SEARCH

SEARCH BY CITATION

Keywords:

  • TISSUE-NONSPECIFIC ALKALINE PHOSPHATASE (TNAP);
  • DENTIN;
  • PYROPHOSPHATE;
  • OSTEOPONTIN;
  • MATRIX VESICLES

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Tissue-nonspecific alkaline phosphatase (TNAP) is expressed in mineralizing tissues and functions to reduce pyrophosphate (PPi), a potent inhibitor of mineralization. Loss of TNAP function causes hypophosphatasia (HPP), a heritable disorder marked by increased PPi, resulting in rickets and osteomalacia. Tooth root cementum defects are well described in both HPP patients and in Alpl−/− mice, a model for infantile HPP. In Alpl−/− mice, dentin mineralization is specifically delayed in the root; however, reports from human HPP patients are variable and inconsistent regarding dentin defects. In the current study, we aimed to define the molecular basis for changes in dentinogenesis observed in Alpl−/− mice. TNAP was found to be highly expressed by mature odontoblasts, and Alpl−/− molar and incisor roots featured defective dentin mineralization, ranging from a mild delay to severely disturbed root dentinogenesis. Lack of mantle dentin mineralization was associated with disordered and dysmorphic odontoblasts having disrupted expression of marker genes osteocalcin and dentin sialophosphoprotein. The formation of, initiation of mineralization within, and rupture of matrix vesicles in Alpl−/− dentin matrix was not affected. Osteopontin (OPN), an inhibitor of mineralization that contributes to the skeletal pathology in Alpl−/− mice, was present in the generally unmineralized Alpl−/− mantle dentin at ruptured mineralizing matrix vesicles, as detected by immunohistochemistry and by immunogold labeling. However, ablating the OPN-encoding Spp1 gene in Alpl−/− mice was insufficient to rescue the dentin mineralization defect. Administration of bioengineered mineral-targeting human TNAP (ENB-0040) to Alpl−/− mice corrected defective dentin mineralization in the molar roots. These studies reveal that TNAP participates in root dentin formation and confirm that reduction of PPi during dentinogenesis is necessary for odontoblast differentiation, dentin matrix secretion, and mineralization. Furthermore, these results elucidate developmental mechanisms underlying dentin pathology in HPP patients, and begin to explain the reported variability in the dentin/pulp complex pathology in these patients. © 2013 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Tissue-nonspecific alkaline phosphatase (TNAP) is an ectoenzyme highly expressed in mineralizing tissues.1 The primary role of TNAP in biomineralization is hydrolysis and reduction of local concentrations of inorganic pyrophosphate (PPi), a potent inhibitor of mineralization. TNAP also dephosphorylates osteopontin (OPN), removing another mineralization-inhibiting determinant from the extracellular matrix.2 Mutations in the human ALPL gene and the resulting loss of TNAP function cause hypophosphatasia (HPP), a heritable disorder resulting in increased PPi leading to the hypomineralization disorders broadly described as rickets and osteomalacia.3 The Alpl−/− mouse (formerly Akp2−/−) is a model for the infantile form of HPP, recapitulating the skeletal manifestations of that disease.4, 5

TNAP activity is high in tissues of the periodontia, the tooth root and supporting dentoalveolar tissues.6, 7 The first link between PPi metabolism and the tooth arose from observations that patients with HPP suffered premature tooth loss, often with intact roots being present.8 Subsequently, it was demonstrated that HPP causes severe aplasia or hypoplasia of root acellular cementum, resulting in deficient attachment of the tooth root to surrounding alveolar bone.9 This loss of cementum was confirmed in two independently generated mouse models null for Alpl with TNAP loss-of-function, the so-called EM line10, 11 and the LJ line.4, 12 Evidence that the tooth is more sensitive to loss of TNAP than the skeleton derives from the clinical symptoms of odontohypophosphatasia, the mildest subtype of HPP, where serum biochemistry is altered but only the dentition is affected.3, 13

Although the effects of TNAP loss-of-function are well described for bone and cementum, the role for TNAP in tooth dentin formation remains unclear, as does the potential for pathological changes in dentin resulting from HPP. In the human case literature, reports have been inconsistent about the presence of dentin phenotypes in HPP patients, with some declaring no dentin defects and others citing widened pulp chambers and thin dentin.9, 14–19 In previous reports on tooth formation in the EM Alpl−/− mice, dentin formation was described as fundamentally sound, with only a mild delay in rate of mineralization;10 however, a more profound dentin mineralization defect was observed in the LJ Alpl−/− model.12, 20 In light of the remarkable effects of HPP on bone and cementum and the variability in reports of associated dentin phenotypes in humans and mice, the role of TNAP in dentin formation remains unclear and deserves further study.

Here, we define the role of TNAP in dentinogenesis through developmental analysis of dentin formation, biomineralization, and the odontoblast molecular profile in the LJ Alpl−/− model in order to better understand the mechanism of dentin defects and whether mineralization defects persist over time. The extracellular matrix protein OPN was demonstrated previously to contribute to skeletal hypomineralization in the Alpl−/− mouse,21 so we analyzed dentinogenesis in double-deficient Alpl−/−; Spp1−/− mice. Finally, we examined the ability of enzyme replacement therapy using mineral-targeting recombinant TNAP to rescue developmental dentin defects in the Alpl−/− mouse.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Animals

Preparation and genotyping of Alpl homozygous knock-out (Alpl−/−) and wild-type (WT; Alpl+/+) mice were described previously.4, 12 Alpl−/− mice were maintained on a mixed background of 129S1/SvImJ and C57BL/6 and prepared by heterozygote breeding. Preparation of Alpl−/−; Spp1−/− double-deficient mice, single knock-outs, and WT littermates was previously described.21 Spp1−/− mice were maintained in a C57BL/6 background, and double-deficient mice lacking both Alpl and Spp1 were obtained by crossing Alpl+/− mice to Spp1+/− mice, with double heterozygote mice in turn used to generate Alpl−/−; Spp1−/− mice. Mice were allowed access to water ad libitum and fed a rodent diet supplemented with pyridoxine (vitamin B6) to suppress seizures in Alpl−/− pups. Animal procedures were approved by the Institutional Animal Care and Use Committee, University of Washington (Seattle, WA, USA) or Sanford-Burnham Animal Care and Use Committee (La Jolla, CA, USA).

