Regional structural characteristics of bovine periodontal ligament samples and their suitability for biomechanical tests

Abstract

Mechanical testing of the periodontal ligament requires a practical experimental model. Bovine teeth are advantageous in terms of size and availability, but information is lacking as to the anatomy and histology of their periodontium. The aim of this study, therefore, was to characterize the anatomy and histology of the attachment apparatus in fully erupted bovine mandibular first molars. A total of 13 teeth were processed for the production of undecalcified ground sections and decalcified semi-thin sections, for NaOH maceration, and for polarized light microscopy. Histomorphometric measurements relevant to the mechanical behavior of the periodontal ligament included width, number, size and area fraction of blood vessels and fractal analysis of the two hard–soft tissue interfaces. The histological and histomorphometric analyses were performed at four different root depths and at six circumferential locations around the distal and mesial roots. The variety of techniques applied provided a comprehensive view of the tissue architecture of the bovine periodontal ligament. Marked regional variations were observed in width, surface geometry of the two bordering hard tissues (cementum and alveolar bone), structural organization of the principal periodontal ligament connective tissue fibers, size, number and numerical density of blood vessels in the periodontal ligament. No predictable pattern was observed, except for a statistically significant increase in the area fraction of blood vessels from apical to coronal. The periodontal ligament width was up to three times wider in bovine teeth than in human teeth. The fractal analyses were in agreement with the histological observations showing frequent signs of remodeling activity in the alveolar bone – a finding which may be related to the magnitude and direction of occlusal forces in ruminants. Although samples from the apical root portion are not suitable for biomechanical testing, all other levels in the buccal and lingual aspects of the mesial and distal roots may be considered. The bucco-mesial aspect of the distal root appears to be the most suitable location.

Introduction

The periodontal ligament (PDL) is the structure that interfaces the teeth with their surrounding alveolar bone. It allows changes in tooth position (e.g. during tooth eruption or under orthodontic pressure) and also has a damping function during chewing (Picton, 1989). It is a thin and highly specialized soft connective tissue located between the root cementum and the alveolar bone (for reviews, see Beertsen et al. 1997; McCulloch et al. 2000). Its structure bears witness to its function in tooth positioning and in the control and dissipation of mechanical stresses as generated during mastication and traumatic impacts. The PDL also prevents resorption and fusion of the root with the surrounding alveolar bone (i.e. ankylosis) (for reviews, see Beertsen et al. 1997; Andersson & Malmgren, 1999). In effect, the PDL, root cementum and bone should be viewed as a structural (Berkovitz, 2004) and functional (Ivanovski et al. 2006; Benatti et al. 2007) unit that allows both physiological and therapeutic tooth movements.

Prediction of tooth movements in terms of direction and magnitude is a central issue in dental biomechanics. In particular, the characterization of the PDL response to mechanical stimuli is essential in developing realistic analytical and numerical tools for the modeling of tooth displacements in orthodontics, periodontics and the interactions between the teeth and their restorations.

Using finite element analyses, a number of equation systems (i.e. the constitutive laws) have been proposed to characterize the biomechanical properties of the PDL (Pietrzak et al. 2002). They were based on data generated from human (Mandel et al. 1986; Pedersen et al. 1991; Toms et al. 2002), rat (Kawarizadeh et al. 2003) porcine (Dorow et al. 2002; Krstin et al. 2002), and bovine tissues (Pini et al. 2002; Sanctuary et al. 2005, 2006; Shibata et al. 2006).

Mechanical testing of the PDL requires a practical experimental model. Most importantly, the teeth and the surrounding bone must be large enough to allow their fractioning into smaller specimens and still allow their fastening to the grips of the mechanical testing devices. Bovine teeth are comparatively large in size and readily available. As such they qualify as a suitable model substrate for experimentation. Yet a basic knowledge of anatomy and histology is a prerequisite for selecting appropriate teeth or sites around the teeth for biomechanical testing. In this regard, however, there is an almost complete absence of information available on normal periodontal anatomy and histology of bovine teeth. To date, only one study has illustrated some structural aspects of bovine PDL (Berkovitz et al. 1997). Unfortunately, many of these are not actually relevant to the biomechanical response of the tissue.

