Analysis of the surface characteristics and mineralization status of feline teeth using scanning electron microscopy

Authors


Dr April DeLaurier, Division of Developmental Biology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK. T: +44 208959 3666 ext. 2095; F: +44 208816 2526; E: adelaur@nimr.mrc.ac.uk

Abstract

External resorption of teeth by odontoclasts is a common condition of unknown origin affecting domestic cats. Odontoclastic resorptive lesions involve the enamel cementum junction (ECJ, cervix) and root surface, leading to extensive loss of enamel, dentine and cementum. This study was undertaken in order to determine whether features of the surface anatomy and mineralization of feline teeth could explain why odontoclastic resorptive lesions are so prevalent in this species. Backscattered electron scanning electron microscopy was used to study enamel, cementum and dentine in non-resorbed, undemineralized teeth from adult cats. Analysis of the ECJ revealed thin enamel and cementum and exposed dentine at this site. Furthermore, enamel mineralization decreased from the crown tip to the ECJ, and dentine mineralization was lowest at the ECJ and cervical root. Analysis of cementum revealed variations in the organization and composition of fibres between the cervical, mid- and apical root although no significant differences in mineralization of cementum were detected between different regions of the root. Reparative patches associated with resorption of cementum by odontoclasts and repair by cementoblasts were present on the root surface. In conclusion, results suggest that the ECJ and cervical dentine could be at a greater risk of destruction by odontoclasts compared with other regions of the tooth. The relationship of these features to the development and progression of resorption now requires further examination.

Introduction

Domestic cats are predisposed to external resorption of the teeth by odontoclasts, leading to destruction of the periodontal attachment and tooth loss (Hopewell-Smith, 1930; Schneck & Osborn, 1976; Schlup, 1982; Lyon, 1992). Feline odontoclastic resorptive lesions (FORLs) are reported to occur in 26–72.5% of cats, and the incidence of the causative disease increases with age (Schlup, 1982; Coles, 1990; Crossley, 1991; Harvey, 1992; van Wessum et al. 1992; Gengler et al. 1995; Lund et al. 1998; Lommer & Verstraete, 2000, 2001; Ingham et al. 2001; Reiter et al. 2005). Clinical and histological studies have identified lesions as external, subgingival defects that initiate on the surface of the tooth root and progress to involve dentine. As the disease progresses, lesions become filled with vascular granulation tissue, and reparative activity may occur (Hopewell-Smith, 1930; Schneck & Osborn, 1976; Schlup & Stich, 1982; Reichart et al. 1984; Okuda & Harvey, 1992; Ohba et al. 1993; Okuda et al. 1995; Gorrel & Larsson, 2002; Roes & Mollenbeck, 2003; Mollenbeck, 2004; Roes, 2004).

Various factors have been suggested to underlie the disease, such as periodontal disease, dietary factors, mechanical stress, developmental tooth defects, breed and viral disease, although none of these factors has been definitively proven to be the direct cause of resorption (see review by Reiter & Mendoza, 2002). Recently, a significant association between the presence of lesions and elevated serum 25-hydroxyvitamin D and low urine specific gravity has been established (Reiter et al. 2005). FORLs have also been reported in exotic cat species (Berger et al. 1995, 1996a; Mendoza et al. 2000). In humans, a similar rare condition has been described as idiopathic resorption (Postlethwaite & Hamilton, 1989; Moody & Muir, 1991; Liang et al. 2003).

Resorption has been reported to involve the enamel cementum junction (ECJ, referred to as ‘neck lesions’), and the tooth root surface apical to the ECJ (Schlup & Stich, 1982; Reichart et al. 1984; Okuda & Harvey, 1992; Ohba et al. 1993; Gauthier et al. 2001; Gorrel & Larsson, 2002; Roes & Mollenbeck, 2003; Harvey et al. 2004 DeLaurier et al. 2005). To date, light microscopy and scanning electron microscopy (SEM) studies of the feline tooth have described the development and structure of enamel, dentine, cementum and periodontal ligament (Boyde, 1964, 1969a, 1989; Forsberg et al. 1969; Silva & Kailis, 1972; Kallenbach, 1976; Nalbandian & Frank, 1980; Hayashi, 1983; Sasaki et al. 1984, 1985; Skobe et al. 1985; Jones & Boyde, 1988; Bishop et al. 1991; Nanci et al. 1992; Okuda & Harvey, 1992; Orsini & Hennet, 1992; Vongsavan & Matthews, 1992; Crossley, 1995; Berger et al. 1996b; Colley et al. 2002). However, few studies have focused on those structures that are critical to the understanding of how tissues may be involved in the progression of FORLs, namely the surface features of the ECJ and the root. As a result, the relationship between these structures and disease remains unclear.

