- Email: [email protected]
Collagen fibre orientation in bovine secondary osteons by collagenase etching
Collagenfibre orientationin bovine secondaryosteonsby collagemseetching M.Green". D.H. Isaac andG.M. Jenkins Department of Materials Engineering, University College of Swansea, Swansea SA2 8PP, UK (Received 5 August 1986; revised 29 April 1987)
The orientation of collagen fibres in bovine secondary osteons has been investigated in the scanning electron microscope (SEM) by removal independently of firstly the mineral component and secondly the collagen fibres. Demineralization of polished transverse sections reveals a lamellar structure for the collagen component but the precise orientation of the collagen in each ring is not unequivocably determined. However, by using a collagenase solution to etch away the collagen component of a polished surface, holes are produced in the mineral revealing the former position of the fibres. The greater rigidity of the mineral component ensures that the structure does not collapse and produce artifacts. A specimen cut so that transverse and longitudinal sections are simultaneously observed allows the relationship between the structural features on each surface to be revealed. Analysis of such micrographs indicates a model for the collagen component of osteons in which the lamellar structure contains fibres with orientations alternately parallel to and circumferential to the long axis of the osteon. Tilting the samples to look directly down the holes shows that the fibres are not precisely longitudinal and circumferential but are tilted from these ideals by a variable angle (typically 20”) the precise angle probably being an important factor related to the in vivu mechanical property requirements. Keywords: Bone, collagen, bone microstructure, SEM. collagenase etching
The spatial arrangement of the collagen fibres in bone has long been a topic of investigation. Kolliker’ expressed the belief that the organic matrix of adult bone was the embryonic skeleton in which calcium salts had been deposited. He further stated that this organic interstitial substance was granular in nature rather than fibrillar. Von Ebnerz, held the view that the spaces between the fibrillar bundles were filled with a substance which contained the mineral salts of the bone, while the fibrils themselves remained noncalcified. Ruth3 examined thin sections of human femoral bone which had been decalcified and treated with various histological stains. He investigated the lamellar structure of osteons and designated the two types of lamellae observed to be compact or diffuse, in accordance with their structural appearance. The compact lamellae were interpreted as bands of circumferentially oriented fibrillae, while the diffuse lamellae were said to be bands of radially oriented fibrillae which were loosely disposed and separated from each other by relatively wide inter-fibrillar spaces filled with a granular substance, namely the mineral. Smith4 also examined sections of the human femur and suggested that the dark lamellae in transverse section contained predominantly circumferential fibres, and corresponded to the light lamellae in the radial sections in which these fibres were cut at right angles. Similarly, he considered ‘Present address: Department of Mechanical and Manufacturing Systems Engineering, UWIST, Cardiff, UK. 0 1987
the dark lamellae in radial sections contained predominantly longitudinal fibres and corresponded to the light lamellae in the transverse sections in which these fibres were again cut at right angles to their long axes. Cooper et a/.5 noted similar features in thin sections of secondary osteons from mongrel dogs. Ascenzi and Bonucc? proposed that osteons could be divided into three categories depending on their appearance under crossed polarized light. The categories were: (1) dark, in which the collagen fibres were oriented parallel to the long axis of the osteon; (2) light, in which the collagen fibres were circumferential in relation to the long axis of the osteon; and (3) intermediate, a mixture of the two. Boyde and Hobdel17 using a scanning electron microscope (SEM) suggested that lamellae were fibrous plates about 3 pm thick separated by inter-lamellar cement bands about 0.1 pm thick, but they found that the fibre direction in a given lamella was very variable. Boyde’, in an extensive review article, introduced the concept of domains for the collagen fibres, a domain being a volume over which the collagen fibre bundle orientation is essentially parallel. His work on human adult femoral bone showed that domains are usually longer along the length of the fibre bundles (typically -8Opm) and narrowerwhen measured across the fibre axis (typically 2030 pm). Black et a/.’ suggested that only the dark osteons proposed by Ascenzi and Bonucci’ were in fact present in bone, the other types being artifacts resulting from the fact that it was almost impossible to be sure that any osteon
Butterworth 8 Co (Publishers) Ltd. 0142-9612/87/060427-06$03.00 Biomaterials
198 7. Vol8 November
427
Bone microstructure: M. Green et al.
under examination had been sectioned at 90” to its long axis. They argued that if the section was not cut at exactly 90” to the long axis of the osteon then fibres parallel to the long axis would not be perpendicular to the plane of the section and would therefore be considered to be part of a spiral around the central canal. Ascenzi and Bonucci” used a series of mechanical tests to support their opinion that there were three types of osteon. Frasca et al.” examined single isolated human osteons which had been mechanically manipulated to expose their inter-lamellar interfaces. They concluded that only the dark osteons proposed by Ascenzi and Bonucci’,” actually conformed to the initial model, the light and intermediate osteons being more complex than first thought. They compared the results of the examination of osteonic specimens undercrossed polarized light following mechanical manipulation” , and concluded that osteons which appeared dark under crossed polars contained fibres with little or no transverse component, whereas light osteons possessed fibre orientations with transverse and longitudinal components. Intermediate osteons were found to contain fibre orientations richer in longitudinal component than in transverse. Frasca et a1.‘3. from X-ray diffraction studies on osteons, concluded that the fibre orientation was very much as they had previously described but that the statistical orientation of the crystallites was much more random than theorientationofthecollagenfibresasseen bySEM.Asthey were of the opinion that the mineral crystallites were intimately linked with the collagen fibrils, they suggested this anomaly might be explained by a degree of misorientation of the fibrils with respect to the fibres which they constitute. Barbos et a1.14 took the investigation a step further by examining the distribution of dark, intermediate and light osteons in sections of human femur. They found that lamellae with longitudinally oriented collagen fibres were found mainly in the lateral and anterior aspects of the bone. Conversely, lamellae whose collagen fibre bundles ran transversely were found mainly in the medial and posterial aspects of the bone. This distribution of lamellae according to theirfibre orientation revealed a degree of correlation with the distribution of stresses within the bone. In a recent paper15 we have shown that it is possibleto demonstrate the in viva position of collagen fibres in bone using the technique of collagenase etching. In this present study the orientation of the collagen fibres in bovine secondary osteons is investigated using collagenase etching.