Therefore the aim of the present study was to assess those anatomical and histological characteristics of the bovine attachment apparatus that affect the mechanical behavior of the PDL.

Materials and methods

Overview

Using freshly prepared sections of bovine molars, the object of the study was to prepare sets of undecalcified ground and decalcified semi-thin sections of the PDL from which the following characteristics were determined:

• size, number and area fraction of blood vessels

• periodontal ligament width;

• profile of the hard tissue interface expressed in terms of fractal dimensions.

The histological and histomorphometric analyses were performed at four different root depths (termed ‘a’ (most apical), ‘b’, ‘c’ and ‘d’ (most coronal)) (Fig. 1) and at six different circumferential locations around the distal and mesial roots of the mandibular first molar, that is (Fig. 2A,B): distal root–distal aspect (‘dd’), distal root–buccal aspect (‘db’), distal root–lingual aspect (‘dl’), mesial root–mesial aspect (‘mm’), mesial root–buccal aspect (‘mb’), and mesial root–lingual aspect (‘ml’).

Fig. 1.

Schematic drawing illustrating the sectioning of the mandibular first molar into four horizontal sections (a–d) of about 1 cm thickness.

Fig. 2.

Low power light micrographs showing (A) the mesial and (B) the distal roots of the first mandibular molar. At all root levels, the horizontal cross-section area was larger at the distal than at the mesial root, the buccal bone plate was thicker than the lingual, and the bone in the interradicular septum was highly trabecular. The six different locations investigated were distal root–buccal aspect (db), distal root–distal aspect (dd), distal root–lingual aspect (dl), mesial root–buccal aspect (mb), mesial root–mesial aspect (mm), and mesial root–lingual aspect (ml). The periodontal ligament (PDL) was rich in blood vessels (C), and cells and connective tissue fibers (D). A–C = ground sections; D = thin section. AB, alveolar bone; C, cementum; D, dentin; P, pulp.

Preparation of block specimens

Ten bovine mandibles (age range 2–4 years) were obtained from two local slaughterhouses (Vuillamy, Cheseaux-sur-Lausanne, Switzerland, and Abattoirs de Clarens, Vevey, Switzerland) immediately following the death of the animals. The mandibles were placed in an ice cooler and, within 30 min, sectioning of the mandibles was initiated in a specially equipped laboratory of the Swiss Federal Institute of Technology in Lausanne, Switzerland, following a method described by Sanctuary (2003). After removal of all remaining soft tissues using a scalpel, the left and right sides of the mandibles supporting the molars, premolars and canines were separated. The specimens were then cut along the long axes of the third premolar and second molar to obtain a block containing an intact first molar. The crowns were removed after sectioning the teeth at the level of the alveolar crest and, last, blocks each containing a mesial and a distal root were produced. The cuts were performed using a heavy industrial band saw (Magnum Bs 0633, Metabo, Nürtingen, Germany) under abundant irrigation with saline. The blocks containing one root were then mounted in a vice and further separated into four horizontal sections (a, b, c, and d) ~1 cm in thickness (Fig. 1). While one hemi-mandible was prepared, the other hemi-mandibles were kept in a refrigerator at 5 °C to minimize tissue degradation. All newly produced horizontal root sections were immediately placed in room temperature fixative solution. Rapidly producing 1-cm-thick horizontal slices yields a more thorough fixation of the tissues than immersing larger specimens. Further tissue processing was performed in the histology laboratory of the Department of Periodontology and Prosthodontics, School of Dental Medicine, University of Berne, Switzerland.