Tissue preparation

For histology, in situ hybridization, and immunohistochemistry of Alpl−/− and WT samples, heads or mandibles were fixed in Bouin's solution overnight and stored with 70% ethanol. Samples of age 8 days postnatal (dpn) or older were demineralized in AFS (acetic acid, formaldehyde, sodium chloride), processed, and embedded for paraffin serial sectioning. Tissues were prepared as previously described for Alpl−/−; Spp1−/− double-deficient mice and controls,21 and enzyme replacement therapy–treated Alpl −/− mice and controls.12 Frontal (buccolingual) and sagittal sections were prepared to examine molar and incisor teeth.

Histological procedures

Standard hematoxylin and eosin (H&E) staining was performed for histological observation. Static histomorphometry using SPOT software (Diagnostic Instruments, Sterling Heights, MI, USA) was used to make calibrated measurements of dentin, predentin, and pulp chambers at fixed distances of 300 µm apical from the cementum-enamel junction of mesial roots of first mandibular molars. Measurements from two central root sections from 3 to 4 animals per genotype per age were used in an independent-samples Student's t test. Immunohistochemistry was performed using an ABC-based kit with a 3-amino-9-ethylcarbazole (AEC) substrate to produce a red reaction product indicating antigen localization, as previously described.22 Primary antibodies included rat anti-human ALPL (R&D Systems, Minneapolis, MN, USA) and LF-175 rabbit anti-mouse OPN (Dr. Larry Fisher, NIDCR, Bethesda, MD, USa). In situ hybridization (ISH) was performed using digoxigenin-labeled (DIG) mouse sense/antisense cRNA probes, employing NBT/BCIP (Nitro blue tetrazolium chloride/5-Bromo-4-chloro-3-indolyl phosphate, toluidine salt) for detection, as previously described.22 Probes included: mouse type I alpha I collagen (Col1) (Benoit de Crombrugghe, MD Anderson Cancer Center), mouse osteocalcin (Ocn) (John Wozney, Genetics Institute), and mouse dentin sialophosphoprotein (Dspp) (Helena Ritchie, University of Michigan).

Micro-computed tomography

Micro-computed tomography (µCT) of nondecalcified hemi-mandible specimens using Skyscan model 1072 (Kontich, Belgium) was performed as previously described.12

Undecalcified histology, transmission electron microscopy, and immunogold labeling

Hemi-mandibles were fixed and processed (without decalcification) in LR White acrylic plastic resin as previously described.23 Thin sections (0.5 µm) for light microscopy were cut on an ultramicrotome, and mineralized areas were visualized by von Kossa staining, followed by counterstaining with toluidine blue. For transmission electron microscopy (TEM), tissues were trimmed, and ultrathin sections (80 nm) were placed on polyvinyl formal- and carbon-coated nickel grids. Grid-mounted tissue sections were processed for colloidal-gold immunocytochemistry by incubation of the sections with a goat anti-mouse primary antibody against OPN (R&D Systems), after which immunolabeling patterns were visualized by incubation with protein A-colloidal gold complex (14 nm gold particles), followed by staining with uranyl acetate and lead citrate, as described previously.23 Stained tissues were examined using a Philips Technai transmission electron microscope operated at an accelerating voltage of 120 kV.

Enzyme replacement therapy protocol for Alpl−/− mice

The production and characterization of a bone-targeted, soluble human TNAP—ENB-0040 or Asfotase Alfa (formerly sALP-FcD10)—has been previously reported, and daily subcutaneous delivery of doses of 2.0 mg/kg and 8.2 mg/kg to Alpl−/− mice from birth onward have been demonstrated to increase serum alkaline phosphatase (ALP) activity, extend life span, and correct the phenotype of the skeleton, cementum, and enamel.12, 20, 24 ENB-0040 is currently in clinical trials for treating infantile and adult forms of HPP.25 Histological sections of molars and incisors of Alpl−/− mice treated with 2.0 mg/kg (ENB-0040 Tx-2.0 group) or 8.2 mg/kg (ENB-0040 Tx-8.2 group) were compared with WT controls at 44 dpn.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Loss of alkaline phosphatase causes root dentin mineralization defects

TNAP was highly expressed in newly differentiating odontoblasts, with elevated expression maintained by odontoblasts during both crown and root formation (Fig. 1A, B). TNAP induction in odontoblasts coincided with secretion and mineralization of the predentin matrix, the initial layer of which is known as the “mantle dentin.”

thumbnail image

Figure 1. Loss of alkaline phosphatase causes root dentin mineralization defects. (A, B) Tissue-nonspecific alkaline phosphatase (TNAP) is highly expressed in odontoblasts (od), and its induction (black arrow in B) coincides with cell polarization, predentin (pd) secretion, and mineralization to dentin (d) (Mouse first molar, 8 dpn). (C, D) Micro-CT demonstrates poorly mineralized molar roots and incisor (*) in 10 dpn Alpl−/− mice compared with WT, in addition to generalized reduction of bone mineralization in the mandible. (E–G) WT molars at 14 dpn show well-developed and mineralized roots, whereas (H) Alpl−/− root dentin suffered from developmental mineralization defects (*) ranging in severity from (I) reduced circumpulpal dentin development and arrested mineralization on the lingual aspect to (J) delay in predentin mineralization to dentin proper on the buccal aspect. Odontoblasts in association with unmineralized dentin in I and J are disorganized and dysmorphic (arrowheads). Regions of unmineralized osteoid accumulation (#) are apparent in surrounding alveolar bone in H. Od = odontoblasts; de = dentin; pd = predentin. Scale bar = 200 mm for panels A, E, and H, 100 µm for B, and 50 µm for F, G, I, and J.