polished sections (Figure 1). The polishing process was carefully monitored using an optical microscope to ensure that at least one osteon in each block was sectioned in both planes. On successful completion of the polishing process the blocks were cut with a fine dental saw to produce specimens of triangular cross-sections. These were then boiled in distilled water for 30 min to denature the collagen, thus enhancing the action of the collagenase. Aliquots (5 ml) of collagenase solution each containing 2 100 units were made up in 0.05 M Tris buffer, pH 7.4 (Sigma Chemicals Ltd: Type 1A, product No. T4378). These solutions also contained 0.1 M calcium chloride, the Ca2+ ions being necessary for activation of the collagenase. The solutions containing the specimens were incubated at 37°C for 1 wk, each bone segment being bathed in 5 ml of enzyme solution which was agitated twice daily. After incubation the specimens were coated with gold and examined in a Jeol 35C SEM and in a Jeol 120C Temscan operating in the scanning mode, at accelerating voltages of 25 and 100 kV respectively.
RESULTS AND DISCUSSION The transverse section of a demineralized bovine secondary osteon is shown in Figure 2. In the low magnification micrograph (Figure 2a) the well documented lamellar structure of the osteon is clearly revealed. The higher magnification (Figure 26) shows more detail of this alternating ring structure. Since the mineral has been removed from these samples the changes observed on moving from one ring to the next must be in the collagen component. There is an impression of a change in the orientation of the collagen fibres between adjacent rings but from this evidence alone the precise orientation of these fibres is uncertain. This is particularly the case when we bear in mind that the removal of the mineral will inevitably result in mobility of the much less rigid organic component. Indeed there is some concern that we have introduced artifacts by removing the mineral and that this has produced a collapsed structure. To overcome these criticisms we have approached,the problem from an alternative standpoint. We would argue that by removing the collagen component one would not expect the rigid mineral to collapse or be affected in any Polished transverse section , _ Final specimen
MATERIALS
AND METHODS Cut to produce final specimen
Demineralization Transverse sections of cortical bone from mature bovine tibia1 diaphyses were polished using carbide papers and polishing alumina. These were then placed in 1 M l-ICI for 4 d to remove the mineral component, dehydrated in ethanol and subsequently dried in air. After coating with gold the samples were examined in a Jeol 35C SEM operating at an accelerating voltage of 25 kV.
Collagen removal Small blocks (approx. 5 mm3) of cortical bone were cut from mature bovine tibia1 diaphyses. These were polished on two of their faces using carbide papers and polishing alumina to produce specimens with adjacent transverse and longitudial
428
8iomaterial.s
1987, Vol8 November
1
E
Longitudinal section
Figure 1 Schematic representation of a bone block which is initially polished on both transverse and longitudinal faces. The final cut to produce the sample for examination in the electron microscope is at 45” to each of these surfaces. This sample has a triangular cross-section and is initially mounted so that the electron beam direction is approx. 45” to each of the surfaces to be examined. Subsequently either surface may be independently observed.
6one microstructure: M. Green et al.
Figure 2 (a) A transverse section through a polished and demineralized bovine tibia1 secondary osteon. The remaining component (collagen) is seen to produce a lamellar structura of alternating rings; (b) A higher magnification micrograph showing more detail of the ring structure of the collagen component.
surface revealing a transverse section to the left of the micrograph and a longitudinal section to the right. The lamellar structure is clearly observed in both sections and the osteon axis is just revealed at the bottom of the micrograph. By rotating this sample, the transverse section may be studied (figure 5b) and subsequently the longitudinal section may be monitored (Figure 5~). The fortuitously placed reference particle approximately one third of the way down the centre of Figure 5a can be seen at the top right of Figure 56 and top left of Figure 5c, so that the positions of dark and light lamellae in each section can be related. Also the relative widths of these lamellae allow a clear analysis to be made. Thus in the transverse section the wider lamellae are light and contain holes arising from removal of collagen fibres which had their long axes approximately parallel to the long axis of the osteon. The corresponding wider lamellae in the longitudinal section show few holes suggesting that in this region there are few circumferential fibres. On the other hand the narrow lamellae in transverse section show few holes whereas in longitudinal section holes are revealed in these lamellae. These observations are consistent with a model for the orientation of collagen fibres in bovine tibia1 secondary osteons in which light lamellae in transverse sections predominantly contain fibres which have their long axes parallel to, or almost parallel to, the long axis of the osteon, alternating with dark lamellae in which the fibre direction is predominantly circumferential, or almost circumferential, to the long axis of the osteon. This is in agreement with the findings of Smith4 and Cooper et a/.5 and similar to the model proposed for intermediate osteons by Ascenzi and Bonucci’. This is represented diagramatically in Figure 6.