The present study was undertaken to describe the structural anatomy of non-resorbed feline teeth and thus provide a basis for future investigation of the relationship between these features with the progression of FORLs. Back scattered electron SEM was used to characterize the surface anatomy of undemineralized, teeth, and to analyse the relative mineralization of enamel, dentine and cementum at different locations on the tooth. We have focused on the surface anatomy of the tooth as FORLs begin on the external surface of the root and progress inwards. Previous microscopic and in vitro studies have demonstrated an increased rate of osteoclastic resorption in tissues which are less mineralized (Reid, 1986; Jones et al. 1995; Lu et al. 1999; Gentzsch et al. 2005). Therefore, the relative mineralization of different regions of the tooth was also analysed, as variations in mineralization between regions of the tooth may determine differences in the rate of destruction of different tissues by odontoclasts.

Materials and methods

Samples

Upper and lower jaws were obtained from a number of veterinary sources. All cats were euthanized for reasons unrelated to this research. Details of age, gender, neuter status, general health and dietary history were not known for most specimens. Intact left and right upper and lower jaws were removed using standard dissection tools and were radiographed using a digital dental radiography unit (RVG™, Trophy, France). Dentitions with radiographic evidence of alveolar bone loss, fractured teeth, resorptive lesions, ankylosis or periapical abscesses were excluded from analysis. All teeth used in the study were from cats with evidence of a permanent dentition where all teeth showed fully closed root apices. Dentitions were fixed in formaldehyde (from 4% paraformaldehyde in phosphate-buffered saline, pH 7.4) for 1 week at 4 °C, washed in distilled H2O and stored in 70% ethanol at room temperature. In total, 56 teeth from the dentitions of seven cats were used for the surface SEM study (Table 1). Thirty-six teeth from the dentitions of ten cats were embedded in PMMA (see below) and the sectioned block surface polished and used for quantitative SEM analysis.

Table 1.  Details of specimens used for surface and section SEM analysis
SideTeethSurfaceSection
Upper jawFirst, second, and third incisors 9 1
Canines 5 3
Second premolar 5 3
Third premolar 3 3
Fourth premolar 6 3
First molar 3 2
Lower jawFirst, second, and third incisors 5 0
Canines 5 3
Third premolar 5 6
Fourth premolar 5 6
First molar 5 6
Total 5636

Preparation of samples for surface analysis

Intact jaws were digested in distilled H2O with 2% Tergazyme™ (alkaline bacterial pronase, Alconox Inc., New York, USA), at 50 °C in a shaker at 50 r.p.m. until no soft tissue could be detected visually (approximately 6 weeks). Samples were rinsed for 1–2 h in running tap water, followed by several rinses in distilled H2O, and then air dried. Teeth were carefully removed from alveolar bone using bone-cutting forceps, such that alveolar bone surrounding the tooth was cut away until the tooth root was fully exposed and could be removed from the bone. Throughout this process, jaws and teeth were examined under a dissecting microscope to assess any signs of pathology or of damage during processing – any such teeth were eliminated from the study. To remove any remaining soft tissue, loose teeth were treated with NaOCl (3% available chlorine) at room temperature for 2 weeks. Samples were thoroughly washed in distilled water, air dried, mounted on aluminium rivets using conductive carbon putty (Leit-C Plast™, Agar Scientific, Stansted, UK) and coated with carbon by evaporation.

Surface analysis using SEM

Teeth were examined using a JEOL 5410LV™ SEM (JEOL, Welwyn Garden City, UK) with an accelerating voltage set at 15 or 20 kV. Images were recorded using digital image acquisition software (Printerface™, GW Electronics, Norcross, Georgia, USA). A calibration standard (Planotech™, Agar Scientific) with intervals of 10 µm was used to calculate field width.

The buccal surfaces (i.e. tooth surfaces facing the cheeks or lips) were examined, except when they were damaged or unclean, in which case the lingual/palatal surfaces (i.e. tooth surfaces facing tongue or palate) were analysed. High-magnification images (500× and 1000×) were recorded of the ECJ, the cervical root, mid-root and apical root surfaces. In cases of multirooted teeth, the buccal or lingual/palatal surfaces of both the mesial and the distal roots were examined. The average diameter of cementum fibres and cementocyte lacunae were measured using a calibration standard used in the microscope.

Embedding and sectioning of specimens

Whole jaws were cleared in xylene in glass jars under vacuum for 15–30 min and then left at room temperature on a shaker in xylene for 7–10 days. The xylene was decanted and the specimens were covered with methylmethacrylate (MMA, BDH, UK). Specimens were degassed under vacuum, lids were secured with plastic film and jars were placed on a rotating platform for 7–10 days. MMA containing a polymerizing catalyst (Azo-isobutyronitrile, BDH, UK, referred to as PMMA) was added and degassed under vacuum. Specimens were left in PMMA at room temperature under normal indoor light conditions until the PMMA was completely set; in most cases, this occurred within 3 weeks. Where PMMA would not set, specimens were placed in a 37 °C incubator overnight or until PMMA appeared to thicken and then allowed to continue to set at room temperature.

