Elastic fibres have been revealed by both elastin immunostaining and conventional histological orcein-staining in the intervertebral discs of the bovine tail. These fibres are distributed in all regions of the disc but their organization varies from region to region. In the centre of the nucleus, long (> 150 µm) elastic fibres are orientated radially. In the transitional region between nucleus and annulus, the orientation of the elastic fibres changes, producing a criss-cross pattern. In the annulus itself, elastic fibres appear densely distributed in the region between the lamellae and also in ‘bridges’ across the lamellae, particularly in the adult. Elastic fibres are apparent within the lamellae, orientated parallel to the collagen fibres of each lamella, particularly in the young (12-day-old) discs. In the region between the disc and the cartilaginous endplate, elastic fibres appear to anchor into the plate and terminate there. The results of this study suggest that elastic fibres contribute to the mechanical functioning of the intervertebral disc. The varying organization of the elastic fibres in the different regions of the disc is likely to relate to the different regional loading patterns.
The principal functions of the intervertebral discs (IV-disc) are mechanical: the discs are always under mechanical load arising principally from muscle activity and ligament tension (Nachemson & Elfstrom, 1970). They thus carry and transmit compressive load throughout the spinal column as well as providing the flexibility that allows the spine to bend and twist (White & Panjabi, 1978). The IV-discs consists anatomically of three distinct regions, the nucleus pulposus, the annulus fibrosus and the cartilaginous endplate. The nucleus is located in the centre of the disc and enclosed laterally by the annulus and vertically by the endplate (Roberts et al. 1989).
The main macromolecular components of the IV-disc are the fibrillar collagens, type I and type II collagen and the large aggregating proteoglycan, aggrecan. However, the disc also contains many minor components such as the small proteoglycans decorin, lumican and biglycan and minor collagens such as collagens III, VI and IX (Roberts et al. 1991; Johnstone et al. 1993; Urban & Roberts, 1997; Sztrolovics et al. 1999). The composition and the organization of these macromolecules differ markedly across the regions. The nucleus is rich in aggrecan and highly hydrated, with a fine collagen network which shows no apparent organization (Szirmai, 1970; Takeda, 1975; Inoue, 1981). The annulus contains less aggrecan than the nucleus and consists predominantly of sheets made from bundles of collagen fibres, which form concentric, cylindrical lamellae around the spinal axis. The angle of the collagen bundles in the lamella alternates between adjacent lamellae, thus forming a cross-woven and reinforced structure (Coventry et al. 1945; Hukins, 1984; Marchand & Ahmed, 1990).
The ability of the disc to bend or extend is thought to depend on the organization of the lamellae which are only loosely interconnected (Takeda, 1975) and can move independently (Szirmai, 1970). Flexion and extension of the spine can result in large deformations of the disc (Coventry et al. 1945) with disc height increasing by up to 60% (Pearcy & Tibrewal, 1984). To ensure adequate recovery of the network organization after those deformations would seem to require the existence of some component with elastic properties in the disc. In most tissues, for example skin, lung, arteries and elastic ligament, such elastic properties are provided by elastic fibres made of elastin and its associated microfibrils, such as fibrillin. The presence of elastic fibres in the IV-disc has been noted in some earlier studies (Buckwalter et al. 1976; Hickey & Hukins, 1981; Johnson et al. 1982, 1984, 1985; Mikawa et al. 1986) but the large-scale organization of the fibres has hitherto not been examined. Therefore, the objective of this study was to identify the distribution and organization of the elastic fibres in the disc in order to provide insight into its possible mechanical roles.
Here we report on the elastic fibre organization in the upper bovine caudal discs. We used these discs because both young and adult healthy discs are readily available and because they resemble human discs in many respects. Their cell phenotype (Horner et al. 2002), their composition, appearance and aspect ratio (Ohshima et al. 1993) and their loaded pressure profiles (Ishihara et al. 1996) are all similar to those of human lumbar discs. Since the compressive loads on the human disc arise mainly from muscle forces (Nachemson & Elfstrom, 1970), it is not surprising that these upper caudal discs, surrounded by large muscles, are also under significant compressive resting loads (Ohshima et al. 1993, 1995).
Materials and methods
Bovine tail discs were obtained from five adult steers (18–24 months old) and five calves (12 days old) within 24 h of slaughter. The four uppermost discs were dissected from the adjacent vertebral bodies. Sagittal or transverse slices (about 5–10 mm in width or length) were cut by hand with a scalpel, snap frozen and stored at –70 °C until used. Frozen sections (30 µm) were cut with a cryostat microtome and mounted on poly lysine-coated microscope slides (BDH, Poole, UK; Cat No. 406/0178/00), then air-dried and stored at room temperature until used. Details of dissection are schematically illustrated in Fig. 1.