significant way. Thus when a bovine secondary osteon is transversely sectioned, polished and etched with collagenase, the sametype of lamellar structure is revealed (Figure 3). The lamellae are seen to be alternately light and dark in appearance. The light lamellae contain many holes up to -200 nm in diameter, with irregular profiles which arise from the removal of collagen fibres which run parallel to, or almost parallel to, the long axis of the osteon. The dark lamellae exhibit few such holes. The longitudinally sectioned collagenase etched osteon of Figure 4 also exhibits alternate light and dark lamellae, the light lamellae containing holes
_rr._ ..
-__., .-. .._. __ . . ..
circumferential to the OS naturally to the question of whether light lamellae in transverse section correspond to the light or dark regions in longitudinal section. If light lamellae in transverse section are also light in longitudinal section, then clearly these would be lamellae with a high collagen content and containing numerous fibres running both longitudinally and circumferentially. Correspondingly lamellae which were dark in both sections would be rich in mineral with few collagen fibres passing through. Alternatively if light lamellae in transverse section become dark in longitudinal section and vice versa the model for the osteon would include lamellae which alternate in the predominance of longitudinal and circumferential fibres respectively. This question may be resolved by examining o&eons which have been sectioned in both the transverse and longitudinal planes. Figure 5 is the result of such an experiment. In Figure 5a the triangularly cut section (see Figure 7) is observed at approx. 45” to each
Figure 3 (a) A transverse section through a bovine tibia1 secondary osteon. The well documented lamellar structure is clearly seen. The light rings contain holes produced by the removal of collagen fibres which had been oriented approximately parallel to the long axis of the osteon; (6) A higher magnification micrograph showing the holes more clearly.
Biomaterials
1987, Vol8 November
429
Bone microstructure: M. Green et al.
Figure 4 [a/ A ~ong~udinal section through a polished and callagenase etched bovine tibiai secondary o&eon. Towards the bottom of the mjcroqraph the inner surface of thefiaversian canalis revealed andprogressing upwards alternate light and dark lamellae are evident. The light rings again contain holes (cf Figure 3) created by the removal of collagen fibres, which in this case wera circumferential to the long axis of the osteon; (&j A higher magnification micrograph of a longitudinal section in which the hole structure within the light lame&e is more clearly seen.
By tilting the specimen in the electron microscope until the electron beam appears to be travelling directly down the holes vacated by the collagen fibres, an estimate may be made of the angles of these holes to the specimen surface and hence the osteon axis. This was carried out on both transverse and longitudinal sections. Observations of the transverse sections revealed that longitudinally oriented fibres were tilted by approximately 20” from the osteon axis, the direction of tilt alternating between one light lamella to the next (Figure 7). Similarly, from longitudinal sections it is seen that the circumferential fibres were tilted by approx. 15” from the plane perpendicular to the osteon axis, the direction of tilt again alternating between one light lamella and the next. To allow for the possibility of the osteon not being sectioned transversely at exactly 90” to its long axis a light lamella was selected which could easily befollowed around the Haversian canal. The specimen was held perpendicular to the electron beam, and an indivjdual hole or small group of holes in the selected light lamella was chosen. The specimen was then tilted until the electron beam appeared to be passing directly into the hole or group of holes in question and the angle of tilt was noted. The specimen was again set to be perpendicular to the electron beam and the procedure was repeated on the same lamella at a similar position on the opposite side of the Haversian canal. In this manner each light lamella in an individual osteon was examined beginning with the innermost and working towards the periphery of the osteon. A similar procedure was applied to the longitudinal sections.
430
Biomaterials
1987, Vol8 November
Figure 5 (a] SEM of the triangularly cut section (Figure I/ observed at about 45” to each surface showing the transverse section (left) and the longitudinal section (right). Note the osteon axis at the bottom of the micrograph, the alternating wide and narrow bands on each surface, and also the fo~uitously placed reference particle on the cut edge about one third from the top in the centre of the micrograph; (bJA highermaqni~cation of the transverse surface approximately perpendicular to the electron beam. Note the reference particle (top right) and the wide rings containing numerous holes from which collagen fibres approximately perpendicular to this surface have been removed. The narrow rings contain no such holes: (c] A higher maqni~cation of the ~onqitud~nal section approxjmate~ perpendicular to the electron beam. The reference part&/e is now at the top left and the wide rings contain few holes whereas the narrow rings reveal holes from which circumferential fibres have been removed
The difficulty experienced in estimating the angle at which the holes meet the plane of the section is demonstrated by an examination of Figure 8. Figure 8a shows part of the light lamella in transverse section. The specimen surface is perpendicular to the electron beam. Figure 86 shows the same area having been tiltad through 16”: the small change in appearance is difficult to appreciate in these mi~rographs
Bone microstructure: M. Green et a/.