Production of undecalcified ground sections

Tissue samples from eight molars were fixed in 4% buffered formalin for 4 days and then processed for the production of undecalcified ground sections. Briefly, the specimens were rinsed in running tap water, dehydrated in ascending concentrations of ethanol, and embedded in methylmethacrylate. The embedded tissue blocks were then split horizontally into approximately thirty 200-µm-thick sections per root using a slow-speed diamond saw (Varicut® VC-50, Leco, Munich, Germany). For the histological and histometric analyses, four sections per root were selected, i.e. one section per root level. Sections that contained enamel (which in bovine teeth extends apical to the alveolar crest) and those that were located too far apically were excluded. Thus, for each root, four sections approximately the same distance from each another and devoid of artifacts were produced, resulting in a total of 64 sections. After mounting the sections onto opaque acrylic glass slabs, they were ground to a final thickness of about 100 µm, polished (Knuth-Rotor-3, Struers, Rodovre/Copenhagen, Denmark) and surface stained with toluidine blue/McNeal (Schenk et al. 1984). The stained ground sections were observed in a Leica M8 stereolupe and a Leica Dialux EB (Leica Microsystems, Glattbrugg, Switzerland). For polarized light microscopy, the ground sections were mounted onto transparent acrylic glass slabs before viewing in the stereolupe.

Production of decalcified semi-thin sections

Five molars were used for the production of semi-thin sections. Following horizontal sectioning of two mandibular first molars as described above, the tissue slabs were fixed for 72 h in 1% glutaraldehyde and 1% (para)formaldehyde, buffered with 0.08 m sodium cacodylate (pH 7.4). Following a brief wash in 0.1 m sodium cacodylate buffer with 5% sucrose, pH 7.4, the tissue samples were exposed to a decalcifying solution containing 4.13% disodium ethylenediamine tetraacetic acid (EDTA). Decalcification was performed at 4 °C room temperature for 3 months with constant stirring. The solution was changed twice weekly. Following decalcification, the tissue slabs were cut with a razor blade into approximately 50 slices per tooth in a corono-apical and vestibulo-oral direction. Following extensive washes in wash buffer solution, 10 slices were processed for tissue maceration with NaOH, a procedure that removes cells and noncollagenous extracellular matrix constituents but preserves collagen structure and position (Kuroiwa et al. 1992; Macchiarelli et al. 2002). Tissue slices were immersed in 10% NaOH under constant stirring for 4 days at room temperature. The NaOH solution was changed daily. Following extensive washes in wash buffer solution, the macerated and non-macerated slices were dehydrated in increasing concentrations of ethanol and embedded in LR White resin (Fluka, Buchs, Switzerland). Semi-thin sections (1 µm thick) were prepared using glass and diamond knives (Diatome, Biel, Switzerland) on a Reichert Ultracut E microtome (Leica), stained with toluidine blue, and observed in a Leica Dialux 22 EB microscope.

Size, number and area fraction of blood vessels

The histological slides were analyzed in a light microscope (Olympus AX 70) at a magnification of 10×. For each of the 6 regions (dd, db, dl, mm, mb, ml) and root levels (a–d), a set of 7–14 pictures (depending on the width and morphology of the PDL) was taken in digital format (640 × 512) to cover a 5 × 5 mm zone. Overlapping pictures of the total PDL surface were produced. Using a commercial imaging software (Photoshop, Adobe Systems, San Jose, CA, USA), the images obtained from each zone were then assembled to picture the entire circumference of the PDL (Fig. 3A). Next, all visible blood vessels (no less than 6 pixel in diameter) were filled with a white background color to allow their detection in the subsequent automated steps (Fig. 3B). The resulting images with white-filled blood vessels were then processed with a custom-made software (LabView, National Instruments, Oak Ridge, TN, USA). After the operator had manually declared a PDL area between the alveolar bone and the root surface as the region of interest (ROI), using threshold-based routines, the software calculated both the surface (in pixels) of the ROI and the total surface of the blood vessels. With the aid of a standard gauge, it was determined that 1 mm corresponded to 770.3 pixels in our system. Hence the program eventually provided the PDL and the blood vessel areas ([px²][mm²]), the blood vessel area fraction (percent of blood vessel total surface vs. PDL surface), the number of blood vessels contained in the ROI and the numerical density (number of blood vessels/PDL area [mm⁻²]). Last, the diameters of all blood vessels were extracted from the image. Since the cross-sections of the blood vessels did not appear as a perfect circles but mostly as ellipses (many blood vessels are not exactly normal in the transverse section), it was decided to set the diameters to the small axes of the ellipses.