Download figure to PowerPoint

Alpl−/− mice featured poorly mineralized tooth molar and incisor roots, in addition to a generalized loss of bone mineralization in the mandible (Fig. 1C, D). Histology of first mandibular molars at 14 dpn revealed that although crown dentin appeared normal in Alpl−/− mice, root dentin suffered from developmental mineralization defects (here seen in decalcified samples but visualized indirectly by loss of pink/purple staining reflecting mineral-bound noncollagenous proteins) ranging in severity from mild to very severe, often within the same molar (Fig. 1E–G for WT, Fig. 1H–J for Alpl−/−). Mild effects included a slight delay in mineralization of the predentin matrix to mineralized dentin matrix proper. This lag in mineralization was observed as approximately 100 to 200 µm of unmineralized predentin in Alpl−/− roots (Fig. 1H, J), whereas in WT controls, mineralization of predentin followed soon after secretion of the matrix (Fig. 1E–G). Notably, delayed dentin mineralization is entirely consistent with descriptions of tooth development in the EM Alpl−/− model.10 More severe dentin mineralization defects manifested as both complete lack of dentin mineralization, as well as reduction in circumpulpal dentin formation, resulting in thin and unmineralized roots (Fig. 1H, I). Reduced dentinogenesis was associated with disorganization of the odontoblast layer and flattened, dysmorphic odontoblasts lacking the typical columnar morphology.

In the majority of Alpl−/− animals analyzed by histology (n = 10), the severe inhibition of dentin mineralization was limited to the lingual aspect of the first molar (n = 8), whereas the contralateral buccal side was mildly affected. In a minority of Alpl−/− animals, both aspects were either severely affected (n = 2) or both mildly affected (n = 2) (Supplemental Fig. S1). When the entire root lacked dentin mineralization, pulp chambers were larger, root lengths were noticeably shorter, and root shape was dysmorphic (Supplemental Fig. S1C). More severe effects on dentinogenesis were associated with more extensive alveolar bone mineralization defects (increased osteoid accumulation), suggesting local differences were involved in the incongruity of buccal and lingual dentin phenotypes.

Histological examination of maxillary first molars and mandibular second molars at 14 dpn confirmed that Alpl−/− affected root but not crown dentin, a phenotypic pattern also observed in the root analog (lingual aspect) of the continuously erupting mouse mandibular incisor (Supplemental Fig. S2).

At 21 dpn, the mildly affected buccal dentin aspects of molars appeared well mineralized, indicating that the delayed mineralization of predentin at 14 dpn was corrected given time, and root development completed normally (Fig. 2). Root dentin mineralization of the lingual aspect remained arrested, with little increase in dentin formation. This equated to a root that continued growing in length but did not advance dentin apposition because of a reduction in circumpulpal dentin formation. An atypical tissue layer formed outside these thin and unmineralized roots, resembling unmineralized cementum, i.e., cementoid. The persistent root dentin defect was reflected in a poorly developed lingual aspect on the Alpl−/− incisor at this age (Fig. 2C, F).

thumbnail image

Figure 2. Arrested dentinogenesis in Alpl−/− molars is maintained over time. At 21 dpn, when (A, B) WT root formation is fundamentally completed, (D, E) Alpl−/− molar root predentin on the lingual aspect remains reduced (*) with little development of circumpulpal dentin, whereas the buccal aspect is well mineralized. Odontoblasts on the lingual side are dysmorphic (arrowhead in E). A layer resembling cementoid (cm) is observed where dentin is severely arrested. A parallel dentin defect (*) in incisor (lingual) root analog is noted when (C) WT is compared with (F) Alpl−/− at 21 dpn. Statistical analysis indicates that Alpl−/− molars feature significantly increased pulp chamber width, as well as increased predentin and decreased dentin width on the lingual (L) (severely affected) aspect at (G) 14 dpn, and nonsignificant trends in the same directions at (H) 21 dpn. Buccal (B) dentin and predentin are not significantly different between Alpl−/− and WT at either age. Od = odontoblasts; de = dentin; pd = predentin; cm = cementoid. Scale bar = 200 µm for panels A, C, D, and F, and 100 µm for B and E.

Download figure to PowerPoint

Histomorphometry of molars at 14 and 21 dpn confirmed that Alpl−/− root buccal dentin thickness did not differ from that of WT controls, whereas lingual dentin thickness was diminished at 14 and 21 dpn, and lingual predentin was significantly thicker and pulp chambers were significantly larger at 14 dpn, versus controls (Fig. 2G, H).

Loss of odontoblast gene expression in Alpl−/− tooth roots

Because of the odontoblast disorganization and dysmorphism observed in association with arrested dentinogenesis, in situ hybridization was used to analyze odontoblast marker gene expression. Although type 1 collagen (Col1) expression was unaltered, expression of odontoblast marker genes osteocalcin (Ocn) and dentin sialophosphoprotein (Dspp) was dramatically reduced in association with cell morphology and mantle dentin mineralization and apposition defects (Fig. 3). Dentin matrix protein 1 (Dmp1) gene expression was not strong enough to conclude any differences; however, DMP1 protein was noticeably reduced in Alpl−/− odontoblasts (Supplementary Fig. S3). Expression of marker genes was strong in Alpl−/− crown odontoblasts, and their reduction was less pronounced in regions where dentin was only mildly affected. In areas where mantle dentin remained completely unmineralized, Ocn was expressed in newly differentiated odontoblasts, but expression was lost in regions where cells lost their characteristic morphology and organization, namely the buccal aspect. Dspp, a later-expressed root odontoblast marker, failed to be expressed at any measurable level in dysmorphic odontoblasts.

thumbnail image

Figure 3. Loss of odontoblast phenotype in Alpl−/− tooth roots. In situ hybridization performed on 14 dpn tissues indicates type 1 collagen (Col1) expression is not different in (A, B) WT versus (C, D) Alpl−/−, although robust expression of odontoblast marker genes osteocalcin (Ocn) and dentin sialophosphoprotein (Dspp) observed in (E, F, I, J) WT is reduced or absent in (G, H, K, L) Alpl−/− root odontoblasts associated with arrested or delayed dentin. Od = odontoblasts. Scale bar = 200 µm for panels A, C, E, G, I, and K, and 100 µm for B, D, F, H, J, and L.