PositIon of
:
Canal
:
Figure 6 Schematic representation of the orientations of collagen fibres in an osteon. Starting with the central canal at the top of the diagram successive lamellae are imagined to be added on progressing downwards. Thus the innermost lamellae have slightly tilted circumferential fibres, and the next lamellae have slightly tilted longitudinal fibres, etc. Note that the angle of tilt from the ‘ideal’ alternates from one circumferential layer to the next (i.e. the first layer istilted slightly up towards the right whereas the third layer is tilted slightly down towards the right) and the same effect is seen for the longitudinal fibres.
Figure 8 (a) A micrograph of part of a light area of a transverse section through an osteon. The surface of the specimen is at 90” to the electron beam; (bJ The same area with the specimen tilted through an angle of 16”. At this angle of tilt the holes now appear to be parallel to the electron beam. However, despite the tilt of the specimen the two views of this area are very similar and changes which are evident by direct observation are difficult to discern in these micrographs.
but
_Exposed transverse
becomes
tilting
much
is observed
more
apparent
directly
scope and can be followed Within
experimental
lamellae long11udlnal
of osteons
in transverse
specimen micro-
error, which was estimated
to be
the angle of tilt was found to be
same for each light lamella examined
For the majority
the
electron
by the observer.
up to 3” for each observation, the
when
in the scanning
in any given osteon.
the angles
of the fibres
sections werefound
in light
to be approx. 20”
from the osteon axis. This figure did however
vary and was
found in one case to be as low as 8” and in another
to be as
high as 35”. All the osteons found
to have the
illustrated proposed,
in
Figure fibres
been suggested
Figure 7 Schematic representation of the collagen fibre orientation in an osteon showing a transverse section at the top of the diagram and a longitudinal section towards the bottom. Circles and arrows represent fibres approximately perpendicular and parallel to the surface respectively. The arrows do not represent any specific direction of the collagen molecules, but rather show the small angles that the collagen fibres make with the revealed surfaces. The heads of the arrows indicate that fibres are tilted slightly outwards from the respective surfaces.
subjected. from variety
Other
only,
models
to which
were
specific of fibre
proportions
but a more
areas
of the
orientations.
of longitudinal
structure
systematic bone
might
In addition
that the angle of tilt from these
important
factor to be considered. described
study well
been It has
results
from
has not been of osteons reveal
more
to the variation
and circumferential
also related to the environmental
have
fibres only or
parts of the bone are
speculate
and the technique
that
in osteons
various
were
orientations
not encountered.
This aspect of the osteonic here,
fibre
of longitudinal
that this variation
the in vivo stresses considered
7.
in this investigation
of collagen
such as the presence
circumferential
Lume'n of Haverslan Canal
examined mixture
fibres
in we
ideals is a further
It is likely that this angle is stress patterns
here is a powerful
in the bone tool for such
investigations.
Biomaterials
198 7, Vol8 November
431
Bone microstructure: M. Green et al.
REFERENCES Kolliker, A, A Manual of Human Microscopic Anatomy,(Amer. Edn), (translated by G. Busk and T. Huxley), Lippincott, Grambo & Co., Philadelphia, 1854, pp 266-344, (Cited by Ruth3) Von Ebner, Ueber den feineren Eau der Knochensubstanz, Sitzungsb. d. Kaisel. Akad. d. Wissensch. Wien 1876, 72, 49-l 38, (Cited by Ruth3) Ruth, E.B., Bone studies. I: Fibrillar structure of adult human bone, Amer. J. Anat. 1947, 80. 35-53 Smith, J.W., The arrangement of collagen fibres in human secondan/ osteons. J. Bone Jt Surg. 1960,428 (no. 3). 588-605 Cooper, RR., Milgram, J.W. and Robinson, R.A., Morphology of the osteon - an electron microscope study, .I. Bone Jt Sorg. 1966,48A (no. 7). 1239-I 271 Ascenzi, A. and Bonucci, E., The compressive properties of single osteons, Anat. Rec. 1968, 161, 377-391 Boyde, A. and Hobdell, M.H., Scanning electron microscopy of lamellar bone, Z i’ellforsch. Midrosk. Anat. 1969, 93, 213-231 Boyde, A., Scanning electron microscope studies of bone, The
432
Biomaterials
198 7, Vol8 November
9
10 11 12 13
14
15
Biochemistry and Physiology of bone, (2nd Edn), (Ed. G.H. Eourne). Vol. 1, Academic Press, 1972, pp 259-310 Black, J., Mattson, R. and Korostoff, E.. Haversian osteons-size, distribution, internal structure and orientation,.I. Biomed. Mater. Res. 1974,6, 299-319 Ascenzi, A. and Bonucci, E., Relationship between ultrastructure and pin test in osteons, C/in. &hop. Rel. Res. 1976, 121, 275-294 Frasca, P., Harper, R.A. and Katz, J.L., Isolation of single osteons and osteon lamellae, Acta Anat. 1976, 95. 122-l 29 Frasca, P., Harper, R.A. and Katz. J.L., Collagen fibre orientation in human secondary osteons, Acta Anat. 1977, 98, l-l 3 Frasca, P.. Harper, R.A. and Katz, J.L., Mineral and collagen fibre orientation in human secondary osteons, .I. Dent. Res. 1978,67 (no. 3). 526-533 Barbos, M.P., Bianco, P. and Ascenzi, A., Distribution of osteonic and interstitial components in the human fermoral shaft with reference to structure, calcification and mechanical properties, Acta Anat. 1983, 115, 178-l 86 Green, M., Isaac, D.H. and Jenkins, G.M., Bone microstructure by collagenase etching, Biomateria/s 1985, 6, 150-l 52
The orientation of collagen fibres in bovine secondary osteons has been investigated in the scanning electron microscope (SEM) by removal independently of firstly the mineral component and secondly the collagen fibres. Demineralization of polished transverse sections reveals a lamellar structure for the collagen component but the precise orientation of the collagen in each ring is not unequivocably determined. However, by using a collagenase solution to etch away the collagen component of a polished surface, holes are produced in the mineral revealing the former position of the fibres. The greater rigidity of the mineral component ensures that the structure does not collapse and produce artifacts. A specimen cut so that transverse and longitudinal sections are simultaneously observed allows the relationship between the structural features on each surface to be revealed. Analysis of such micrographs indicates a model for the collagen component of osteons in which the lamellar structure contains fibres with orientations alternately parallel to and circumferential to the long axis of the osteon. Tilting the samples to look directly down the holes shows that the fibres are not precisely longitudinal and circumferential but are tilted from these ideals by a variable angle (typically 20”) the precise angle probably being an important factor related to the in vivu mechanical property requirements. Keywords: Bone, collagen, bone microstructure, SEM. collagenase etching