Specimen pretreatment with hyaluronidase and collagenase
Tissue sections were rehydrated with phosphate-buffered saline (PBS) and fixed with 10% formalin in PBS for 10 min. In order to visualize the elastic fibres it was necessary to remove glycosaminoglycans (GAG) and collagen before staining, since both were found to have a masking effect on elastic fibre staining. Thus fixed sections were treated with hyaluronidase (Sigma Chemical Co.,: Cat No. H6254, 4800 U mL−1) overnight at room temperature to remove glycosaminoglycans partially. All sections (except calf disc nucleus sections) were further treated with purified collagenase (CN Biosciences Ltd, Nottingham, UK; Calbiochem. Cat No. 234134) for 30 min (nucleus samples) or 2 h (annulus samples) at 37 °C to remove collagen partially.
Orcein staining is one of the conventional histological methods for demonstrating elastic fibres. Fixed and enzyme-treated sections were kept in 70% ethanol for 5 min before moving into the orcein staining solution (1% orcein, Sigma Chemical Co.: Cat No. 07505 + 1% concentrated HCl in 80% ethanol; Drury & Wallington, 1980). After 40–50 min staining at room temperature, slides were washed with 70% ethanol and then with tap water for approximately 5 min. Serial ethanol solutions (70% × 1; 90% × 1; 100% × 2) were used to dehydrate the sections. The sections were then washed twice with xylene and mounted with DPX (BDH; Cat No. 361254D) mounting medium.
Fixed and pretreated sections were first incubated in 0.3% hydrogen peroxide in PBS for 30 min at room temperature in order to block endogenous peroxidase. After washing with PBS three times, the sections were then treated with normal goat serum (Vector Laboratories Ltd, Peterborough, UK: Vector elite ABC kit, Cat No. pk610; 1:50 diluted with 3% bovine serum albumin in PBS) for 20 min to block the non-specific binding. The sections were then incubated with a rabbit polyclonal anti-alpha human-elastin antibody (Biogenesis, Poole UK: Cat No. 4060-1054) at room temperature for 2 h. Adjacent sections were incubated with PBS or normal rabbit serum as the negative controls. After washing with PBS three times, the sections were then incubated with secondary biotinylated antirabbit goat IgG (Vector Laboratories Ltd: Vector elite ABC kit, Cat. No. pk6101) at room temperature for 30 min and then washed a further three times in PBS. The elastic fibres were visualized by using the streptavidin–biotin detection system (Vector Laboratories Ltd; Vector elite ABC kit, Cat No. pk610 1) and the substrate of diaminobenzidine tetrachloride (DAB) (Sigma Chemical Co.: Cat No. D8001). The sections were then dehydrated and mounted as described above.
General collagen organization
The use of polarized light demonstrated an organized collagen network forming concentric lamellae in the annulus region (Fig. 2a). The concentric collagen lamellae in the outer annulus (OA) showed more distinct boundaries than those of inner annulus (IA), suggesting that the collagen network in the OA is more densely organized than in the IA, in line with previous biochemical analyses of the disc (Eyre & Muir, 1976; Oshima et al. 1993). There was no apparent organized collagen network in the nucleus pulposus (N).
Elastic fibre network
Elastic fibres were observed in different regions of calf disc using both conventional orcein staining and elastin immunostaining. These two techniques gave a similar pattern of staining and demonstrated the elastic network throughout the disc.
Both orcein staining (Fig. 2b) and elastin immunostaining (Fig. 2c) revealed long (> 150 µm) and straight elastic fibres appearing to run parallel across the 30-µm-thick nucleus sections; the fibres could be considerably longer in the tissue than in the essentially two-dimensional section. These fibres appeared to have a defined orientation, running radially from the nucleus centre towards the disc periphery as shown by the arrows in Fig. 2(a–c). In addition, some fibres could be seen in cross-section (Fig. 2b,c: black arrow), suggesting that there is an organized network of elastic fibres in the nucleus running both radially from the centre of the nucleus towards the annulus and also running vertically from endplate to endplate.
(ii) Inner annulus
Moving from the central nucleus region towards the IA region, both orcein staining (Fig. 2d) and elastin immunostaining (Fig. 2e) showed that the organization of the elastic fibres changed; here fibres were no longer primarily radially distributed but formed a criss-cross pattern. Cross-sections of fibres were also apparent (Fig. 2d, black arrow) as in the nucleus (Fig. 2b), suggesting that some vertical elastic fibres were also present.