Download figure to PowerPoint

Matrix vesicle initiation of mineralization in Alpl−/− mantle dentin

Matrix vesicles (MVs) have been suggested to be important in the initiation of mineralization in mantle dentin.26, 27 TEM was used to determine if MV formation, distribution, initiation of mineralization, and rupture, contributed to the dentin mineralization defect in the Alpl−/− mice. The region of interest was unmineralized apical mantle dentin of WT and Alpl−/− molars (Fig. 4A, B). The number, distribution and morphology of MVs in the mantle dentin of Alpl−/− mice appeared normal compared with WT controls (data not shown), showing varying degrees of mineralization as they progressed from the unmineralized state to the mineralized state, and then to rupture of the MV membrane where the mineral deposits are then exposed to the extracellular milieu (Fig. 4C–J). OPN is an extracellular matrix protein and member of the SIBLING protein family expressed in mineralized and nonmineralized tissues,28 with potent actions on regulating hydroxyapatite crystal growth in mineralizing extracellular matrices.2, 29–33 Immunogold labeling demonstrated OPN localization to ruptured MVs (Fig. 5; see also Fig. 6 for OPN immunohistochemistry) as reported previously34 but not to unruptured MVs. However, unlike predentin in WT mice, Alpl−/− predentin did not feature a mineralization front, did not proceed to mineralize, and failed to become mineralized dentin proper (Fig. 4B), essentially aborting at the ruptured MV stage.

thumbnail image

Figure 4. Undecalcified histology and TEM of matrix vesicles in molar mantle dentin of Alpl−/− mice. Generally unmineralized mantle dentin in the most apical portion of molar roots (arrows in A and B) was sectioned and stained to examine the ultrastructure of matrix vesicles (MVs). (C–J) TEM revealed that in Alp−/− mantle dentin, the number and distribution of MVs is not different from that in WT controls, with MVs being present at varying stages of mineralization from early to late stage, here depicted on a scale of 1 to 3, with 3 being the most mineralized MVs. De = dentin; * = mineralization inhibited dentin.

Download figure to PowerPoint

thumbnail image

Figure 5. Immunogold labeling for osteopontin in mantle dentin of Alpl−/− mice. Immunogold labeling demonstrates high levels of OPN localization at sites of mineralization-ruptured MVs (stage 3 of mineralization in a of range 1 of 3, where 3 is most advanced), but not at nonruptured MVs, in mantle dentin of Alpl−/− mouse molars.

Download figure to PowerPoint

thumbnail image

Figure 6. Role of osteopontin in dentin defects of alkaline phosphatase-deficient mice. (A, B) In WT tissues at 14 dpn, OPN localization is observed in bone (b), acellular cementum (c), and expressed by odontoblast cells (od). (C, D) In Alpl−/− tissues, OPN is concentrated in the thin region of mantle dentin in generally unmineralized roots (black arrow in D), as well as being localized to newly differentiated odontoblasts. OPN is absent from the root surface in Alpl−/−molars, reflecting a lack of cementum. Compared with (E, F) WT and (I, J) Spp1−/− mice, those mice deficient for both Alpl and Spp1 (K, L) exhibit root dentin defects (*) indistinguishable from (G, H) Alpl−/− mice at 10 dpn. Od = odontoblasts; de = dentin; b = bone; c = cementum. Scale bar = 200 µm for panels A, C, E, G, I, and K, and 100 µm for B, D, F, H, J, and L.

Download figure to PowerPoint

Role of osteopontin in dentin defects of Alpl−/− mice

Systemic circulating OPN concentration is increased in the Alpl−/− mouse, and it was demonstrated that elevated OPN contributes to the skeletal mineralization pathology caused by increased PPi in the Alpl−/− mice.21 In light of these previous findings and our immunogold label demonstrating the presence of OPN at sites of ruptured MVs in unmineralized root mantle dentin matrix in Alpl−/− mice (Fig. 5), we further examined the contribution of OPN to Alpl−/− dentin deficiencies, first by immunostaining for OPN in Alpl−/− tooth histological sections, and second by crossing Alpl−/− mice onto the Spp1−/− (OPN-encoding gene) background. In WT molars, OPN localization was observed in bone, acellular cementum, and odontoblast cells (Fig. 6A, B). In Alpl−/− tissues, OPN staining was concentrated in some regions of alveolar bone (Fig. 6C, D), consistent with previous demonstrations of increased OPN at other skeletal locations.21–35 OPN protein was also heavily concentrated in the thin region of mantle dentin in Alpl−/− molars (Fig. 6D, arrow). Although OPN localization in Alpl−/− root odontoblasts was weaker in general, the newly differentiating odontoblasts near the root apex stained strongly for OPN, only losing OPN reactivity in association with dentin mineralization defects, in a pattern matching loss of Ocn and Dspp gene expression in those cells (Fig. 3). OPN immunostaining was lacking from the root surface, reflecting the lack of cementum.12, 36

Mice lacking both Alpl and Spp1 exhibited root dentin defects indistinguishable from Alpl−/− mice using histology on decalcified sections (Fig. 6E–L). Thus, despite a likely role for OPN in inhibition of mantle dentin mineralization in Alpl−/− mice, ablation of OPN was not sufficient to rescue the observed dentin defects.