The spatial arrangement of the collagen fibres in bone has long been a topic of investigation. Kolliker’ expressed the belief that the organic matrix of adult bone was the embryonic skeleton in which calcium salts had been deposited. He further stated that this organic interstitial substance was granular in nature rather than fibrillar. Von Ebnerz, held the view that the spaces between the fibrillar bundles were filled with a substance which contained the mineral salts of the bone, while the fibrils themselves remained noncalcified. Ruth3 examined thin sections of human femoral bone which had been decalcified and treated with various histological stains. He investigated the lamellar structure of osteons and designated the two types of lamellae observed to be compact or diffuse, in accordance with their structural appearance. The compact lamellae were interpreted as bands of circumferentially oriented fibrillae, while the diffuse lamellae were said to be bands of radially oriented fibrillae which were loosely disposed and separated from each other by relatively wide inter-fibrillar spaces filled with a granular substance, namely the mineral. Smith4 also examined sections of the human femur and suggested that the dark lamellae in transverse section contained predominantly circumferential fibres, and corresponded to the light lamellae in the radial sections in which these fibres were cut at right angles. Similarly, he considered ‘Present address: Department of Mechanical and Manufacturing Systems Engineering, UWIST, Cardiff, UK. 0 1987
the dark lamellae in radial sections contained predominantly longitudinal fibres and corresponded to the light lamellae in the transverse sections in which these fibres were again cut at right angles to their long axes. Cooper et a/.5 noted similar features in thin sections of secondary osteons from mongrel dogs. Ascenzi and Bonucc? proposed that osteons could be divided into three categories depending on their appearance under crossed polarized light. The categories were: (1) dark, in which the collagen fibres were oriented parallel to the long axis of the osteon; (2) light, in which the collagen fibres were circumferential in relation to the long axis of the osteon; and (3) intermediate, a mixture of the two. Boyde and Hobdel17 using a scanning electron microscope (SEM) suggested that lamellae were fibrous plates about 3 pm thick separated by inter-lamellar cement bands about 0.1 pm thick, but they found that the fibre direction in a given lamella was very variable. Boyde’, in an extensive review article, introduced the concept of domains for the collagen fibres, a domain being a volume over which the collagen fibre bundle orientation is essentially parallel. His work on human adult femoral bone showed that domains are usually longer along the length of the fibre bundles (typically -8Opm) and narrowerwhen measured across the fibre axis (typically 2030 pm). Black et a/.’ suggested that only the dark osteons proposed by Ascenzi and Bonucci’ were in fact present in bone, the other types being artifacts resulting from the fact that it was almost impossible to be sure that any osteon
Butterworth 8 Co (Publishers) Ltd. 0142-9612/87/060427-06$03.00 Biomaterials
198 7. Vol8 November
427
Bone microstructure: M. Green et al.
under examination had been sectioned at 90” to its long axis. They argued that if the section was not cut at exactly 90” to the long axis of the osteon then fibres parallel to the long axis would not be perpendicular to the plane of the section and would therefore be considered to be part of a spiral around the central canal. Ascenzi and Bonucci” used a series of mechanical tests to support their opinion that there were three types of osteon. Frasca et al.” examined single isolated human osteons which had been mechanically manipulated to expose their inter-lamellar interfaces. They concluded that only the dark osteons proposed by Ascenzi and Bonucci’,” actually conformed to the initial model, the light and intermediate osteons being more complex than first thought. They compared the results of the examination of osteonic specimens undercrossed polarized light following mechanical manipulation” , and concluded that osteons which appeared dark under crossed polars contained fibres with little or no transverse component, whereas light osteons possessed fibre orientations with transverse and longitudinal components. Intermediate osteons were found to contain fibre orientations richer in longitudinal component than in transverse. Frasca et a1.‘3. from X-ray diffraction studies on osteons, concluded that the fibre orientation was very much as they had previously described but that the statistical orientation of the crystallites was much more random than theorientationofthecollagenfibresasseen bySEM.Asthey were of the opinion that the mineral crystallites were intimately linked with the collagen fibrils, they suggested this anomaly might be explained by a degree of misorientation of the fibrils with respect to the fibres which they constitute. Barbos et a1.14 took the investigation a step further by examining the distribution of dark, intermediate and light osteons in sections of human femur. They found that lamellae with longitudinally oriented collagen fibres were found mainly in the lateral and anterior aspects of the bone. Conversely, lamellae whose collagen fibre bundles ran transversely were found mainly in the medial and posterial aspects of the bone. This distribution of lamellae according to theirfibre orientation revealed a degree of correlation with the distribution of stresses within the bone. In a recent paper15 we have shown that it is possibleto demonstrate the in viva position of collagen fibres in bone using the technique of collagenase etching. In this present study the orientation of the collagen fibres in bovine secondary osteons is investigated using collagenase etching.