Following complete polymerization, embedded teeth were sectioned using a diamond saw (Isomet™, Buehler, Coventry, UK) for SEM analysis. Incisors, canines, premolars and molars were sectioned longitudinally, slightly offset from the centre of roots (Fig. 1A). Both the mesial and the distal roots of multirooted premolars and molars were sectioned and analysed separately. The specimen surface was ground using polishing paper (400–1200 grit, MetPrep, Coventry, UK) until the surface was as close to the central plane of the root as possible. The surface was polished on a cloth with 6-µm diamond paste, followed by 1-µm paste (MetPrep). Multiple blocks were fastened to an aluminium plate and sputter-coated with carbon as described above. Halogenated dimethacrylate standards of known backscattering coefficients (see below) were added to the plate for analysis of relative backscattering of samples (Banerjee & Boyde, 1998).

Figure 1.

Schematic diagrams illustrating the sectioning and analysis of methacrylate-embedded teeth. (A) Teeth were sectioned longitudinally, slightly offset from the centre of roots (plane of section indicated by dotted lines). Both the mesial and the distal roots of multirooted premolars and molars were sectioned and analysed separately. The specimen surface was ground until the surface was as close to the central plane of the root as possible. (B) Locations where mineralization was measured in sectioned teeth. 1. Crown tip enamel (surface enamel, inner enamel, and enamel dentine junction) and crown tip dentine. 2. Mid-crown enamel (surface enamel and enamel dentine junction) and mid-crown dentine. 3. Enamel and dentine and the enamel cementum junction. 4. Cervical root cementum and dentine. 5. Mid-root cementum (surface cementum and cementum dentine junction) and dentine. 6. Apical root cementum (surface cementum and cementum dentine junction) and dentine.

BSE SEM analysis of embedded specimens

PMMA blocks were examined using backscattered electron (BSE) mode, on a Zeiss DSM (digital scanning microscope; 962™, Zeiss, Welwyn Garden City, UK) using an annular solid-state BSE detector (KE Electronics, Toft, Comberton, Cambs., UK). Blocks were examined at an accelerating voltage of 20 kV, a beam current of 0.5 nA and working distance 17 mm, giving 11 mm sample to BSE detector clearance. Images were recorded under external computer control (Kontron Elektronik, Munich, Germany).

Both buccal and lingual/palatal surfaces of longitudinal sections of teeth were examined. High-magnification images (50×, 150×, 500×) were recorded at six locations on the tooth surface including the crown tip, buccal and lingual/palatal ECJs, the cementum at the cervical root, mid-root and at the apical root. Images were taken for analysis of features of the tooth in section, including tissue thickness, the average diameter of features of enamel and dentine, and mineralization.

The thickness of enamel (crown tip, mid-crown and ECJ within 250 µm coronally) and cementum (ECJ within 250 µm apically) at the cervical root, mid-root and root apex were measured using SigmaScan Pro™ image analysis software (SPSS Science, Chicago, USA). Differences in tissue thickness and distance were compared statistically between locations on the tooth using anova (SPSS™, SPSS Science).

Analysis of mineralization

Three halogenated dimethacrylate standards of known backscattering coefficients covering the backscattering range of mineralized dental tissues were attached to the plate and were used to calibrate specimens (Table 2) (for details of method, see Boyde et al. 1995; Banerjee & Boyde, 1998; Howell & Boyde, 1998). The grey levels of standards were used to scale tooth images so that a monobrominated standard (MBr) represented a grey level of 0 (black) and a tetrabrominated standard (Br4) represented a grey level of 255 (white). Relative grey level was used as an indicator of mineralization; less mineralized tissues appear dark grey, and more mineralized tissues appear light grey or white.

Table 2.  Backscattering coefficient of halogenated dimethacrylate standards, PMMA, bone, dentine, and enamel
 Mean backscattering coefficient
Monobromo: C22H25O10Br0.125*
Tetrabromo: C22H24O10Br40.195*
PMMA: H8C5O20.082*
Bone and dentine0.135–0.151*
Enamel0.179

Measurement of enamel, dentine and cementum mineralization was calculated from standardized grey level images analysed using SigmaScan Pro™ image analysis software. Grey levels were analysed from 50× magnification images. Mineralization was measured at 17 locations on the tooth, including the crown tip enamel (surface enamel, inner enamel and enamel at the enamel dentine junction, referred to as EDJ) and crown tip dentine, mid-crown enamel (surface enamel and enamel at the EDJ) and mid-crown dentine, the ECJ enamel and dentine, the cervical root cementum and dentine, the mid-root cementum (surface cementum and cementum at the cementum dentine junction, referred to as CDJ) and mid-crown dentine, and the apical root cementum (surface cementum and cementum at the CDJ) and apical root dentine (Fig. 1B). Grey levels were calculated by measuring mean pixel intensity over an area of 202 pixels (approximate area = 1800 µm2). Grey level values were compared statistically between locations on the tooth using anova. Owing to limited sample numbers of each tooth type, pixel intensity measurements could not be compared statistically between teeth.

Results

Surface characteristics of the tooth

The enamel was assessed in 34 teeth. In all cases, the surface was primarily smooth and featureless, with no evidence of perikymata or features associated with the external ends of enamel prisms (rods). Ten teeth (29%) had wear on the crown, revealing the prism structure underlying surface enamel (Fig. 2A). Nine teeth (26%) showed areas of developing enamel surface morphology, including enamel prisms and interpit enamel, in the fissures between cusps (Fig. 2B). In three teeth (9%), a band of ‘cobbled’ enamel, in which the prism-free surface enamel otherwise covering the tooth was missing, was detected 50–100 µm coronally from the ECJ (Fig. 2C). In a tooth fractured post-mortem, longitudinal enamel prisms were present, organized in Hunter–Schreger bands (Fig. 2D). In this tooth, a layer of prism-free enamel at the surface of the crown was also observed.

