department for Orthopedics and Trauma Surgery, MedBaltic, Neumünster, Germany
  4. 4European Institute for Molecular Imaging, University of Münster, Germany
  5. 5Department of Nuclear Medicine, University Hospital Münster, Münster, Germany
  6. 6Centre for Orthopaedics and Trauma Surgery Bad Bramstedt, University hospital Hamburg-Eppendorf, Germany
  7. 7Institute for Experimental Musculoskeletal Medicine, University Hospital Münster, Münster, Germany
  1. Correspondence to Jessica Bertrand, Institute of Experimental Musculoskeletal Medicine, University Hospital Münster, Domagkstraße 3, D-48149 Münster, Germany; bertrand{at}


Objective To analyse the function of nucleotide pyrophosphatase phosphodiesterase (NPP1), a member of the pyrophosphate pathway, in osteoarthritis (OA).

Methods mRNA expression of NPP1, ANK ankylosing protein and tissue non-specific alkaline phosphatase was assessed by quantitative PCR. NPP1 protein levels were analysed in mouse and human cartilage samples. Bone metabolism was analysed by F18-positron emission tomography-scanning and µCT in ttw/ttw mice. Ttw/ttw mice are mice carrying a loss-of-function mutation in NPP1. Calcification of articular cartilage was assessed using von Kossa staining and OA severity using the Mankin score. Cartilage remodelling was investigated by type X collagen immunohistochemistry.

Results Expression of NPP1, but not the other members of this pathway, inversely correlated with cartilage calcification and OA severity in mouse and humans. Proinflammatory cytokines downregulated the expression of NPP1, demonstrating an influence of inflammation on matrix calcification. Ttw/ttw mutant mice, carrying a loss-of-function mutation in NPP1, exhibit increased bone formation process in joints compared with wild types. Ttw/ttw mice also developed spontaneous OA-like changes, evaluated by histological analysis and in vivo imaging. Ectopic calcifications were associated with increased expression of collagen X in the cartilage.

Conclusion The authors conclude that OA is characterised by the reactivation of molecular signalling cascades involving proinflammatory cytokines, thereby regulating the pyrophosphate pathway which consequently leads to cartilage ossification, at least in part resembling endochondral ossification.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Matrix calcification occurs physiologically during the longitudinal growth of bones through endochondral ossification at the growth plate. Under pathophysiological conditions, such as atherosclerosis or osteoarthritis (OA), matrix calcification is restarted for unknown reasons.1 Our previous work showed that in OA, ectoptic calcification of cartilage with basic calcium phosphate (BCP) is strongly associated with the hypertrophic differentiation of chondrocytes.2 Interestingly, alterations of the inorganic pyrophosphate (PPi) pathway have been associated with the pathological calcification of extracellular matrix.3 4 Thus, high extracellular inorganic phosphate (Pi) levels lead to the accumulation of BCP crystals and prevent the generation of calcium pyrophosphate dehydrate crystals (CPPD). Under physiological conditions, PPi potently suppresses BCP crystal deposition and propagation.5 Three molecules have been identified as central regulators of PPi metabolism: the tissue non-specific alkaline phosphatase (TNAP), which hydrolyses PPi,6 the multiple-pass transmembrane protein (ANK) ankylosing protein, which mediates intracellular to extracellular channelling of PPi,7 and the nucleotide pyrophosphatase phosphodiesterase (NPP1), which generates PPi from nucleoside triphosphates.8 The activity of NPP1 is a major component of chondrocyte PPi generation by cleavage of ATP.9 Interestingly, it has been shown that NPP1 is downregulated in calcified plaques in atherosclerosis,10 and mutations in the NPP1 gene are correlated with hand OA, and being the main gene related with generalised arterial calcification of infancy.11 In the present study, we have analysed the expression of NPP1 in OA cartilage and used NPP1-mutant mice to study its effects upon the development of OA-like changes.