Fig. 3.

One example showing consecutive pictures (A) before and (B) after all visible blood vessels in the periodontal ligament (PDL) were manually selected and filled with a white background color to ease their detection in the subsequent automated measuring process. AB, alveolar bone; C, cementum. A, B = ground sections.

Periodontal ligament width

The PDL width was assessed in four teeth from two animals for each root levels (a–d) and each regions (dd, db, dl, mm, mb, ml). Using the graphic software package, within each ROI, a line was drawn through the periodontal ligament. The length of the line was then converted to millimeters, and the PDL total surface calculated during the blood vessel measurements was divided by the line length to yield the PDL width.

Boundaries between PDL and bone and between PDL and cementum

The fractal dimensions of the two boundary tissues were assessed to quantify the tortuous windings of their paths. The following expressions were used to obtain the fractal dimension of the selected specimens (Krappraff, 1986; Borodich, 1997):

(1)

L(ɛ) is the length of the line obtained when using the measuring unit (or scale) ɛ. K and D are parameters. L(ɛ) is experimentally obtained by

(2)

where N(ɛ) is the number of times that the unit ɛ fits into de measured curve.

The real length L of the curve is thus defined as

(3)

It has been empirically determined that the numbers N(ɛ) satisfy the law

(4)

from which relation (1) can be derived.

The exponent D in equation (1) is referred to as the fractal dimension of the line. When D ≈ 1 the boundary is smooth, whereas for larger Ds the tortuousness of the border is increased.

D is experimentally determined by plotting L(ɛ) vs. ɛ on logarithmic x- and y-scales and then measuring the slope of the line obtained (slope = 1 – D).

In the present study, four different specimens per location were examined. ɛ was taken equal to 1, 1/3, 1/9 and 1/27. These unit gauges were manually fitted to the intricacies of the bone and cementum boundaries. The number of times (N) that the gauge fitted the curve was recorded.

Results

Descriptive histology

For all root levels, the distal root of the mandibular first molar was wider than the mesial root (Fig. 2A,B). Both bone thickness and density varied according to site. Irrespective of the root depth, the buccal and lingual bone of the alveolar process consisted of thick cortical bone plates, whereas the interdental bone septum was extremely thin, sometimes even non-existent. The interradicular bone septum was trabecular and porous – the porosity increasing in the apical direction. The PDL contained a variable number of large blood vessels (Fig. 3A) and was rich in cells and fibers (Fig. 3B). The principal PDL fibers were anchored as Sharpey's fibers inside the alveolar bone (Fig. 4A,C) and the cementum (Fig. 4B,D). Tissue maceration prior to embedding in LR White resulted in the removal of the cells, leaving the collagen bundles intact (Fig. 4C,D). Due to active or previously active remodeling activity, fiber insertion into bone was not present at all sites (Fig. 5A,B). In general, the morphology of the alveolar bone was indicative of high bone turnover. In contrast, the remodeling activity in the cementum appeared to be lower than in the alveolar bone (Fig. 5A,B). Resorption lacunae were hardly detected along the cementum surface (Fig. 5C). Soft tissue channels, however, were frequently observed in the cementum, particularly in the apical root portion (Fig. 5D). These soft tissue channels contained blood vessels. The presence of Howship's lacunae without (Fig. 5E) and with (Fig. 5F) new cementum deposition was indicative of some remodeling activity in the deeper layers of the cementum. Along the entire root length, the cementum was of a mixed type, that is, containing lacunae with both cementocytes and Sharpey's fibers (Fig. 5B,E,F).

Fig. 4.

Thin sections without (A,B) and with (C,D) prior maceration illustrating the insertion (arrows) of principal periodontal ligament (PDL) fibers into the alveolar bone (AB) and the cementum (C). The maceration process has removed the cells and some extracellular matrix constituents but not the collagen fibers.

Fig. 5.