polished sections (Figure 1). The polishing process was carefully monitored using an optical microscope to ensure that at least one osteon in each block was sectioned in both planes. On successful completion of the polishing process the blocks were cut with a fine dental saw to produce specimens of triangular cross-sections. These were then boiled in distilled water for 30 min to denature the collagen, thus enhancing the action of the collagenase. Aliquots (5 ml) of collagenase solution each containing 2 100 units were made up in 0.05 M Tris buffer, pH 7.4 (Sigma Chemicals Ltd: Type 1A, product No. T4378). These solutions also contained 0.1 M calcium chloride, the Ca2+ ions being necessary for activation of the collagenase. The solutions containing the specimens were incubated at 37°C for 1 wk, each bone segment being bathed in 5 ml of enzyme solution which was agitated twice daily. After incubation the specimens were coated with gold and examined in a Jeol 35C SEM and in a Jeol 120C Temscan operating in the scanning mode, at accelerating voltages of 25 and 100 kV respectively.
RESULTS AND DISCUSSION The transverse section of a demineralized bovine secondary osteon is shown in Figure 2. In the low magnification micrograph (Figure 2a) the well documented lamellar structure of the osteon is clearly revealed. The higher magnification (Figure 26) shows more detail of this alternating ring structure. Since the mineral has been removed from these samples the changes observed on moving from one ring to the next must be in the collagen component. There is an impression of a change in the orientation of the collagen fibres between adjacent rings but from this evidence alone the precise orientation of these fibres is uncertain. This is particularly the case when we bear in mind that the removal of the mineral will inevitably result in mobility of the much less rigid organic component. Indeed there is some concern that we have introduced artifacts by removing the mineral and that this has produced a collapsed structure. To overcome these criticisms we have approached,the problem from an alternative standpoint. We would argue that by removing the collagen component one would not expect the rigid mineral to collapse or be affected in any Polished transverse section , _ Final specimen
MATERIALS
AND METHODS Cut to produce final specimen
Demineralization Transverse sections of cortical bone from mature bovine tibia1 diaphyses were polished using carbide papers and polishing alumina. These were then placed in 1 M l-ICI for 4 d to remove the mineral component, dehydrated in ethanol and subsequently dried in air. After coating with gold the samples were examined in a Jeol 35C SEM operating at an accelerating voltage of 25 kV.
Collagen removal Small blocks (approx. 5 mm3) of cortical bone were cut from mature bovine tibia1 diaphyses. These were polished on two of their faces using carbide papers and polishing alumina to produce specimens with adjacent transverse and longitudial
428
8iomaterial.s
1987, Vol8 November
1
E
Longitudinal section
Figure 1 Schematic representation of a bone block which is initially polished on both transverse and longitudinal faces. The final cut to produce the sample for examination in the electron microscope is at 45” to each of these surfaces. This sample has a triangular cross-section and is initially mounted so that the electron beam direction is approx. 45” to each of the surfaces to be examined. Subsequently either surface may be independently observed.
6one microstructure: M. Green et al.
Figure 2 (a) A transverse section through a polished and demineralized bovine tibia1 secondary osteon. The remaining component (collagen) is seen to produce a lamellar structura of alternating rings; (b) A higher magnification micrograph showing more detail of the ring structure of the collagen component.
surface revealing a transverse section to the left of the micrograph and a longitudinal section to the right. The lamellar structure is clearly observed in both sections and the osteon axis is just revealed at the bottom of the micrograph. By rotating this sample, the transverse section may be studied (figure 5b) and subsequently the longitudinal section may be monitored (Figure 5~). The fortuitously placed reference particle approximately one third of the way down the centre of Figure 5a can be seen at the top right of Figure 56 and top left of Figure 5c, so that the positions of dark and light lamellae in each section can be related. Also the relative widths of these lamellae allow a clear analysis to be made. Thus in the transverse section the wider lamellae are light and contain holes arising from removal of collagen fibres which had their long axes approximately parallel to the long axis of the osteon. The corresponding wider lamellae in the longitudinal section show few holes suggesting that in this region there are few circumferential fibres. On the other hand the narrow lamellae in transverse section show few holes whereas in longitudinal section holes are revealed in these lamellae. These observations are consistent with a model for the orientation of collagen fibres in bovine tibia1 secondary osteons in which light lamellae in transverse sections predominantly contain fibres which have their long axes parallel to, or almost parallel to, the long axis of the osteon, alternating with dark lamellae in which the fibre direction is predominantly circumferential, or almost circumferential, to the long axis of the osteon. This is in agreement with the findings of Smith4 and Cooper et a/.5 and similar to the model proposed for intermediate osteons by Ascenzi and Bonucci’. This is represented diagramatically in Figure 6.
significant way. Thus when a bovine secondary osteon is transversely sectioned, polished and etched with collagenase, the sametype of lamellar structure is revealed (Figure 3). The lamellae are seen to be alternately light and dark in appearance. The light lamellae contain many holes up to -200 nm in diameter, with irregular profiles which arise from the removal of collagen fibres which run parallel to, or almost parallel to, the long axis of the osteon. The dark lamellae exhibit few such holes. The longitudinally sectioned collagenase etched osteon of Figure 4 also exhibits alternate light and dark lamellae, the light lamellae containing holes
_rr._ ..