One hundred and twenty patients with primary end-stage OA undergoing total knee arthroplasty at Rheumaklinik Bad Bramstedt were prospectively included in this study. The study was approved by the ethics committee of the Ärztekammer Schleswig-Holstein, Bad Segeberg, Germany. All patients in the study gave full written informed consent for participation prior to the operative procedure.

Isolation of RNA and quantitative real-time PCR

Total RNA was extracted from C28/I2 cells stimulated for 24 h with 10 ng/ml IL-1 β or 100 ng/ml tumour necrosis factor α (TNFα) R&D (Wiesbaden, Germany), Invitrogen (Karlsruhe, Germany), QIAGEN (Hilden, Germany), Sigma Aldrich (Hamburg, Germany). Five hundred ng of total RNA from each sample was reverse transcribed using Thermoscript Reverse transcription kit (Invitrogen). Quantitative PCR was performed with hot-start DNA polymerase (QIAGEN) in the presence of 0.1×SYBR green (Sigma-Aldrich) and 0.2×ROX dye (Invitrogen). NPP1 primer (forward GGG TTC CTC TCC CCA CCACAA CTA and reverse CCATGCACACAGCTCTCGCTG T) β-actin primer (forward CACGGCTGCTTCCAGCTC and reverse CACAGGACTCCATGCCCAG) were used or detection was performed using the following predesigned TaqMan gene expression assays, following the manufacturer's guidelines: Hs99999905_m1 GAPDH, Hs01054040_m1 NPP1, Hs00986657_m1 ANKH and Hs01029144_m1 TNAP.

Animal experiments

NPP1-mutant ttw/ttw mouse has been described by Okawa et al.12 Induction of OA was performed as described by Glasson et al.13 Joints of wild type and ttw/ttw mice were harvested at the age of 21 weeks and frontal sections were taken through the entire joint.

Histological analysis

Cartilage sample preparation and scoring was performed as described2 using the Mankin score.14 Von Kossa stainings of knee sections and human cartilage samples were performed to assess the calcification. Quantification of mineralisation was performed using digital contact radiography as published previously.2 Paraffin sections were analysed after pepsin digestion with primary antibodies directed against human/mouse ENPP1 (PAB6879, Abnova Heidelberg, Germany) and collagen X (ab58632, Abcam Cambridge, UK) using the Vectastain ABC Elite Kit (Vector Laboratories Loerrach, Germany).


Animals were anaesthetised with isoflurane (1.8%) and placed on a heating pad to keep body temperature stable. F18-sodium fluoride 9.6–10.4 MBq was injected intravenously 1 h prior to positron emission tomography (PET). PET list mode data were acquired for 15 min using the 32-module quadHIDAC scanner (Oxford Positron Systems Oxfordshire, UK). The effective resolution is 0.7 mm (full width at half maximum) in transaxial and axial directions using an iterative resolution recovery reconstruction algorithm. PET data were reconstructed into a single image volume for each mouse (voxel size 0.4×0.4×0.4 mm3). X-ray CT was performed with a spatial resolution of 15 µm using small animal CT (Inveon).

Statistical analysis

All data were expressed as mean±SEM. The correlation between continuous numerical data was analysed by Spearman's correlation coefficient. The Student's t-test was used to determine significance in a dependent pairwise comparison. A p value less than 0.05 was considered significant.


NPP1 expression is downregulated in human OA

Analysing the severity of OA in human patients, we found lower NPP1 protein levels in cartilage with increased OA severity and matrix mineralisation (figure 1A). The quantification of these staining revealed an inverse relationship between cartilage calcification and NPP1 protein (figure 1B). At mRNA level, we also found that NPP1 inversely correlated (p=0.048) with calcification, whereas the expression of ANK (p=0.4) and TNAP (0.31) mRNA did not show such correlation (figure 1C). Based on the data of Lotz et al,15 we next analysed the effects of the proinflammatory cytokines IL-1β and TNFα on the expression of NPP1 and found a 50% downregulation of NPP1 by both cytokines (figure 1D).