Enzyme replacement therapy rescues the dentin phenotype in Alpl−/− mice

The production and characterization of a bone-targeted, soluble, human form of TNAP, ENB-0040, has previously been reported, and daily subcutaneous delivery of doses of 2.0 mg/kg and 8.2 mg/kg to Alpl−/− mice from birth has demonstrated increased serum ALP, extended life span of the mice, and correction of the phenotype of the skeleton and tooth crown and root.12, 20, 24 Compared with WT mice at 44 dpn, Alpl−/− mice treated with either 2.0 mg/kg or 8.2 mg/kg/day exhibited normal molar root dentin formation and mineralization (Fig. 7). Interestingly, dentin of the incisor root analog was not rescued at either dose.

thumbnail image

Figure 7. Enzyme replacement therapy rescued the dentin phenotype in alkaline phosphatase-deficient mice. Compared with (A) WT molars at 44 dpn, molars of Alpl−/− mice treated with either (B) 2.0 mg/kg or (C) 8.2 mg/kg/day exhibit normal root dentin (de) mineralization. (D–F) Incisor root analog dentin of the Alpl−/− is not rescued at either dose and remained unmineralized (*) with disorganized odontoblasts at 44 dpn. Scale bar = 200 µm.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

The results of the current study demonstrate a developmental defect in dentinogenesis in the Alpl−/− mouse model of HPP. The dentin defect is root-targeted and varies from delayed mineralization in the mildest case, to arrest of root mantle dentin mineralization, lack of circumpulpal dentin, and odontoblast differentiation defects in the most severe manifestations. The mildly delayed mineralization was corrected given time, but when dentinogenesis was significantly perturbed at the mantle dentin phase, defects persevered throughout molar tooth development. Although HPP case reports do not uniformly identify dentin defects, findings in the Alpl−/− mouse are consistent with reports of widened pulp chambers, thin root dentin, and “shell teeth.” Our data further reveal that the Alpl−/− dentin defect resulted from inability of MV-initiated mineralization foci to expand into a mineralization front and that increased OPN accumulation likely contributed to the aborted mineralization of Alpl−/− mantle dentin. Importantly, administration of bioengineered TNAP showed that early intervention allows for rescue of dentinogenesis and restoration of molar mineralization.

An important role for TNAP in root dentin development

The dramatic induction of TNAP expression as pulp cells differentiated into polarized, secretory odontoblasts suggests a key role for TNAP in mineralizing the predentin matrix. TNAP expression remains robust in crown and root odontoblasts throughout development, although studies here reconfirm that the root bears the brunt of the Alpl−/− dentin phenotype. This conserved pattern in molars and continually erupting incisors argues for a real difference in the necessity for TNAP in root versus crown dentin, and provides evidence against the possibility that crown dentin is spared owing to its prenatal development under maternal (i.e., Alpl heterozygote) influence. Differences in crown versus root dentin matrix composition and mineralization37–39 may be related to this observation.

The results here confirm observations of severely hypomineralized root dentin in the LJ model for HPP,12, 20 whereas the previous report on tooth formation in Alpl−/− mice from the EM colony reported a much milder delay in the rate of root dentin mineralization,10 despite similar severity in skeletal phenotypes between the two independently generated lines.5 However, the analysis is not so straightforward. In addition to previously unreported regions of severely arrested and unmineralized root mantle dentin identified here, we also observed a milder dentin delay, often on the contralateral aspect of the same Alpl−/− molar that was consistent with the defect reported by Beertsen and colleagues. Thus, differences in severity of tooth phenotype between HPP patients, and even between Alpl−/− littermates in inbred mice, may be compounded by a large degree of intertooth and intratooth variations. Possible reasons for these dissimilarities include local tissue differences in concentrations of PPi and OPN, as well as levels of other mineral-binding proteins. TNAP coordinates with other factors including PHOSPHO1, progressive ankylosis protein (ANK), and ectonucleotide pyrophosphatase phosphodiesterase 1 (NPP1) to determine local mineral metabolism,35, 36, 40 and nuances of these interrelationships warrant further analysis. Ultimately, evidence supports that the mantle dentin phase is very sensitive to disturbance in the mineralization process, and when foci cannot merge to a coherent mineralization front, circumpulpal dentinogenesis is severely affected.

Changes in root odontoblasts resulting from loss of alkaline phosphatase

Establishing the nature and extent of the dentin defect in Alpl−/− mice allowed us to identify an odontoblast cellular defect in situ, associated with those areas of severe failure of dentin mineralization. Odontoblasts lost cell polarity, became flattened and disorganized, and lost expression of key markers of differentiation, Ocn, Dspp, and DMP1. These changes suggest a process akin to dedifferentiation, where cells still expressed high levels of type I collagen but lost phenotypic odontoblast characteristics. Whether these cell changes contribute to dentin defects or arise as a result of them is under investigation. Recently we reported that pulp cells from HPP-diagnosed subjects were unable to acquire an odontoblast phenotype and to promote mineralization in vitro, which was partially restored by phosphate/PPi modulation.41

The role for osteopontin in inhibition of Alpl−/−dentin mineralization

OPN is an extracellular matrix protein with mineral-inhibiting properties in its phosphorylated form.32, 42 OPN is expressed by odontoblasts and found in the circulation.43, 44 Elevated OPN contributes to the increased PPi in inhibiting bone mineralization in Alpl−/− mice,21 and we sought to determine if there was a pathological role for OPN in Alpl−/− dentin hypomineralization. Two observations argued for such a role. First, OPN was abundant at sites of ruptured matrix vesicles (MV) in Alpl−/− mantle dentin, suggesting that OPN contributes to inhibiting the merger of mineral foci into a mineralization front (“secondary mineralization” that takes place after MV rupture). Second, immunostaining at the light microscopy level identified concentrated OPN in the mantle dentin of unmineralized Alpl−/− roots, at the location where mineralization normally starts to propagate after MV rupture. Although OPN localization was generally reduced in Alpl−/− root odontoblasts (matching the loss of Ocn and Dspp expression), early odontoblasts featured high OPN staining, suggesting that these cells deposited OPN in the mantle dentin before regressive changes. Additional OPN contributions to the mantle dentin matrix might derive from the circulation,43 especially because systemic OPN increased in Alpl−/− mice.21

Although our immunolabeling data support a role for OPN in inhibiting Alpl−/− dentin mineralization, overlaying the TNAP deficiency of the Alpl−/− mice on an Spp1−/− (OPN-deficient) background resulted in a tooth root dentin phenotype comparable to the Alpl−/− mouse, with no indication of rescue. This suggests that PPi, or possibly a factor downstream of PPi, may be the more central mineralization-inhibiting factor, and removal of OPN is not sufficient to overcome the PPi-mediated abortion of mineral crystal growth in the dentin. It should be noted that Spp−/− mice also show high PPi levels, and mice deficient for both Alpl and Spp1 feature increased PPi compared with the single knock-outs.21 Nevertheless, we cannot exclude contributions of other mineralization-inhibiting factors or other types of perturbation in the observed dentin defect, an aspect that remains under investigation. We are currently exploring whether relative correction of PPi in Alpl−/−;Ank−/− and Alpl−/−;Enpp1−/− double-deficient mice rescues dentinogenesis in parallel fashion to bone.35