Figure 2.

Surface and section features of enamel (EP = enamel prisms). (A) Enamel prisms of a left upper first molar exposed through wear (scale bar = 25 µm). (B) Developing enamel surface morphology of a lower right fourth premolar showing enamel prisms (pit floor) and interpit enamel (IP, scale bar = 20 µm). (C) Developing enamel surface morphology showing ‘cobbles’ (CB) surrounded by prism-free enamel on the crown surface coronal to the enamel cementum junction (ECJ) in a left upper fourth premolar (scale bar = 25 µm). (D) Enamel of a lower left third premolar fractured post-mortem, showing horizontal sections through enamel prisms, which are orientated in alternate directions forming Hunter–Schreger bands, the prism-free surface enamel layer (PFE), and the enamel dentine junction (EDJ) (scale bar = 50 µm).

Among teeth with exposed dentine tubules at the cervical root (Fig. 3A), the area of exposed dentine extended from the ECJ for distances that ranged from 20 µm up to several millimetres apically. In a tooth that was fractured post-mortem, dentine tubules radiated from the pulp chamber, and could be detected extending to the EDJ (Fig. 3B).

Figure 3.

Surface and section features of dentine. (A) Dentine tubules (DT) exposed at the cervical root surface adjacent to cementum featuring extrinsic fibre bundles (EFB) on a left upper third premolar. (B) Upper left third premolar fractured post-mortem, showing dentine tubules (DT) radiating from the pulp chamber to the enamel dentine junction (EDJ) (D = dentine, E = enamel). All scale bars = 25 µm.

The ECJ structure was studied in 50 teeth, although the enamel margin could only be assessed in 28 teeth as there was cracking and fracture of enamel at this location among remaining teeth. The enamel margin had a ‘cobbled’ appearance and comprised circular projections 3–10 µm in diameter, extending in a band 10–50 µm coronally from the ECJ in 18 (64%) teeth (Fig. 4A), and was porous in nine (32%) teeth (Fig. 4B). In one sample (4%), cementum overlapped enamel, and the enamel margin could not be assessed (Fig. 4C). The root surface at the ECJ was assessed in all 50 teeth. In the majority of cases (34 teeth), the ECJ was associated with a porous root surface with exposed dentine tubules (68%, Fig. 4B). Fibrillar cementum with dense extrinsic fibre bundles 2–5 µm in diameter contacting enamel occurred in the remaining 16 teeth (32%, Fig. 4A).

Figure 4.

Surface features of the enamel cementum junction (ECJ) (E = enamel, RS = root surface). (A) ‘Cobbled’ (CB) enamel at the ECJ associated with a root surface featuring cementum with extrinsic fibre bundles (EFB) on a lower left third premolar (scale bar = 25 µm). (B) Enamel margin (EM) with enamel tubules (ET) and exposed dentine tubules (DT) on the root surface on a lower right canine (scale bar = 25 µm). (C) Cementum (CE) overlapping enamel on the root surface of an upper right fourth premolar (scale bar = 50 µm).

The entire root surface of the tooth was examined to determine if there were differences in surface anatomy between cervical, mid-root and apical regions. Extrinsic and intrinsic fibre cementum was detected throughout the root surface. Extrinsic fibre cementum was composed of bundles of fibres 2–5 µm in diameter (Fig. 5A). Individual fibres within these bundles could usually not be discerned. Fibre bundles were either densely organized (i.e. no space between fibres) or spaced 5–25 µm apart. In a tooth fractured post-mortem, projecting extrinsic fibre bundles could be traced from the CDJ to the root surface (Fig. 5B).

Figure 5.

Features of extrinsic and intrinsic fibre cementum, and cementocyte lacunae on root surfaces (EFB = extrinsic fibre bundles, IF = intrinsic fibres, CL = cementocyte lacunae). (A) Intrinsic fibre cementum distributed between projecting extrinsic fibre bundles on a left lower third premolar (scale bar = 25 µm). (B) Cementum surface of a left lower third premolar fractured post-mortem showing extrinsic fibre bundles that can be traced from the cementum dentine junction (CDJ) to the root surface (RS) (scale bar = 25 µm). (C) Patch of remodelling intrinsic fibre cementum on the root surface of an upper right fourth premolar. Cementocyte lacunae are present among intrinsic fibres (scale bar = 25 µm). (D) Cementocyte lacunae among extrinsic and intrinsic fibres on an upper right first molar. Canaliculi can be observed in the base of lacunae (white arrow, scale bar = 10 µm).

