- Front Endocrinol (Lausanne). 2012 Aug 9;3:98. doi: 10.3389/fendo.2012.00098. eCollection 2012. Periosteal Sharpey's fibers: a novel bone matrix regulatory system?
Sharpey's
"perforating" fibers (SF) are well known skeletally in tooth anchorage.
Elsewhere they provide anchorage for the periosteum and are less well
documented.
Immunohistochemistry has transformed their potential significance by identifying their collagen type III (CIII) content and enabling their mapping in domains as permeating arrays of fibers (5-25 μ thick), protected from osteoclastic resorption by their poor mineralization.
As periosteal extensions they are crucial in early skeletal development and central to intramembranous bone healing, providing unique microanatomical avenues for musculoskeletal exchange, their composition (e.g., collagen type VI, elastin, tenascin) combined with a multiaxial pattern of insertion suggesting a role more complex than attachment alone would justify.
A proportion permeate the cortex to the endosteum (and beyond), fusing into a CIII-rich osteoid layer (<2 μ thick) encompassing all resting surfaces, and with which they apparently integrate into a
PERIOSTEAL-SHARPEY FIBER-ENDOSTEUM (PSE) structural continuum.
This intraosseous system behaves in favor of bone loss or gain depending upon extraneous stimuli (i.e., like Frost's hypothetical "mechanostat"). Thus, the birefringent fibers are sensitive to humoral factors (e.g., estrogen causes retraction, rat femur model), physical activity (e.g., running causes expansion, rat model), aging (e.g., causes fragmentation, pig mandible model), and pathology (e.g., atrophied in osteoporosis, hypertrophied in osteoarthritis, human proximal femur), and with encroaching mineral particles hardening the usually soft parts. In this way the unobtrusive periosteal SF network may regulate bone status, perhaps even contributing to predictable "hotspots" of trabecular disconnection, particularly at sites of tension prone to fatigue, and with the network deteriorating significantly before bone matrix loss.
Immunohistochemistry has transformed their potential significance by identifying their collagen type III (CIII) content and enabling their mapping in domains as permeating arrays of fibers (5-25 μ thick), protected from osteoclastic resorption by their poor mineralization.
As periosteal extensions they are crucial in early skeletal development and central to intramembranous bone healing, providing unique microanatomical avenues for musculoskeletal exchange, their composition (e.g., collagen type VI, elastin, tenascin) combined with a multiaxial pattern of insertion suggesting a role more complex than attachment alone would justify.
A proportion permeate the cortex to the endosteum (and beyond), fusing into a CIII-rich osteoid layer (<2 μ thick) encompassing all resting surfaces, and with which they apparently integrate into a
PERIOSTEAL-SHARPEY FIBER-ENDOSTEUM (PSE) structural continuum.
This intraosseous system behaves in favor of bone loss or gain depending upon extraneous stimuli (i.e., like Frost's hypothetical "mechanostat"). Thus, the birefringent fibers are sensitive to humoral factors (e.g., estrogen causes retraction, rat femur model), physical activity (e.g., running causes expansion, rat model), aging (e.g., causes fragmentation, pig mandible model), and pathology (e.g., atrophied in osteoporosis, hypertrophied in osteoarthritis, human proximal femur), and with encroaching mineral particles hardening the usually soft parts. In this way the unobtrusive periosteal SF network may regulate bone status, perhaps even contributing to predictable "hotspots" of trabecular disconnection, particularly at sites of tension prone to fatigue, and with the network deteriorating significantly before bone matrix loss.
KEYWORDS:
collagen type III; collagen type VI; elastin; endosteal membrane; matrix biochemical domains; skeletal aging; tenascincollagen type VI, elastin, tenascin) combined with a multiaxial pattern of insertion suggesting a role more complex than attachment alone would justify. A proportion permeate the cortex to the endosteum (and beyond), fusing into a CIII-rich osteoid layer (<2 μ thick) encompassing all resting surfaces, and with which they apparently integrate into a PERIOSTEAL-SHARPEY FIBER-ENDOSTEUM (PSE) structural continuum. This intraosseous system behaves in favor of bone loss or gain depending upon extraneous stimuli (i.e., like Frost’s hypothetical “mechanostat”). Thus, the birefringent fibers are sensitive to humoral factors (e.g., estrogen causes retraction, rat femur model), physical activity (e.g., running causes expansion, rat model), aging (e.g., causes fragmentation, pig mandible model), and pathology (e.g., atrophied in osteoporosis, hypertrophied in osteoarthritis, human proximal femur), and with encroaching mineral particles hardening the usually soft parts. In this way the unobtrusive periosteal SF network may regulate bone status, perhaps even contributing to predictable “hotspots” of trabecular disconnection, particularly at sites of tension prone to fatigue, and with the network deteriorating significantly before bone matrix loss.