Figure 1

(A) Staining of human cartilage samples with Mankin score 1 (0–4), Mankin score 2 (5–9) and Mankin score 3 (10–13) with safranin-orange, von Kossa (black colour) and immunohistological for NPP1 (brown colour using 3-3′-diamino benzidene). (B) Quantification of von Kossa staining and quantification of NPP1 staining using image J, revealed an increase in calcification with increasing Mankin score (Mankin 1=0.02±0.02, Mankin 2=0.75±0.07, Mankin 3=1.39±0.15, p<0.0001) and an inverse correlation of the Mankin score with NPP1 staining (Mankin 1=84.49±4.93, Mankin 2=21.65±5.53, Mankin 3=11.05±5.57, p<0.0001). (C) Quantitative RT-PCR analysis of TNAP-mRNA, ANK-mRNA and NPP1-mRNA expression in human cartilage samples showed an inverse correlation of NPP1 expression (p=0.048) with cartilage calcification and no regulation for TNAP (p=0.50) and ANK (p=0.31). (D) Quantitative RT-PCR to detect NPP1 expression after stimulation of monolayer C28 chondrocytes with IL-1β (10 ng/ml) and TNFα (10 ng/ml) in comparison to unstimulated controls showed a downregulation of NPP1 (unstimulated=11.08±0.91, IL-1β=5.38±1.23, TNFα=6.23±1.0) expression after stimulation. ANK, ankylosing protein; IL, interleukin; NPP1, nucleotide pyrophosphatase phosphodiesterase 1; RT-PCR quantitative real time PCR, reverse transcriptase PCR; TNAP, tissue non-specific alkaline phosphate; TNF, tumour necrosis factor.

Calcification in ttw/ttw mice leads to OA-like changes in articular cartilage

In line with these human data we found a reduction of NPP1 protein after the induction of OA-like changes in mice using immunohistological staining (figure 2A). To study the functional contribution of NPP1, we then used the ttw/ttw mouse that carries a loss-of-function mutation in the NPP1 gene and displays postnatal development of progressive intervertebral ankylosis, peripheral joint hyperostosis, arterial and articular cartilage calcification and increased vertebral cortical bone formation.12 Using F18-PET we found that ttw/ttw mice exhibit enhanced bone-forming activity in areas such as the elbow joints, the intervertebral discs of the thoracic spine and in cartilage of non-weight-bearing areas, including ear cartilage (figure 2B), suggesting that mechanical stress is not required for induction of calcification. Micro CT analysis revealed the formation of osteophytes at the attachment sites of the ligaments (asterisks) (figure 1C). Investigating this calcification using histology, we found severe calcification of the articular cartilage of the tibial plateau in ttw/ttw mice (41.2% of total cartilage area), whereas no calcification of articular cartilage was found in wt mice (asterisks, figure 2D). To test the hypothesis that calcification is directly linked to the development of OA, we assessed OA severity in the ttw/ttw mice and found a drastic increase in OA-like cartilage degeneration in Mankin 12.5 mice in comparison with wild type (Mankin 4.8) animals (figure 2D). Intriguingly, calcification in the ttw/ttw mice was associated with increased expression of the hypertrophic marker collagen X in the articular cartilage area of ttw/ttw mice (88.31±2.14 n=4 in ttw/ttw mice and 21.84±5.24 n=4 in wt mice) indicating hypertrophic differentiation of chondrocytes (figure 2D).