-__., .-. .._. __ . . ..
circumferential to the OS naturally to the question of whether light lamellae in transverse section correspond to the light or dark regions in longitudinal section. If light lamellae in transverse section are also light in longitudinal section, then clearly these would be lamellae with a high collagen content and containing numerous fibres running both longitudinally and circumferentially. Correspondingly lamellae which were dark in both sections would be rich in mineral with few collagen fibres passing through. Alternatively if light lamellae in transverse section become dark in longitudinal section and vice versa the model for the osteon would include lamellae which alternate in the predominance of longitudinal and circumferential fibres respectively. This question may be resolved by examining o&eons which have been sectioned in both the transverse and longitudinal planes. Figure 5 is the result of such an experiment. In Figure 5a the triangularly cut section (see Figure 7) is observed at approx. 45” to each
Figure 3 (a) A transverse section through a bovine tibia1 secondary osteon. The well documented lamellar structure is clearly seen. The light rings contain holes produced by the removal of collagen fibres which had been oriented approximately parallel to the long axis of the osteon; (6) A higher magnification micrograph showing the holes more clearly.
Biomaterials
1987, Vol8 November
429
Bone microstructure: M. Green et al.
Figure 4 [a/ A ~ong~udinal section through a polished and callagenase etched bovine tibiai secondary o&eon. Towards the bottom of the mjcroqraph the inner surface of thefiaversian canalis revealed andprogressing upwards alternate light and dark lamellae are evident. The light rings again contain holes (cf Figure 3) created by the removal of collagen fibres, which in this case wera circumferential to the long axis of the osteon; (&j A higher magnification micrograph of a longitudinal section in which the hole structure within the light lame&e is more clearly seen.
By tilting the specimen in the electron microscope until the electron beam appears to be travelling directly down the holes vacated by the collagen fibres, an estimate may be made of the angles of these holes to the specimen surface and hence the osteon axis. This was carried out on both transverse and longitudinal sections. Observations of the transverse sections revealed that longitudinally oriented fibres were tilted by approximately 20” from the osteon axis, the direction of tilt alternating between one light lamella to the next (Figure 7). Similarly, from longitudinal sections it is seen that the circumferential fibres were tilted by approx. 15” from the plane perpendicular to the osteon axis, the direction of tilt again alternating between one light lamella and the next. To allow for the possibility of the osteon not being sectioned transversely at exactly 90” to its long axis a light lamella was selected which could easily befollowed around the Haversian canal. The specimen was held perpendicular to the electron beam, and an indivjdual hole or small group of holes in the selected light lamella was chosen. The specimen was then tilted until the electron beam appeared to be passing directly into the hole or group of holes in question and the angle of tilt was noted. The specimen was again set to be perpendicular to the electron beam and the procedure was repeated on the same lamella at a similar position on the opposite side of the Haversian canal. In this manner each light lamella in an individual osteon was examined beginning with the innermost and working towards the periphery of the osteon. A similar procedure was applied to the longitudinal sections.
430
Biomaterials
1987, Vol8 November
Figure 5 (a] SEM of the triangularly cut section (Figure I/ observed at about 45” to each surface showing the transverse section (left) and the longitudinal section (right). Note the osteon axis at the bottom of the micrograph, the alternating wide and narrow bands on each surface, and also the fo~uitously placed reference particle on the cut edge about one third from the top in the centre of the micrograph; (bJA highermaqni~cation of the transverse surface approximately perpendicular to the electron beam. Note the reference particle (top right) and the wide rings containing numerous holes from which collagen fibres approximately perpendicular to this surface have been removed. The narrow rings contain no such holes: (c] A higher maqni~cation of the ~onqitud~nal section approxjmate~ perpendicular to the electron beam. The reference part&/e is now at the top left and the wide rings contain few holes whereas the narrow rings reveal holes from which circumferential fibres have been removed
The difficulty experienced in estimating the angle at which the holes meet the plane of the section is demonstrated by an examination of Figure 8. Figure 8a shows part of the light lamella in transverse section. The specimen surface is perpendicular to the electron beam. Figure 86 shows the same area having been tiltad through 16”: the small change in appearance is difficult to appreciate in these mi~rographs
Bone microstructure: M. Green et a/.
PositIon of
:
Canal
:
Figure 6 Schematic representation of the orientations of collagen fibres in an osteon. Starting with the central canal at the top of the diagram successive lamellae are imagined to be added on progressing downwards. Thus the innermost lamellae have slightly tilted circumferential fibres, and the next lamellae have slightly tilted longitudinal fibres, etc. Note that the angle of tilt from the ‘ideal’ alternates from one circumferential layer to the next (i.e. the first layer istilted slightly up towards the right whereas the third layer is tilted slightly down towards the right) and the same effect is seen for the longitudinal fibres.