Intrinsic fibres (1–2 µm wide on average) were present among extrinsic fibre bundles (Fig. 5A), and sometimes formed extrinsic-fibre-free ‘patches’, which increased in number towards the root apex (Fig. 5C). Cementocyte lacunae were associated with intrinsic fibre cementum and their average diameter was 10 µm, although some larger lacunae (15–20 µm) were also observed (Fig. 5D).

In general, at the cervical root, the cementum was composed of dense extrinsic fibre bundles 2–5 µm in diameter, which was acellular near the ECJ, but had increasing numbers of cementocytes towards the mid-root. Towards the apex of the root, extrinsic fibre bundles increased in diameter (up to 10 µm) and became less densely packed with intrinsic fibres and cementocyte lacunae present between the extrinsic fibre bundles. Although the number of different types of teeth was limited in this study, the anatomical features described above were observed in all tooth types.

Anatomy of embedded and sectioned teeth

Enamel prisms and the structure of Hunter–Schreger bands were observed in all 36 embedded and sectioned teeth (Fig. 6A). The average pit floor width of enamel prisms was 5 µm, and the thickness of interpit enamel was approximately 1 µm. In 18 (50%) teeth, enamel tubules were visible within 25 µm of the EDJ (Fig. 6B). A layer of prism-free enamel 5–20 µm thick was present on the surface of all teeth (Fig. 6C).

Figure 6.

Features of enamel in sectioned teeth (EP = enamel prisms, IP = interpit ameloblastic enamel, EDJ = enamel dentine junction). (A) Enamel prisms surrounded by interpit enamel at the EDJ of a lower right canine. Enamel prisms are organized in alternating Hunter–Schreger bands (scale bar = 50 µm). (B) Enamel tubules (ET) at the EDJ of a lower right first molar (scale bar = 50 µm). (C) Prism-free enamel (PFE) at the crown surface of a lower right canine (scale bar = 20 µm).

The buccal and lingual/palatal surfaces of the ECJ were examined in 28 teeth. The enamel and root surface was examined within an area that extended 250 µm coronally and apically from the ECJ in all specimens. Enamel had edge-to-edge contact with cementum on the buccal sides of 22 (79%) teeth and lingual/palatal sides of 23 (82%) teeth, either abutting cementum (Fig. 7A) or extending as a thin projection to meet cementum (Fig. 7B). Enamel joined exposed dentine on the buccal sides of six (21%) teeth and lingual/palatal sides of five (18%) teeth (Fig. 7C).

Figure 7.

Features of the enamel cementum junction (ECJ) structure in sectioned teeth (CE = cementum, E = enamel). (A) Edge-to-edge contact between enamel and cementum in a lower left third premolar. (B) Enamel extending as a thin projection to meet cementum in a lower left third premolar. (C) Enamel margin (EM) joining exposed dentine (D) in a lower left canine. Scale bars = 50 µm.

Within dentine, tubules were 1–2 µm in diameter and spaced 7–10 µm apart. At the CDJ, the branching terminal ends of tubules < 1 µm in diameter were present (Fig. 8A). The granular layer of Tomes was present within dentine deep to the CDJ in all teeth. A region of parallel rows of von Korff fibres was visible within dentine at the EDJ in eight (22%) teeth (Fig. 8B).

Figure 8.

Features of dentine and cementum in sectioned teeth (E = enamel, D = dentine, CE = cementum, CDJ = cementum dentine junction). (A) Dentine tubules (DT) at the CDJ in a lower left first molar. Fine tubules are present branching off from larger tubules (asterisk and arrow). The granular layer of Tomes (GLT) is present in dentine deep to the CDJ. (B) Von Korff fibres (VKF) of dentine at the enamel dentine junction (EDJ) of the crown tip of a right mandibular canine. (C) Extrinsic fibre bundles (EFB) extending from the CDJ to the root surface (RS), associated with cementocyte lacunae (CL) in a right mandibular fourth premolar. All scale bars = 50 µm.

Extrinsic fibre bundles were present in all teeth at all locations on the root, extending from the CDJ to the root surface (Fig. 8C). Cementocyte lacunae, 7–10 µm in diameter, were incorporated into cementum, increasing in number and density towards the apex, from 20–500 µm apart near the cervical root to 20–150 µm apart at the root apex.

Enamel was significantly thicker at the crown tip compared with the mid-crown enamel and enamel at the ECJ (within 250 µm coronal to the ECJ; P < 0.001, Table 3). Enamel at the mid-crown was also significantly thicker than enamel at the ECJ (P < 0.001). Cementum thickness increased significantly from the cervical root to the root apex (Table 4). No significant differences were observed between buccal and lingual/palatal cementum thickness at the ECJ.

Table 3.  Thickness of enamel at the crown tip, mid-crown, and enamel cementum junction (ECJ, 250 µm coronal of junction) in embedded and sectioned teeth
LocationNumberMean thickness (SD) (µm)
  • *

    Significantly thicker than enamel at the mid-crown, P < 0.001.

  • Significantly thicker than enamel at the ECJ, P < 0.001.

Crown tip29297 (103)*
Mid-crown35176 (62)
250 µm coronal of ECJ35 55 (22)
Table 4.  Thickness of cementum at the enamel cementum junction (ECJ, 250 µm apical of junction), cervical root, mid-root and apical root in embedded and sectioned teeth
LocationNumberMean thickness (SD) (µm)
  • *

    Significantly thicker than cementum at the ECJ, P < 0.001.