Ground (A, C–F) and thin (B) sections illustrating the remodeling activity of the alveolar bone (AB) and the cementum (C). At most sites (A, B, D), the surface contour was smoother in cementum than in bone. The irregular and jagged surface contour of the alveolar bone, the marked differences in the staining intensity between bone matrix compartments (A), and the presence of osteoclasts (arrows) were indicative of a high bone remodeling activity. In contrast, the cementum surface revealed only sparse resorption lacunae (C). However, soft tissue channels (asterisks) were frequently observed in the apical cementum (D). These soft tissue channels contained blood vessels (BV). The presence of both Howship's lacunae (arrowheads) and newly deposited cementum matrix (NC) indicated remodeling activity in the depth of the cementum layer (E, F). D: dentin; PDL: periodontal ligament.

Figure 6 shows representative micrographs illustrating the six locations investigated. All six revealed large variations in PDL width, area fraction of blood vessels, number of blood vessels, amount and orientation of collagen fiber bundles, and surface contour of both alveolar bone and cementum. A qualitative assessment of the six locations at four different root depths showed the following: (1) In location dl (Fig. 6A), the cementum and bone contours were rather jagged and undermined. Blood vessels were large and numerous; many of them were found in indentations of the mineralized tissues. Some principal PDL fibers were functionally oriented. (2) In location dd (Fig. 6B), the cementum and bone contours appeared in variable conrations. Blood vessels were numerous but, in comparison with location dl, fewer were functionally oriented. (3) In location db (Fig. 6C), the cementum contour was variable and the bone contour was often jagged. Blood vessels were numerous where the PDL was thin and less numerous in zones where the PDL was thick. Many of the blood vessels were located in hard tissue indentations and a sizeable amount of the collagen fibers were oriented in parallel bundles, particularly at wide PDL sites. (4) In location ml (Fig. 6D), the cementum and bone contours, the number of blood vessels, and connective tissue fiber orientation were variable. (5) In location mm (Fig. 6E), the cementum contour was flat and the bone contour was a little jagged. The number of blood vessels was variable and many collagen fibers were functionally oriented. (6) In location mb (Fig. 6F), the cementum and bone contours were essentially flat, the number of blood vessels was variable, and some of the collagen fibers were functionally oriented.

Fig. 6.

Ground sections illustrating the six locations investigated: (A) dl, (B) dd, (C) db, (D) ml, (E) mm, and (F) mb. Large variations in periodontal ligament (PDL) width, area fraction of blood vessels (BV), number and size of blood vessels, amount and orientation of collagen fibers, and surface contour of both the alveolar bone (AB) and the cementum (C) may be seen.

Histomorphometry

Blood vessel surface

The area fraction of blood vessels is presented in Fig. 7. The lowest and highest absolute values for the area fraction of blood vessels were 7.0% and 34.2%, respectively, and the values pooled per region varied from 11.9% (dl) to 23.5% (dd). anova and Scheffe's analyses revealed a significant (P < 0.05) difference between locations, that is, the more coronal the section, the larger the area fraction of blood vessels. The lowest and highest absolute numbers of blood vessels per mm² were 26 and 112, respectively, and the data pooled per region ranged from 28.0 (dl) to 49.72 (ml). The blood vessel diameters at the buccal locations were significantly greater (P < 0.05) than at all other locations, and the smallest blood vessels were found at the lingual location of the distal root.

Fig. 7.

Percent area fraction of blood vessels for the four root levels, a (apical) to d (coronal).

PDL width

The PDL widths are presented in Fig. 8. The mean PDL width for all 95 measurements was 551±125 µm (mean ± SD) with the lowest and highest values measuring 221 µm and 785 µm, respectively. For all locations, there was an increase of the PDL width from coronal to apical.

Fig. 8.

Periodontal ligament width for the four root levels, a (apical) to d (coronal).

Fractal dimensions of the cementum and bone surfaces

The fractal dimensions of the cementum and bone contours are shown in Table 1. The fractal dimensions of the cementum and alveolar bone ranged from 1.01 to 1.10 and from 1.06 to 1.22, respectively. For all locations except ‘dd’ the fractal dimensions of the bone contour were higher than those of the cementum surface.

Table 1.