Pyrophosphate removal is necessary for complete mineralization of root mantle dentin

The findings reported here shed new light on the function of TNAP in dentinogenesis and the nature of the dentin defect in Alpl−/− mice. Several roles have been hypothesized for TNAP in biomineralization, including provision of ionic phosphate for mineralization, clearance of mineral inhibitor PPi, dephosphorylation of mineralization-inhibiting proteins such as OPN, and participation in ion transfer or purinergic signaling. These need not be mutually exclusive of one another; however, we propose that the loss of PPi clearance is the primary cause for dentin defects in the Alpl−/− mice. Corroborating evidence for this can be found in studies demonstrating the effects on tooth formation of 1-hydroxyethylidene-1,1-bisphosphonate (HEBP). HEBP is, in a sense, a durable (nonhydrolyzable) PPi analog that also prevents crystal growth. Administration of HEBP to mice, rats, and guinea pigs inhibited root mantle dentin in a manner remarkably similar to that seen in Alpl−/− mice here.45–47 In a description that would aptly portray Alpl−/− tooth roots, Beertsen and colleagues reported that after HEBP administration to mice, mineralization of new dentin was inhibited, mantle dentin was formed but unmineralized, circumpulpal dentin was absent, and there were “regressive changes in the associated odontoblast layer” including loss of morphology and organization and reduced phosphoprotein secretion.45 Similar inhibition of root dentin mineralization was observed after forced expression of mineralization-inhibiting matrix Gla protein (MGP) in teeth.48

Our analyses, in conjunction with HEBP studies, strongly support TNAP as a PPi removal agent in dentinogenesis, where Alpl−/− mice suffer from physicochemical inhibition of dentin matrix mineralization because of increased PPi levels. The sensitivity of the mantle dentin phase to PPi inhibition can be explained by a dependence on MV-mediated mineralization, whereas circumpulpal dentin may rely primarily upon the actions of extracellular matrix proteins.26, 27 Defective proliferation of mineral crystals outside of MVs also led to hypomineralization in long bones and cartilage in Alpl−/− mice.49 An unresolved aspect is the ability of crown dentin to form properly in Alpl−/− mice, as well as under HEBP treatment.45 A plausible explanation lies in the differences in developmental influences on crown versus root dentin, including intrinsic differences in odontoblast populations, and interactions with different epithelial cells, and in the case of crown, with enamel matrix. Notably, odontoblasts adjacent to crown enamel or mineralized portions of root dentin in Alpl−/− teeth retain their odontoblast characteristics, leading us to speculate that regressive changes in Alpl−/− odontoblasts may stem from disrupted matrix-cell interactions with hypomineralized dentin.

Intriguingly, the HEBP-treatment studies also report a layer of unmineralized cementum matrix (“hyperplastic cementum”) associated with unmineralized root dentin,46, 47, 50 identical in appearance to the “cementoid” layer found at later time points in Alpl−/− mouse molar teeth. We speculate this to be a functional response of the periodontal compartment to loss of dentin competence in Alpl−/− mice, and this hypothesis is the subject of ongoing studies.

Rescue of dentin mineralization in Alpl−/− mice by injection of mineral-targeting recombinant TNAP (ENB-0040) confirmed that PPi reduction is necessary for root dentin mineralization. Improvement of Alpl−/− mouse skeletal and dental mineralization by ENB-004012, 20, 24 indicates that to some degree, TNAP activity or PPi reduction is important in physiological mineralization of all the hard tissues, although cell- and tissue-specific differences are indicated. The modulation of PPi represents a potential paradigm shift in considering therapies directed at dento-osseous regeneration.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

JLM is a consultant for Enobia Pharma, Inc., now Alexion Pharmaceuticals. MDM is a consultant for, and received research funding from, Enobia Pharma, Inc., now Alexion Pharmaceuticals. All other authors state that they have no conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

This research was supported in part by the Intramural Research Program of the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) of the National Institutes of Health (NIH). A portion of this research was completed when BLF, FHN, ABT, and MJS were affiliated with the University of Washington School of Dentistry, Department of Periodontics (Seattle, WA, USA). The authors thank Jirawan Wade for preparing histological sections, Lydia Malynowsky for her help with immunogold labeling and electron microscopy, Daisy Matsa Dunn for in situ hybridization, and Tracy Popowics for critical reading of the manuscript. Grant funding: NIH R01DE15109 (MJS), NIH R01 AR47908 and R01 DE12889 (JLM), and CIHR MOP-97858 (MDM).