  • Significantly thicker than cementum at the cervical root, P < 0.001.

  • Significantly thicker than cementum at the mid-root, P < 0.001.

250 µm apical of buccal ECJ22 11.1 (5.5)
250 µm apical of ligual/palatal ECJ20 12.3 (5.5)
Cervical root31 18.3 (13.6)*
Mid-root34 79.9 (62.2)*
Apical root27286.1 (160.3)*

Mineralization of feline dental tissues

Analysis of grey levels showed significant differences in mineralization between surface crown tip and inner crown tip enamel with enamel at the EDJ (P < 0.001, Table 5). No significant difference could be detected between crown tip surface enamel and inner crown tip enamel. At the mid-crown, a significant difference (P < 0.05) in mineralization was detected between the surface enamel and enamel at the EDJ. However, no significant difference was observed between mean crown tip enamel mineralization (including surface, inner and junctional enamel) and that of the mean mid-crown (including surface and junctional enamel). The mineralization of both the mean crown tip and the mean mid-crown enamel was significantly higher than the mineralization of enamel at the ECJ (P < 0.001).

Table 5.  Mineralization of the crown tip, mid-crown and enamel cementum junction (ECJ, 250 µm coronal of junction) enamel in embedded and sectioned teeth. Mineralization is represented by the mean grey level and standard deviation (SD)
Location NumberMean grey level (SD)
  • *

    Significantly higher mineralization compared with enamel at the crown tip EDJ, P < 0.001.

  • Significantly higher mineralization compared with enamel at the mid-crown EDJ, P < 0.05.

  • Significantly higher mineralization compared with enamel at the ECJ.

Crown tipSurface26180 (6.7)*
Inner24182 (4.0)*
Enamel dentine junction (EDJ)27174 (6.7)
Mean22179 (4.5)
Mid-crownSurface27181 (6.4)
Enamel dentine junction (EDJ)27176 (9.2)
Mean27178 (5.5)
250 µm coronal to ECJMean28173 (4.4)

There was no significant difference between crown tip and mid-crown dentine mineralization, although both were found to be significantly higher than the dentine at the ECJ (P < 0.001 and 0.05, respectively, Table 6). The dentine at the cervical root had the lowest mineralization of all regions of the tooth. The dentine at the ECJ was significantly more mineralized than the dentine at the cervical root (P < 0.05). However, the dentine at the mid-root and root apex was significantly more mineralized than the dentine at the ECJ and cervical root (P < 0.001). There was no significant difference in mineralization of cementum at any site on the root (Table 7). Cementum mineralization could not be analysed quantitatively at the ECJ as it was too thin to be measured at this site.

Table 6.  Mineralization of dentine at the crown tip, mid-crown, enamel cementum junction (ECJ, 250 µm coronal of junction), cervical root, mid-root and apical root, in embedded and sectioned teeth. Mineralization is represented by mean grey level and standard deviation (SD)
LocationNumberMean grey level (SD)
  • *

    Significantly higher mineralization than dentine at the ECJ, P < 0.001.

  • Significantly higher mineralization than dentine at the ECJ, P < 0.05.

  • Significantly higher mineralization than dentine at the cervical root, P < 0.05.

  • §

    Significantly higher mineralization than dentine at the ECJ and cervical root, P < 0.001.

Crown tip dentine25117 (17.1)*
Mid-crown dentine29114 (16.1)
Dentine 250 µm coronal of ECJ27106 (6.4)
Cervical root dentine25100 (7.4)
Mid-root dentine29120 (21.0)§
Apical root dentine21123 (18.1)§
Table 7.  Mineralization of cementum at the cervical root, mid-root and root apex in embedded and sectioned teeth. Mineralization is represented by mean grey level and standard deviation (SD)
Location NumberMean grey level (SD)
Cervical root 13117 (12.3)
Mid-rootCementum dentine junction (CDJ)14116 (4.3)
Surface cementum15122 (15.8)
Mean22118 (9.7)
Apical rootCementum dentine junction (CDJ)18121 (13.5)
Surface cementum18122 (12.1)
Mean21122 (10.0)

Discussion

An advantage of SEM compared with conventional histology is that it does not require demineralization of tissues, and allows the direct study of mineralized tissue surfaces. Previously, SEM has been used in studies of cat teeth to study enamel development (Boyde, 1964, 1969a,b), to measure the elemental composition of enamel, dentine and cementum (Colley et al. 2002), and to describe FORLs (Berger et al. 1996b; Gauthier et al. 2001; DeLaurier et al. 2005). In the present study, several features of teeth have been described which are consistent with previous descriptions of cat teeth made using other imaging methods. However, we have identified other structural features of cat teeth that have not been previously reported but may be significant for understanding FORLs.