Fractal dimensions (D) for cementum and bone at the six peripheral locations

Location	Cementum contour	Bone contour (#specimens)
dL (4)	1.10±0.05	1.22±0.08
dd (4)	1.09±0.06	1.06±0.03
db (4)	1.08±0.05	1.18±0.05
mL (4)	1.05±0.02	1.07±0.04
mm (2)	1.01±0.02	1.14±0.04
mb (4)	1.01±0.01	1.04±0.01

Polarized light microscopy

Polarized light microscopy was performed to better visualize the density and architecture of the PDL collagen fiber bundles. The qualitative assessment showed that both fiber density and orientation varied greatly among sites (Fig. 9). In general, fibers were more densely packed and displayed a higher degree of orientation at root levels b, c, and d (Fig. 9A,B), whereas the apical PDL (i.e. at level a) was wider and contained fewer and less organized connective tissue fibers (Fig. 9C,D).

Fig. 9.

Ground sections viewed under polarized light showing non-uniform fiber density and orientation. The periodontal ligament (PDL) fiber system connecting the cementum (C) with the surrounding alveolar bone (AB) was either straight (A), angulated (B), or arranged in a criss-cross pattern (C, D). In contrast to the root levels b–d (A, B), the apical periodontal ligament presented a low fiber density and a lower degree of fiber organization (C, D). D, dentin.

Discussion

The present study is part of a research line aimed at characterizing the mechanical characteristics of PDL (Pietrzak et al. 2002; Pini et al. 2004; Justiz, 2004; Shibata et al. 2006). The testing principle consists in loading samples under well defined conditions, that is, by controlling 1) the origin and structure of the tissue, 2) the geometry of the specimen, 3) the mechanical preconditioning of the sample, 4) the load profile applied to the tissue and 5) the physical environment (moisture, atmospheric pressure). The data sets thus generated then provide the basis for equations systems (i.e. the constitutive laws) which simulate the response of the three compartments controlling the dynamics of a tooth under load, that is, 1): the fibrous component (collagen and elastic fibers), 2) the ground substance and 3) the vasculature. Ideally, such equation sets should be ‘robust’ in that the equations that model the behavior of the tissue should be controlled by parameters that reflect the structure and properties of the tissue (and should not be written as any ‘best fit’ function). Further, for a given tissue type, the basic response should be modeled within the equation system. Refinements pertaining to species should be obtained by adjusting tissue-specific parameters.

The drawback of this approach is that it discounts the stabilizing effect of the vasculature and does not (as yet) establish a link between local tissue response and global tooth movements. Both these issues have been discussed in previous reports (Pini et al. 2002; Sanctuary et al. 2005; Sanctuary et al. 2006).

Bovine vs. human teeth

The PDL is part of a multicomponent unit forming the tooth attachment apparatus. Therefore, a functional characterization of the PDL cannot be confined to the soft tissue but must take the two bordering mineralized tissues, that is, the alveolar bone and the cementum, into consideration. For the first time, the present study characterized the PDL of the bovine mandibular first molar both morphologically and morphometrically. The various histological techniques used provided a comprehensive insight into the complex structural organization of the tooth attachment apparatus in bovine mandibular first molars and demonstrated marked species-specific particulars and differences with human tissues.

The number of fibers, their thickness and density, structural organization and insertion modus into both alveolar bone and cementum are all parameters that may affect the attachment function. At a number of sites, the periodontium of bovine teeth shows histological signs of active or post-active remodeling activity and suggests that the local attachment function is weakened at such sites. The higher fractal dimensions of the bone surface compared with the cementum support the qualitative observations of a higher remodeling activity in the alveolar bone, since an increase in the number of resorption lacunae results in a more irregular surface profile. If the remodeling activity is confined to tiny spots, the overall attachment function may not be affected. However, no predictable distribution pattern of resorptive activity could be deduced from the present material. When compared to bovines, human teeth have a lower incidence of remodeling activity. This observation is probably related to different mastication patterns as cattle are ruminant herbivores with extreme horizontal tooth displacements during grinding. Although most remodeling activity was observed in the alveolar bone, the root cementum also showed histological signs of moderate remodeling activity, albeit to a much lesser degree than in the alveolar bone. The permanent eruption of bovine molars results in the formation of thick layers of cementum on the roots and the cervical portion of the crown (Ainamo, 1970). This continuous process, which requires rapid cementum deposition, may contribute to the high remodeling activity in the bovine periodontium. An interesting finding was that the cementum, particularly in the apical root portion, contained soft tissue channels with blood vessels. This is in contrast to human teeth which are not known to present vascularized soft tissue channels inside the cementum (for a review, see Bosshardt, 2005). It suggests that cementum in the apical root portion has more bone-like features in bovine than in human teeth. The presence of blood vessels may enable more cementocytes to survive in the deeper layers and/or to play a role in cementum remodeling.