Authors' roles: Study design: BLF, KJN, MDM, JLM, and MJS. Acquisition and analysis of data: BLF, KJN, HWT, ABT, MCY, SN, and MDM. Interpretation of data: BLF, KJN, FHN, SN, MCY, MDM, JLM, and MJS. Manuscript writing: BLF and KJN. Final approval of the manuscript: all authors.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Millan JL. Mammalian Alkaline Phosphatases: from Biology to Applications in Medicine and Biotechnology. Weinheim, Germany: Wiley-VCH; p. 337. 2006.
  • 2
    Addison W, Azari F, Sørensen E, Kaartinen M, McKee M. Pyrophosphate inhibits mineralization of osteoblast cultures by binding to mineral, up-regulating osteopontin, and inhibiting alkaline phosphatase activity. J Biol Chem. 2007;282(21):158723.
  • 3
    Whyte M. Hypophosphatasia and the role of alkaline phosphatase in skeletal mineralization. Endocr Rev. 1994;15(4):43961.
  • 4
    Narisawa S, Fröhlander N, Millán J. Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev Dyn. 1997;208(3):43246.
  • 5
    Fedde K, Blair L, Silverstein J, Coburn S, Ryan L, Weinstein R, Waymire K, Narisawa S, Millán J, MacGregor G, Whyte M. Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J Bone Miner Res. 1999;14(12):201526.
  • 6
    Groeneveld M, Everts V, Beertsen W. A quantitative enzyme histochemical analysis of the distribution of alkaline phosphatase activity in the periodontal ligament of the rat incisor. J Dent Res. 1993;72(9):134450.
  • 7
    Groeneveld M, Everts V, Beertsen W. Alkaline phosphatase activity in the periodontal ligament and gingiva of the rat molar: its relation to cementum formation. J Dent Res. 1995;74(7):137481.
  • 8
    Bruckner R, Rickles N, Porter D. Hypophosphatasia with premature shedding of teeth and aplasia of cementum. Oral Surg Oral Med Oral Pathol. 1962;15:135169.
  • 9
    van den Bos T, Handoko G, Niehof A, Ryan L, Coburn S, Whyte M, Beertsen W. Cementum and dentin in hypophosphatasia. J Dent Res. 2005;84(11):10215.
  • 10
    Beertsen W, VandenBos T, Everts V. Root development in mice lacking functional tissue non-specific alkaline phosphatase gene: inhibition of acellular cementum formation. J Dent Res. 1999;78(6):12219.
  • 11
    Waymire KG, Mahuren JD, Jaje JM, Guilarte TR, Coburn SP, MacGregor GR. Mice lacking tissue non-specific alkaline phosphatase die from seizures due to defective metabolism of vitamin B-6. Nat Genet. 1995;11(1):4551.
  • 12
    McKee MD, Nakano Y, Masica DL, Gray JJ, Lemire I, Heft R, Whyte MP, Crine P, Millán JL. Enzyme replacement therapy prevents dental defects in a model of hypophosphatasia. J Dent Res. 2011;90(4):4706.
  • 13
    Reibel A, Manière M, Clauss F, Droz D, Alembik Y, Mornet E, Bloch-Zupan A. Orodental phenotype and genotype findings in all subtypes of hypophosphatasia. Orphanet J Rare Dis. 2009;4:6.
  • 14
    Macfarlane J, Swart J. Dental aspects of hypophosphatasia: a case report, family study, and literature review. Oral Surg Oral Med Oral Pathol. 1989;67(5):5216.
  • 15
    Hu J, Plaetke R, Mornet E, Zhang C, Sun X, Thomas H, Simmer J. Characterization of a family with dominant hypophosphatasia. Eur J Oral Sci. 2000;108(3):18994.
  • 16
    Olsson A, Matsson L, Blomquist H, Larsson A, Sjödin B. Hypophosphatasia affecting the permanent dentition. J Oral Pathol Med. 1996;25(6):3437.
  • 17
    Baab D, Page R, Ebersole J, Williams B, Scott C. Laboratory studies of a family manifesting premature exfoliation of deciduous teeth. J Clin Periodontol. 1986;13(7):67783.
  • 18
    el-Labban N, Lee K, Rule D. Permanent teeth in hypophosphatasia: light and electron microscopic study. J Oral Pathol Med. 1991;20(7):35260.
  • 19
    Chapple I. Hypophosphatasia: dental aspects and mode of inheritance. J Clin Periodontol. 1993;20(9):61522.
  • 20
    Millán J, Narisawa S, Lemire I, Loisel T, Boileau G, Leonard P, Gramatikova S, Terkeltaub R, Camacho N, McKee M, Crine P, Whyte M. Enzyme replacement therapy for murine hypophosphatasia. J Bone Miner Res. 2008;23(6):77787.
  • 21
    Harmey D, Johnson K, Zelken J, Camacho N, Hoylaerts M, Noda M, Terkeltaub R, Millán J. Elevated skeletal osteopontin levels contribute to the hypophosphatasia phenotype in Akp2(−/−) mice. J Bone Miner Res. 2006;21(9):137786.
  • 22
    Foster BL, Nagatomo KJ, Bamashmous SO, Tompkins KA, Fong H, Dunn D, Chu EY, Guenther C, Kingsley DM, Rutherford RB, Somerman MJ. The progressive ankylosis protein regulates cementum apposition and extracellular matrix composition. Cells Tissues Organs. 2011;194(5):382405.
  • 23
    McKee M, Nanci A. Postembedding colloidal-gold immunocytochemistry of noncollagenous extracellular matrix proteins in mineralized tissues. Microsc Res Tech. 1995;31(1):4462.
  • 24
    Yadav MC, de Oliveira RC, Foster BL, Fong H, Cory E, Narisawa S, Sah RL, Somerman M, Whyte MP, Millán JL. Enzyme replacement prevents enamel defects in hypophosphatasia mice. J Bone Miner Res. 2012;27(8):172234.
  • 25
    Whyte MP, Greenberg CR, Salman NJ, Bober MB, McAlister WH, Wenkert D, Van Sickle BJ, Simmons JH, Edgar TS, Bauer ML, Hamdan MA, Bishop N, Lutz RE, McGinn M, Craig S, Moore JN, Taylor JW, Cleveland RH, Cranley WR, Lim R, Thacher TD, Mayhew JE, Downs M, Millán JL, Skrinar AM, Crine P, Landy H. Enzyme-replacement therapy in life-threatening hypophosphatasia. N Engl J Med. 2012;366(10):90413.
  • 26
    Stratmann U, Schaarschmidt K, Wiesmann HP, Plate U, Höhling HJ, Szuwart T. The mineralization of mantle dentine and of circumpulpal dentine in the rat: an ultrastructural and element-analytical study. Anat Embryol (Berl). 1997;195(3):28997.
  • 27