Figure 8 (a) A micrograph of part of a light area of a transverse section through an osteon. The surface of the specimen is at 90” to the electron beam; (bJ The same area with the specimen tilted through an angle of 16”. At this angle of tilt the holes now appear to be parallel to the electron beam. However, despite the tilt of the specimen the two views of this area are very similar and changes which are evident by direct observation are difficult to discern in these micrographs.
but
_Exposed transverse
becomes
tilting
much
is observed
more
apparent
directly
scope and can be followed Within
experimental
lamellae long11udlnal
of osteons
in transverse
specimen micro-
error, which was estimated
to be
the angle of tilt was found to be
same for each light lamella examined
For the majority
the
electron
by the observer.
up to 3” for each observation, the
when
in the scanning
in any given osteon.
the angles
of the fibres
sections werefound
in light
to be approx. 20”
from the osteon axis. This figure did however
vary and was
found in one case to be as low as 8” and in another
to be as
high as 35”. All the osteons found
to have the
illustrated proposed,
in
Figure fibres
been suggested
Figure 7 Schematic representation of the collagen fibre orientation in an osteon showing a transverse section at the top of the diagram and a longitudinal section towards the bottom. Circles and arrows represent fibres approximately perpendicular and parallel to the surface respectively. The arrows do not represent any specific direction of the collagen molecules, but rather show the small angles that the collagen fibres make with the revealed surfaces. The heads of the arrows indicate that fibres are tilted slightly outwards from the respective surfaces.
subjected. from variety
Other
only,
models
to which
were
specific of fibre
proportions
but a more
areas
of the
orientations.
of longitudinal
structure
systematic bone
might
In addition
that the angle of tilt from these
important
factor to be considered. described
study well
been It has
results
from
has not been of osteons reveal
more
to the variation
and circumferential
also related to the environmental
have
fibres only or
parts of the bone are
speculate
and the technique
that
in osteons
various
were
orientations
not encountered.
This aspect of the osteonic here,
fibre
of longitudinal
that this variation
the in vivo stresses considered
7.
in this investigation
of collagen
such as the presence
circumferential
Lume'n of Haverslan Canal
examined mixture
fibres
in we
ideals is a further
It is likely that this angle is stress patterns
here is a powerful
in the bone tool for such
investigations.
Biomaterials
198 7, Vol8 November
431
Bone microstructure: M. Green et al.
REFERENCES Kolliker, A, A Manual of Human Microscopic Anatomy,(Amer. Edn), (translated by G. Busk and T. Huxley), Lippincott, Grambo & Co., Philadelphia, 1854, pp 266-344, (Cited by Ruth3) Von Ebner, Ueber den feineren Eau der Knochensubstanz, Sitzungsb. d. Kaisel. Akad. d. Wissensch. Wien 1876, 72, 49-l 38, (Cited by Ruth3) Ruth, E.B., Bone studies. I: Fibrillar structure of adult human bone, Amer. J. Anat. 1947, 80. 35-53 Smith, J.W., The arrangement of collagen fibres in human secondan/ osteons. J. Bone Jt Surg. 1960,428 (no. 3). 588-605 Cooper, RR., Milgram, J.W. and Robinson, R.A., Morphology of the osteon - an electron microscope study, .I. Bone Jt Sorg. 1966,48A (no. 7). 1239-I 271 Ascenzi, A. and Bonucci, E., The compressive properties of single osteons, Anat. Rec. 1968, 161, 377-391 Boyde, A. and Hobdell, M.H., Scanning electron microscopy of lamellar bone, Z i’ellforsch. Midrosk. Anat. 1969, 93, 213-231 Boyde, A., Scanning electron microscope studies of bone, The
432
Biomaterials
198 7, Vol8 November
9
10 11 12 13
14
15
Biochemistry and Physiology of bone, (2nd Edn), (Ed. G.H. Eourne). Vol. 1, Academic Press, 1972, pp 259-310 Black, J., Mattson, R. and Korostoff, E.. Haversian osteons-size, distribution, internal structure and orientation,.I. Biomed. Mater. Res. 1974,6, 299-319 Ascenzi, A. and Bonucci, E., Relationship between ultrastructure and pin test in osteons, C/in. &hop. Rel. Res. 1976, 121, 275-294 Frasca, P., Harper, R.A. and Katz, J.L., Isolation of single osteons and osteon lamellae, Acta Anat. 1976, 95. 122-l 29 Frasca, P., Harper, R.A. and Katz. J.L., Collagen fibre orientation in human secondary osteons, Acta Anat. 1977, 98, l-l 3 Frasca, P.. Harper, R.A. and Katz, J.L., Mineral and collagen fibre orientation in human secondary osteons, .I. Dent. Res. 1978,67 (no. 3). 526-533 Barbos, M.P., Bianco, P. and Ascenzi, A., Distribution of osteonic and interstitial components in the human fermoral shaft with reference to structure, calcification and mechanical properties, Acta Anat. 1983, 115, 178-l 86 Green, M., Isaac, D.H. and Jenkins, G.M., Bone microstructure by collagenase etching, Biomateria/s 1985, 6, 150-l 52