Site-specific characteristics of the bovine attachment apparatus

In contrast to the surface-stained ground sections viewed under bright field illumination, polarized light microscopy allowed a more detailed visualization of the PDL connective tissue fibers. These images demonstrated a high degree of randomness in terms of fiber density and orientation ranging from compact and highly organized architectures to rather low fiber density and organization. In the most apical root segment, fiber density was lowest and its architecture showed the least degree of structuring. There was one site, however, which quite consistently demonstrated a high degree of fiber organization. This site was located at the bucco-mesial aspect of the distal root at levels b, c and d. Since the fibers are predictably organized and the buccal bone plate is very thick, these sites may be regarded as the most suitable locations for harvesting samples for mechanical testing.

A remarkable difference between the bovine mandibular first molar and human teeth is the distribution pattern of types of cementum. While the coronal half of human roots is covered with acellular extrinsic fiber cementum, and the apical root portion and the furcation is covered with cellular mixed stratified cementum, there was only one type of cementum found in bovine teeth, that is, a mixed type consisting of embedded cells (i.e. cementocytes) and Sharpey's fibers. This has implications on stress distribution and dissipation around the roots as, due to its high number of inserting Sharpey's fibers, acellular extrinsic fiber cementum is a major contributor to the attachment function (Bosshardt & Selvig, 1997). In general, in the bovine molar, there appears to be a lower but more homogeneous number of fibers inserting into the cementum along the root.

The maceration of the tissue samples removed cells and parts of the noncollagenous proteins but leaving the collagen structure intact. Thus far, the NaOH maceration has been used to visualize fiber architecture in scanning electron microscopy. To our knowledge, this is the first time that maceration was used in combination with tissue embedding, sectioning and optical microscopy. This methodology may be applied in future studies to assist in calculating the collagen fiber area fraction in soft connective tissues using computer-aided image processing.

The measurements of size, number and area fraction of blood vessels showed values that varied greatly from one location to another. The higher area fraction of blood vessels in the coronal PDL portion may be explained by a contribution of branches of the interalveolar and interradicular arteries, which extend coronally inside the spongiosa of the alveolar process (Schroeder, 1986).

Like the blood vessel measurements, large variations were noted in PDL width. All measurements, however, were clearly superior to the PDL width of human teeth, which is in the order of 100–300 µm (Coolidge, 1937; Schroeder, 1986).

The highest values in the bovine thus are two to three times higher than in humans. It is known that human teeth respond to increased functional loads by a widening of their PDL space (Schroeder, 1986). Therefore the greater PDL width of the bovine teeth may have been caused by higher functional loads in herbivores. In human teeth, the PDL is always thinnest at the mid-level of the root (Schroeder, 1986) whereas in the present study a consistent increase in PDL width from coronal to apical was noted. This observation implies that samples for biochemical testing should be extracted from similar root levels (although large differences were observed between animals as well).

Conclusions

The present study provided the baseline information on the tooth attachment apparatus in bovine mandibular first molars which should assist in planning and conducting future biomechanical studies of the PDL. The histological observations and the histomorphometric data indicate that the morphology of the PDL is highly inhomogeneous around a given tooth and between teeth without predictable pattern between sites and animals. While the most apical root portion is not recommended for harvesting samples for biomechanical tests, the buccal and lingual aspects of the mesial and distal root – along with the bucco-mesial aspect of the distal root – appear to be the most suitable locations.