Figure 2

(A) Immunohistological staining of cartilage samples for nucleotide pyrophosphatase phosphodiesterase 1 (NPP1) expression in knee joint sections of mice with and without induction of osteoarthritis (OA) (brown colour). Quantification of NPP1 staining revealed a significant reduction of staining upon induction of OA (uninduced=15.27±1.13 n=5, induced=7.73±0.45 n=5, p=0.0008). (B) Ttw/ttw mice showed an increased bone-forming activity in F-18-fluoride-positron emission tomography in comparison with wild type mice. The coloured picture indicates an enhanced radioactivity in elbow joints, the intervertebral discs of the thoracic spine (yellow, red colour) and auricles (blue colour), whereas bone metabolism in the long bones was unchanged. (C) Micro CT of knee joints revealed destruction of ttw/ttw articular surfaces and formation of osteophytes indicated in the picture by asterisks. (D) Histological staining using von Kossa combined with safranin-orange showed severe calcification of cartilage in ttw/ttw mice (40%) (white asterisks). This was accompanied by increased Mankin score (wt=3.50±0.65 n= 4, ttw/ttw=12.00±0.41 n=4, p<0.0001) and staining for collagen X in cartilage area in ttw/ttw mice (wt=21.84±5.24 n=4, ttw/ttw=88.31±2.14 n=4, p<0.0001).


Chondrocyte hypertrophy occurs physiologically during endochondral ossification; however, in healthy articular cartilage, chondrocytes maintain a stable phenotype and resist proliferation and differentiation. By contrast, articular chondrocytes from OA joints develop terminal differentiation and hypertrophy.16 Recent studies have demonstrated that (HIF) hypoxia inducable factor -2α-dependent hypertrophic differentiation contributes to OA progression and that chondrocyte hypertrophy is sufficient to induce matrix mineralisation.2 However, it has not been clear whether this is an epiphenomenon or whether matrix mineralisation contributes to OA progression. Therefore, we tested the hypothesis that the pyrophosphate pathway is involved in cartilage calcification in OA and that this calcification itself can trigger the development of OA-like changes. We show that NPP1 is downregulated in human and mouse OA cartilage, correlating with increased cartilage calcification. This decrease in NPP1 is associated with decreased extracellular pyrophosphate and, consequently, with the deposition of BCP crystals, which fits with the known function of NPP1 as an inhibitor of BCP crystal deposition. Downregulation of NPP1 in OA can be induced by proinflammatory cytokines such as IL-1β, which is the main cytokine in OA and TNFα, thereby promoting the production of BCP crystals. Interestingly, enhancing matrix mineralisation by altering pyrophosphate metabolism, and thereby promoting BCP crystal production, appears to be sufficient to drive OA progression. Additionally, we have shown in our previous paper that chondrocytes start producing BCP when getting hypertrophic, establishing a pathogenic role for this phenomenon. It will be important to establish whether, inhibiting cartilage mineralisation will be sufficient to halt OA progression. NPP1-deficient ttw/ttw mice displayed high expression of collagen X indicating late differentiation of chondrocytes. However, the calcifying activity in the ttw/ttw mice seems not to be dependent upon mechanical stimulation, as the non-weight-bearing cartilage also undergoes calcification. Therefore, this suggests that mineralisation, in turn, is sufficient to initiate hypertrophy, and to establish a vicious cycle resulting in cartilage breakdown. One difference between endochondral ossification during embryogenesis and OA is the role, in OA, of inflammatory cytokines.17 Interestingly, calcification and inflammation are common features of atherosclerosis and OA.10 There is accumulating evidence that atherosclerosis and OA share some similarities, which are especially reflected in the ttw/ttw mouse model, which is a model for both diseases, and the fact that there is a genetic link between mutations in NPP1 and hand OA and generalised arterial calcification of infancy.11 Taken together, our data show that calcification of the articular cartilage is not simply an associate feature secondary to dystropic cartilage degeneration, but is driven by an active metabolic process associated with hypertrophic chondrocyte differentiation, suggesting a direct role in OA progression.


This work was supported by the Interdisciplinary Center of Clinical Research (IZKF core unit SmAP), Münster, Germany and funded by the DFG (Pa689/12).


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • JB and YN are joint first authors and FD and TP are joint last authors.

  • Funding This work was supported by the Interdisciplinary Centre of Clinical Research (IZKF core unit SmAP), Münster Germany and funded by the DFG (Pa689/12).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the ethics committee of the Ärztekammer Schleswig- Holstein, Bad Segeberg, Germany.

  • Provenance and peer review Not commissioned; externally peer reviewed.