    Takano Y, Sakai H, Baba O, Terashima T. Differential involvement of matrix vesicles during the initial and appositional mineralization processes in bone, dentin, and cementum. Bone. 2000;26(4):3339.
  • 28
    Fisher L, Fedarko N. Six genes expressed in bones and teeth encode the current members of the SIBLING family of proteins. Connect Tissue Res. 2003; 44(Suppl 1): 3340.
  • 29
    Murshed M, McKee MD. Molecular determinants of extracellular matrix mineralization in bone and blood vessels. Curr Opin Nephrol Hypertens. 2010;19(4):35965.
  • 30
    Hunter GK, Hauschka PV, Poole AR, Rosenberg LC, Goldberg HA. Nucleation and inhibition of hydroxyapatite formation by mineralized tissue proteins. Biochem J. 1996;317(Pt 1):5964.
  • 31
    Boskey AL, Spevak L, Paschalis E, Doty SB, McKee MD. Osteopontin deficiency increases mineral content and mineral crystallinity in mouse bone. Calcif Tissue Int. 2002;71(2):14554.
  • 32
    Boskey AL, Maresca M, Ullrich W, Doty SB, Butler WT, Prince CW. Osteopontin-hydroxyapatite interactions in vitro: inhibition of hydroxyapatite formation and growth in a gelatin-gel. Bone Miner. 1993;22(2):14759.
  • 33
    Speer M, McKee M, Guldberg R, Liaw L, Yang H, Tung E, Karsenty G, Giachelli C. Inactivation of the osteopontin gene enhances vascular calcification of matrix Gla protein-deficient mice: evidence for osteopontin as an inducible inhibitor of vascular calcification in vivo. J Exp Med. 2002;196(8):104755.
  • 34
    McKee M, Zalzal S, Nanci A. Extracellular matrix in tooth cementum and mantle dentin: localization of osteopontin and other noncollagenous proteins, plasma proteins, and glycoconjugates by electron microscopy. Anat Rec. 1996;245(2):293312.
  • 35
    Harmey D, Hessle L, Narisawa S, Johnson K, Terkeltaub R, Millán J. Concerted regulation of inorganic pyrophosphate and osteopontin by akp2, enpp1, and ank: an integrated model of the pathogenesis of mineralization disorders. Am J Pathol. 2004;164(4):1199209.
  • 36
    Foster BL, Nagatomo KJ, Nociti FH, Fong H, Dunn D, Tran AB, Wang W, Narisawa S, Millán JL, Somerman MJ. Central role of pyrophosphate in acellular cementum formation. PLoS One. 2012;7(6):e38393.
  • 37
    Beertsen W, Niehof A. Root-analogue versus crown-analogue dentin: a radioautographic and ultrastructural investigation of the mouse incisor. Anat Rec. 1986;215(2):1068.
  • 38
    Steinfort J, van den Bos T, Beertsen W. Differences between enamel-related and cementum-related dentin in the rat incisor with special emphasis on the phosphoproteins. J Biol Chem. 1989;264(5):28405.
  • 39
    Ahmad M, Iseki H, Abduweli D, Baba O, Tabata MJ, Takano Y. Ultrastructural and histochemical evaluation of appositional mineralization of circumpulpal dentin at the crown - and root-analog portions of rat incisors. J Electron Microsc (Tokyo). 2011;60(1):7987.
  • 40
    Yadav MC, Simão AM, Narisawa S, Huesa C, McKee MD, Farquharson C, Millán JL., Loss of skeletal mineralization by the simultaneous ablation of PHOSPHO1 and alkaline phosphatase function: a unified model of the mechanisms of initiation of skeletal calcification. J Bone Miner Res. 2011;26(2):28697.
  • 41
    Rodrigues TL, Foster BL, Silverio KG, Martins L, Casati MZ, Sallum EA, Somerman MJ. Nociti FH Jr. Hypophosphatasia-associated deficiencies in mineralization and gene expression in cultured dental pulp cells obtained from human teeth. J Endod. 2012;38(7):90712.
  • 42
    Hunter GK, Kyle CL, Goldberg HA. Modulation of crystal formation by bone phosphoproteins: structural specificity of the osteopontin-mediated inhibition of hydroxyapatite formation. Biochem J. 1994;300(Pt 3):7238.
  • 43
    VandenBos T, Bronckers A, Goldberg H, Beertsen W. Blood circulation as source for osteopontin in acellular extrinsic fiber cementum and other mineralizing tissues. J Dent Res. 1999;78(11):168895.
  • 44
    Butler W, Ritchie H. The nature and functional significance of dentin extracellular matrix proteins. Int J Dev Biol. 1995;39(1):1699.
  • 45
    Beertsen W, Niehof A, Everts V. Effects of 1-hydroxyethylidene-1, 1-bisphosphonate (HEBP) on the formation of dentin and the periodontal attachment apparatus in the mouse. Am J Anat. 1985;174(1):83103.
  • 46
    Alatli-Kut I, Hultenby K, Hammarström L. Disturbances of cementum formation induced by single injection of 1-hydroxyethylidene-1,1-bisphosphonate (HEBP) in rats: light and scanning electron microscopic studies. Scand J Dent Res. 1994;102(5):2608.
  • 47
    Takano Y, Sakai H, Watanabe E, Ideguchi-Ohma N, Jayawardena C, Arai K, Asawa Y, Nakano Y, Shuda Y, Sakamoto Y, Terashima T. Possible role of dentin matrix in region-specific deposition of cellular and acellular extrinsic fibre cementum. J Electron Microsc (Tokyo). 2003;52(6):57380.
  • 48
    Kaipatur N, Murshed M, McKee M. Matrix Gla protein inhibition of tooth mineralization. J Dent Res. 2008;87(9):83944.
  • 49
    Anderson H, Sipe J, Hessle L, Dhanyamraju R, Atti E, Camacho N, Millán J, Dhamyamraju R. Impaired calcification around matrix vesicles of growth plate and bone in alkaline phosphatase-deficient mice. Am J Pathol. 2004;164(3):8417.
  • 50
    Alatli I, Hammarström L. Root surface defects in rat molar induced by 1-hydroxyethylidene-1,1-bisphosphonate. Acta Odontol Scand. 1996;54(1):5965.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
jbmr_1767_sm_SupplFig1.tif13466KSupplementary Figure 1
jbmr_1767_sm_SupplFig2.tif15478KSupplementary Figure 2
jbmr_1767_sm_SupplFig3.tif6030KSupplementary Figure 3
jbmr_1767_sm_SupplFigsLegend.doc21KSupplementary Figures Legend

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.