Surface enamel was primarily smooth, lacking evidence of perikymata (incremental growth lines) and surface prisms, consistent with observations made in the cat and dog by Skobe et al. (1985). This prism-free layer was also observed in embedded and sectioned teeth, overlying the deeper prismatic enamel. This smooth surface is formed as a result of slowed incremental growth of enamel prisms by ameloblasts, which is associated with a change in the shape of the Tomes process, resulting in a lack of distinction between pit-floor and interpit enamel (Boyde, 1989). The absence of perikymata in the cat reflects rapid growth of the crown. In section, the average diameter of enamel prisms was 5 µm, which is consistent with the diameter of prisms described in humans and in dogs, and in previous studies of the cat (Boyde & Reith, 1968; Boyde, 1969a, 1989; Skobe et al. 1985; Boyde et al. 1988).

The high prevalence of exposed enamel prisms on occlusal surfaces and bulbous parts of the crown associated with wear marks suggests that the smooth enamel is normally worn away throughout life to expose underlying enamel prisms. The presence of exposed ‘pockets’ of cobbled enamel in the fissures of crowns and near the ECJ that were not associated with wear has not been described in previous studies of cat tooth enamel. The cellular process associated with the formation of this feature is unclear, but it may be the result of premature termination of enamel matrix secretion by ameloblasts. At cusp infoldings, adjacent populations of ameloblasts may converge and ‘strangulate’ due to space constraints and competition for nutrients from adjacent blood vessels, resulting in a lack of maturation of enamel prisms and failure to form a smooth enamel layer (Boyde, 1989).

Several previously unreported patterns of organization of the enamel margin and root surface were observed at the ECJ. A ‘cobbled’ enamel margin was frequently observed, but from the surface study it was unclear whether ‘cobbles’ represented a pattern of terminal-phase enamel secretion by ameloblasts forming spherical structures, or a thin layer of smooth enamel overlying dentine calcospherites. The presence of a thin enamel layer overlying dentine was confirmed in the study of sectioned teeth, suggesting ‘cobbles’ represent enamel overlying dentine globules.

In tooth surfaces examined, the predominant pattern of the ECJ was of a gap between enamel and cementum (present in 68% of teeth), followed by edge-to-edge contact between enamel and cementum (32%). In sectioned teeth, the predominant ECJ patterns were of edge-to-edge contact between enamel and cementum (present in 79% buccal and 82% lingual/palatal sides of teeth), and a gap between enamel and cementum exposing dentine (present in 21% buccal and 18% lingual/palatal sides of teeth). In teeth where the surface of the root was studied, only one tooth (4%) showed cementum overlapping enamel, while in sectioned teeth, no specimens featured cementum overlapping enamel at the ECJ. This contrasts with observations of the ECJ in humans, in which SEM analysis of the entire circumference of the ECJ revealed that edge-to-edge contact between enamel and cementum, and cementum overlapping enamel are the predominant patterns, while exposed dentine is least frequently observed (Schroeder & Scherle, 1988). Differences in the distributions of ECJ patterns between the surface and section analyses in the present study are probably due to variations in sampling, as teeth from the same cats were not used for both analyses.

Exposed dentine at the erstwhile ECJ on intact root surfaces and in sectioned teeth was a significant observation of this study. It could be a developmental feature of teeth, as this pattern has also been described in human teeth not affected by disease (Schroeder & Scherle, 1988). Alternatively, exposed dentine could be the result of attachment loss of the periodontal ligament (Grzesik & Narayanan, 2002). Exposed dentine lacking the protective cover of cementum may be at risk of the destructive activity of odontoclasts. In humans, tooth resorption may be associated with damage or deficiency of cementum (Jones & Boyde, 1988; Malek et al. 2001; Bilgin et al. 2004; Coyle et al. 2006). Mineralization of dentine was found to be lower than that of adjacent cementum, and its mineralization is lower at the cervical root and at the ECJ compared with other areas of the tooth. If dentine is exposed to active odontoclasts, it may be at risk of an increased rate of resorption compared with a root surface where the cementum covering the tooth is present.

Analysis of enamel thickness showed a significant decrease from the crown tip to the ECJ, consistent with previous findings using light microscopy (Crossley, 1995). Mineralization of dentine was lowest at the cervical root and ECJ while the crown tip dentine did not have a significantly higher mineralization compared with mid-crown dentine. This is consistent with previous observations from microindentation and mineral composition analysis of feline teeth (Sasaki et al. 1984; Hayashi & Kiba, 1989).

The underlying reason for lower mineralization of enamel and dentine at the ECJ and cervical root is unclear. Reduced mineralization of enamel may be a developmental feature, as enamel at the ECJ, which is the last to be formed in the tooth, undergoes a shorter period of maturation–mineralization compared with the rest of the crown. Enamel maturation is primarily controlled by a small population of maturation-stage ameloblasts that undergo cyclical changes in function and morphology (Boyde & Reith, 1983a,b). These cells remain upon the surface of enamel at the ECJ for a shorter period than they do on the surface of crown enamel, which forms earlier. This could lead to enamel being less mineralized at the ECJ. This, combined with low mineralization of dentine, could be related to resorption of the ECJ in FORLs (Reichart et al. 1984; Gauthier et al. 2001; Roes & Mollenbeck, 2003; DeLaurier et al. 2005). It has previously been shown through in vitro studies and analysis of mineralization that there is a relationship between the rate of resorption and the mineral concentration; less mineralized tissues are resorbed more rapidly (Reid, 1986; Jones et al. 1995; Lu et al. 1999; Gentzsch et al. 2005). The relationship between the low mineralization of tissues at the ECJ and cervical root compared with other regions of the tooth requires further investigation.