Keywords: collagen type III, collagen type VI, tenascin, elastin, matrix biochemical domains, skeletal aging, endosteal membrane
PERIOSTEUM AND SHARPEY’S FIBERS
PERIOSTEUM
This
strong, encapsulating skeletal membrane containing osteoprogenitor
cells consists of an outer fibrillar layer and an inner cellular layer
that is usually poorly defined unless actively engaged in osteoid
apposition. Despite its relatively low visual impact it defines vital
developmental boundaries. Extending from it are the Sharpey’s fibers
that ensure adhesion to the outer cortex and to tendons and ligaments,
themselves perceived as modified periosteum (Hurle et al., 1989).
While the unstressed periosteum seems biochemically quiescent, short
bursts of loading stimulates the rapid induction of enzyme activity
within discrete periosteal and bone matrix domains (Skerry et al., 1989) apparently by the mediation of signals to selected regions.
SHARPEY’S FIBERS
These delicate optical features (Figures Figures11 and and22)
described as “perforating fibers” by William Sharpey, cross matrix
lamellae and are particularly abundant in the alveolar socket of the
teeth (Sharpey et al., 1867). Also reporting them at this time was H. Muller (Quain’s Elements of Anatomy, 1867) who recognized the elastic nature of the fibers and a tendency to “escape calcification.” Later Weidenreich (1923),
citing Koelliker (1886), confirmed their poorly mineralized status, and
although they were apparently short and superficial he was of the
opinion that they influenced not only the external anatomy but also the
internal bone structure. From another quarter were reports by Tomes (1876) and Black (1887) that embedded Sharpey’s fibers constituted the cemento-alveolar fibers of the periodontal ligament, and in due course Cohn (1972) mapped their passage through the cementum and on across the entire thickness of the alveolar wall (Quigley, 1970). Other related reports followed, such as that by Jones and Boyde (1974)
outlining further their presence in the cranial sutures and muscle
attachments as well as in tooth sockets. However, the subsequent
literature focused almost exclusively on Sharpey’s fibers functioning as
the periodontal ligament and how this special dental structure altered
with age both organically and inorganically, weakening its tooth-holding
capacity. The detrimental changes observed included fibrosis, increased
cellularity, and progressive calcification (Sloan et al., 1993).
Photomicrograph
of a typical representative array of three periosteal Sharpey’s fibers
(black arrows), each about 15 μm thick, and extending from the
periosteum (P), through the bone (B) toward the endosteum outside which
is the marrow ...
Photomicrograph
of two typical representative collagen type III-rich Sharpey’s fibers,
about 10 μm thick, and fluorescing positive within the negative
calcified bone matrix. Human proximal femur. FITC-immunostain for CIII, UV epifluorescence ...
At
the present time, sufficient evidence is now accumulating to suggest
that the relative neglect of those abundant Sharpey’s fibers located
away from the dentition may be unjustified. In redressing the balance in
favor of their structural significance elsewhere in the skeleton, and
complementing the classification of Johnson (1987), Al-Qtaitat (2004), 2007 identified two types of Sharpey’s fibers (see also Al-Qtaitat et al., 2010),
one coarse (8–25 μm thick) and the other fine (<8 μm thick). Their
entry angle into the subperiosteal bone was multiaxial. It included the
almost horizontal (i.e., tangential) fibers especially common with age
and often found among inserting muscle fascicles, functionally
propagating biomechanical exchange across the periosteum. It also
included the perpendicular (i.e., vertical) fibers, frequently crossing
the cortex to the cancellous region and generally of the coarse type in
bundles <40 μm thick, functionally adding complexity to the
muscle-to-bone interface that may influence bone atrophy, augmentation,
and remodeling. In addition were the oblique fibers, these being the
most numerous and predominant in the young skeleton, functionally
mediating exchange between the periosteum and outer cortex and providing
soft tissue anchorage. While some of these insertions apparently ended
abruptly (like rows of short, regular parallel stitches), it was the
proportion that traversed to the medulla, some becoming intertrabecular,
others with dispersed intra-osseous fan-like termini that were of
special interest. Added to this was their unusual profile in transverse
section, which was not the simple circle expected but showed sharply
defined surface indentations and configurations ranging from a
horseshoe-shape to a “hollow” core (Aaron and Skerry, 1994).