(iii) Outer annulus
In the region between the IA and OA, elastic fibres no longer appeared to run in a criss-cross pattern, but were orientated parallel to the collagen fibres within each lamella (Fig. 2f,g). With hyaluronidase pretreatment alone, only a few elastic fibres were visible within the lamellae (Fig. 2f). After pretreating the specimen with collagenase, orcein staining revealed an abundant elastic fibre network previously masked by collagen within the lamellae. Elastic fibres were found to run predominantly parallel to the collagen fibres within the outer annulus lamella (Fig. 2g).
General collagen organization
The sagittal sections viewed with polarized light demonstrate collagen lamellae running vertically from one growth plate to the next in the outer annulus while forming a lateral bulge in the inner annulus (Fig. 3a). No collagen organization was visible in the growth plate itself. In the nucleus, oval, concentric layers around the central nucleus were apparent.
Elastic fibre network (i) Nucleus
In the lateral region of the nucleus (Fig. 3b), elastic fibres appeared shorter (< 50 µm) than those seen in transverse sections of the disc (Fig. 2b,c); their predominant direction was radial but the fibres bulged towards the inner annulus in a similar manner to the layers of collagen which also bulged outwards towards the inner annulus (Fig. 3a). In the vertical region of the nucleus (Fig. 3c), elastic fibres appeared shorter but almost parallel to each other in the section, forming vertical layers running towards the central nucleus. The organization of elastic fibres in layers appears similar to the layers of collagen seen around the nucleus under polarized light (Fig. 3a). In the central nucleus region (Fig. 3d), elastic fibres appear long (> 200 µm), straight and parallel to each other in the 30-µm section. The elastic fibres here were radially distributed towards the central nucleus, corresponding to the pattern seen in the transverse section (Fig. 2c).
(ii) Growth plate
Figure 3(e) shows that the elastic fibres of the disc anchor into growth plate and terminate there; no elastic fibres were observed in the cartilaginous growth plate itself. The dark brown spots (orcein-stained) apparent in the GP region (black arrow) are possibly the stained nuclei of growth plate chondrocytes.
In the annulus region, elastic fibres appeared most densely distributed between the lamellae, running in general parallel to the direction of the lamellae (Fig. 3f,g). Fine fibres were also seen within the lamellae but here appeared to lie perpendicular rather than parallel to the direction of the lamellae (Fig. 3g). The elastic fibres between the lamellae were thicker than within the lamellae.
In general, the size of a bovine adult disc is about 2–3 times larger than that of the 12-day calf disc from the same spinal level. The adult disc had thicker lamellae than the calf. However, the organization of the lamellae of the adult disc was similar to that of calf disc except elastin cross-bridges within lamellae were frequently apparent. These bridges crossed the lamellae vertically and also obliquely and were observed in both transverse and sagittal sections (Figs 4e and 5f).
Transverse-section (i) Nucleus
Elastic fibres (< 50 µm) appeared to be straight and parallel in the section and mostly were seen running radially from the central nucleus region (Fig. 4a,b). The elastic fibres were mostly shorter than those found in similar sections of the calf disc (Figs 2c and 3d) and appeared to form distinct layers. In the outer region of the nucleus, the direction of fibres was sometimes random (Fig. 4c); also, fibres in criss-cross pattern, but finer than those seen in the calf disc, were often apparent (Fig. 4d). In some sections (Fig. 4c,d, black arrow), immunostained spots were visible. From their size and appearance, they are likely to be nucleus cells (Errington et al. 1998).
Elastic fibres were most densely distributed between the lamellae in the adult, as found in the calf disc. However, in the adult disc, dense cross-bridges of elastic fibres, running perpendicular to and across the lamellae, were apparent (Fig. 4e). Elastic fibres of the cross-bridge were not regularly distributed. They appeared long (> 50 µm) (Fig. 4f), but were sometimes shorter and criss-crossed (not shown). Within the lamellae, long elastic fibres (> 200 µm) were apparent, particularly in the outer lamellae, running parallel to the collagen fibres (Fig. 4g). In the inner annulus, the elastic fibres within the lamellae appeared finer and shorter than in the outer region (Fig. 4h) and than those lying between the lamellae (Fig. 4h).
Sagittal section (i) Nucleus
In the lateral region of the nucleus towards the annulus, elastic fibres appeared to be vertically orientated and parallel but were also seen to bulge laterally outwards towards the annulus (Fig. 5a). In the central region of the nucleus, elastic fibres were parallel to each other within the section and were orientated towards the midpoint of the central nucleus (Fig. 5b). In this region the elastic fibres appeared to be organized in layers in a similar manner to those in the calf disc (Fig. 3c). In the region between the end plate and the disc, elastic fibres were parallel and vertically orientated towards the endplate (Fig. 5c). Stained cells were also apparent (black arrow).
As seen in the calf disc, elastic fibres were most densely located in between the lamellae (Fig. 5d). There were also dense cross-bridges crossing the lamellae (black arrow, Fig. 5e).