Previous analysis of FORLs has demonstrated that resorption also involves root surfaces (Gorrel & Larsson, 2002). However, the surface of the non-resorbed tooth was not examined by Gorrel & Larsson (2002) as the teeth were examined after sectioning. In the present study we addressed this by analysing the surface anatomy and have shown that the organization and composition of cementum fibres varies across the root. Extrinsic fibre bundle diameters were within the range described for humans, although the average diameter of bundles in the cat (2–5 µm) is lower than in humans (5–7 µm) (Jones & Boyde, 1972). Extrinsic fibre bundle diameter has been previously described in the tiger to be 4–7 µm (Jones, 1973). Feline cementocyte lacunae were within the size range of lacunae observed in human cementum (Jones, 1981). Features of root cementum in sectioned teeth were consistent with surface findings, including evidence of dense extrinsic fibre bundles and cementocytes embedded in cementum matrix. The size and arrangement of extrinsic fibre bundles varied by location on the tooth surface of the cat and, as in humans, the organization of extrinsic fibre bundles in cementum is probably associated with an adaptation of the periodontal ligament fibres to organize along lines of mechanical force (Hassell, 1993). In the cat, extrinsic fibre bundles were generally dense and small at the cervical root, becoming larger, more widely spaced, organized in rows, and associated with more cellular intrinsic fibre cementum towards the apex. This is consistent with the pattern in humans where small, dense fibres orientated perpendicular to the root surface at the cervical root are associated with opposing lateral compressive forces and protecting deeper periodontal ligament (Hassell, 1993). In the cat, as in other carnivores, the periodontal fibres have been described previously as numerous, dense and orientated obliquely to the root surface. Their attachment to the tooth is more apical than their attachment to alveolar bone and is adapted to stabilizing the tooth under heavy loads (Forsberg et al. 1969).

We also found that cementum was significantly thinner at the ECJ than at the root apex. These findings support observations made in the tooth surface SEM study that remodelling of the root, associated with greater thickening of cementum, increases towards the apex. This pattern of cementum thickening towards the root apex has previously been described as a normal feature of feline and human teeth (Forsberg et al. 1969; Bilgin et al. 2004). The relationship between cementum thickness and resorption could not be addressed in this study, but thin cementum, like thin enamel, may be at greater risk of removal by odontoclasts compared with elsewhere on the tooth. For example, cementum thinness will determine how quickly odontoclasts reach underlying, less mineralized dentine. Because dentine may be resorbed faster than cementum, this may lead to the more rapid progression of FORLs in this region.

Interestingly, there was no significant difference in mineralization of cementum between the CDJ and the surface, and between different locations on the root, which suggests that formation and mineralization of cementum occurs uniformly in feline teeth. In many teeth, cellular intrinsic fibre ‘patches’ were observed on the root surface, which indicate active remodelling and repair of the root by cementoblasts.

Inter-individual variation was not explored in this study, nor were the effects of age on root remodelling and the structure of the ECJ. However, these variables merit future study as it has been suggested that the susceptibility to root resorption varies among cats and there is an increased risk of the condition with age (Schlup, 1982; Harvey, 1992; van Wessum et al. 1992; Gengler et al. 1995; Lund et al. 1998; Ingham et al. 2001). Furthermore, an analysis of the anatomical variation between tooth types would be a valuable contribution towards understanding why some teeth are more predisposed to resorption than others (Schlup, 1982; Coles, 1990; Harvey, 1992; van Wessum et al. 1992; Lund et al. 1998; Ingham et al. 2001; Harvey et al. 2004). Analysis of mineralization of teeth among different species might contribute towards understanding why cats are uniquely predisposed to tooth resorption.

In conclusion, this study provides a detailed SEM analysis of the microscopic anatomy of adult feline teeth, and of the relative mineralization of enamel, cementum and dentine using quantitative backscattered SEM. Analysis of the ECJ revealed thin enamel and cementum and exposed dentine at this site. Furthermore, enamel mineralization decreased from the crown tip to the ECJ, and dentine mineralization was lowest at the ECJ and cervical root. These results indicate that the ECJ and cervical dentine may be at a greater risk of destruction by odontoclasts compared with other regions of the tooth in FORLs. The relationship between these features with FORLs requires further study.

Acknowledgements

We would like to thank Sheila J. Jones, for assistance with interpreting the surface SEM data, Peter Howell for help with the analysis of mineralization, and Mo Arora for help with preparation of specimens – all from the Department of Anatomy, UCL. This work formed part of a PhD thesis awarded to A.D. (University of London 2003). All SEM studies were performed in the Department of Anatomy, UCL. The facility for the determination of mineralization density at the microscopic scale was provided by the MRC.

Ancillary