In this study, we found elastic fibres were distributed through all the regions of the IV-disc of the bovine tail. Although elastic fibres have been seen in the disc in some earlier studies (Buckwalter et al. 1976; Johnson et al. 1982, 1984, 1985), details of their organization have not been described previously, possibly because they were masked by the dense GAG and collagen matrix. It is apparent from the results of this study that the network is very extensive and its organization differs between disc regions and changes with age. A detailed stereological investigation, beyond the scope of this study, is required to describe the three-dimensional structure of the network fully. However, some understanding of its organization was obtained from the work described here. We found that, in contrast to the random organization of the collagen network described by others (e.g. Inoue & Takeda, 1975), the nucleus contained a highly organized network of elastic fibres. Elastic fibres appeared long, straight and orientated radially in the central nucleus (Fig. 2b,c); these radial fibres were also intersected by elastic fibres orientated vertically (Fig. 2b), possibly running from the nucleus to the endplate where they appeared to anchor (Fig. 3e). The organization of the radial fibres changed towards the annulus to form a criss-cross pattern at the nucleus annulus junction (Fig. 2d,e) suggesting the nucleus was enclosed by a mesh of elastic fibres. In the annulus, elastic fibres appeared most densely clustered between the lamellae and were orientated predominantly parallel to the collagen fibres within the lamellae (Fig. 2g). In the adult disc, the lamellae were interrupted by dense cross-bridges composed of elastic fibres (Figs 4e and 5e). A schematic illustration of the elastic fibre network is shown in Fig. 6. Initial studies indicate a similar organization of elastic fibres exists in the annulus of human lumbar discs (not shown).
It is likely that the organization of the elastic fibre network shown here influences the mechanical behaviour of the intervertebral disc. Elastic fibres have been found present in the extracellular matrix of many tissues such as skin, lung, arteries and elastic ligament. They occur in various forms in different tissues, such as concentric sheets or lamellae in arteries, rod-like fibres in elastic ligament and as a three-dimensional meshwork of fine fibres in the elastic cartilage of the ear (Cleary & Gibson, 1983, 1996). Different forms of elastic fibre organization appear to be related to the specialized functions of elastin in the various tissues.
The organization of the elastic fibre network in the disc suggests that it too has a specialized mechanical role though at present we can only speculate on its function. The elastic network running through the nucleus and the elastic fibre meshwork enclosing the nucleus may contribute to maintaining its structure. The network of long straight radial fibres running through the nucleus may also be related to the ability of the gel-like nucleus to transmit load to all the parts of the annulus equally (Nachemson, 1960; Szirmai, 1970). Thus, the elastic fibres of the nucleus may be capable of being stretched radially when the load is transmitted to the annulus and then contribute to the recovery of the original network conformation after deformation. Elastic fibres throughout the nucleus in the plane perpendicular to the adjacent vertebral bodies could help realignment of the network after flexion, torsion or extension. In the annulus, the organization of the elastic fibres, which are concentrated at the region between the lamellae, may also relate to realignment after load-induced deformation of the disc. The change in disc height on extension or flexion far exceeds the maximum extensibility of collagen fibrils (Viidik, 1973) and is thought to depend on slip between the adjacent collagen lamellae (Szirmai, 1970; Takeda, 1975). High concentrations of hyaluronan have been observed between the lamellae (Inkinen et al. 1999); it has been suggested that hyaluronan contributes to this slip by acting as a lubricant. The elastic fibres lying between the lamellae and anchored within them (Fig. 4e) may also contribute, particularly by making this slip reversible and hence may enable recovery of annular lamellae organization after deformation. The elastin bridges crossing the lamellae apparent in the older discs and which have hitherto not been described to our knowledge (Figs 4e and 5e) may also contribute to network flexibility.
It has been reported that elastin makes up approximately 1.7% of the dry weight of both the annulus and the nucleus of human intervertebral discs (Mikawa et al. 1986). Previously it has been difficult to see how elastic fibres have a significant mechanical role with elastin making up such a small proportion of the total tissue matrix. However, it is apparent from this study (e.g. Figs 3f and 4e) that the fibres are not evenly distributed and in some regions, such as in the interlamellar space and in the cross-bridges, their concentration appears relatively high. Thus, here they could play a mechanical role in, for instance, recovery of lamellar organization after load-induced deformation of the collagen–elastic network. In addition, elastic fibres may connect physically through cross-weaving or chemically cross-linking with other matrix components, particularly collagen. Indeed, elastic fibres seen parallel to the collagen fibres within lamellae (Fig. 2g) indicate the close relationship between these two components. Thus, with such coupling, even a low density of elastic fibres could play a significant mechanical role.
This work was supported by the Wellcome Trust and the Arthritis Research Campaign (U0511).