Further
examination using an established histochemistry test for elastin
(Verhoeff’s stain) supported the observation of Muller above that
(unlike collagen type I, CI) they have elastic properties that can
absorb strain. Moreover, the elastin staining was not uniform but
suggested the discrete contours of a spiral encircling some of the
individual coarse fibers (Aaron and Skerry, 1994).
The mechanical properties of elastin are unique. Unlike non-extensible
collagen it can be stretched, recoils, branches, and imparts
flexibility. However, it has been rarely documented in bone (Johnson and Low, 1981; Keene et al., 1991),
except, that is, at sites of tendon and ligament insertion, and its
presence will alter the biophysical properties of the Sharpey’s fibers.
IMMUNOHISTOCHEMISTRY OF SHARPEY’S FIBERS
It
required a technological advance to demonstrate the otherwise hidden
scale of Sharpey fiber permeation and to establish their biochemical
composition more extensively (see Aaron and Shore, 2004b
for technical details). Polarized light showed a highly birefringent
nature consistent with collagen, but little else could be deduced by
simple staining (Smith, 1960), with for example picro-sirius red stain, or by the Goldner tetrachrome method (Aaron and Shore, 2004b). The prospect was transformed by the introduction of heavy duty cryomicrotomy (see for example, Aaron and Carter, 1987; Carter et al., 1989),
combined with the increasing availability of a widening range of
specific fluorescent antibodies. Prior to this, the organic matrix
biochemistry was based on tissue homogenates and extracts. The new
method enabled a structural face to be applied. This identified
previously unsuspected matrix sub-divisions, showing a mosaic of
biochemically distinct domains, defined by boundaries and with
differential aptitudes for signal trafficking through, for example,
endochondrally derived versus intramembranous regions. Perhaps foremost
among these potentially transducing macomolecules is collagen type III
(CIII). It is this together with amounts of collagen type VI (CVI),
tenasin, fibronection, and elastin, that are now known to characterize
the Sharpey’s fibers, meaning that structurally they are considerably
more complex than was previously supposed, and especially complex for
structures fulfilling the relatively uncomplicated function of anchorage
traditionally assigned to them.
An advance
in their histochemical differentiation had pre-empted their
immunohistochemistry with descriptions of certain “argyrophilic” matrix
fibers (Nowack et al., 1976; Carter et al., 1991
for references), which in retrospect were found to be coincident in
distribution with CIII immunostaining. As with the more prominent CI, so
CIII is also found in all interstitial connective tissues but in
contrast there was little evidence for its occurrence in bone, with the
exception of the earliest mesenchymal condensations (Pratt, 1959; Page et al., 1986), and its well documented appearance in alveolar bone (Becker et al., 1986). The application of CIII immunostaining (Figure Figure22) has now transformed this state of affairs (Wang et al., 1980), leaving in no doubt its discrete structural affinity for the birefringent Sharpey’s fibers.
Collagen type III
It
is recognized that different collagens, e.g., CI and CIII can be
present in the same fibril to modulate its physical properties. Like CI,
the structure of CIII consists of long (300 nm) uninterrupted triple
helices, chemically distinguished from CI by an increased level of 4-Hyp
and the occurrence of cysteine, facilitating disulphide bond formation.
In contrast to the high tensile strength of CI fibers, those of CIII
are thinner and less orderly (Kielty et al., 1993)
and they are prevalent in tissues with clear elastic properties,
including skin, aorta, lung, and gut. As well as being argyrophilic,
above, these fibers were known histologically as reticulin fibers, and
were especially associated with epithelial basement membrane stability,
where their contribution to organ containment cannot be overestimated.
It was reported by Bailey et al. (1993)
that in normal human bone CIII content averaged 4–5%, with 3% in
osteoporotic bone. Similarly in culture conditions osteoblast-like cells
have been said to secrete about 6% CIII (Aufmokolk et al., 1985; see Luther, 1998
for references). The collagens CI, II, and III are all translated from
mRNA coding for pre-proα chains with similar, but not identical, N- and
C-terminal extensions. The partnership of CIII with CVI is reported to
provide exceptional stability (e.g., Hulmes, 1992; Sherwin et al., 1999)
and this combination in Sharpey’s fibers must have fundamental
implications for their persistence in a tissue with the versatility and
turnover of bone.
Collagen type VI
A structural association between CIII and CVI in bone was reported by Becker et al. (1986), and CVI was said to be reduced in osteoporosis (OP; Bailey et al., 1993)
although the implications were not clear. CVI is microfibrillary,
composed of a short triple helical axis and globular termini, creating
its typical dumbell shape. It has many adhesive RGD sequences and like
CIII has stabilizing disulfide bonds (Hulmes, 1992). It has been suggested that the removal of CVI is a factor that may permit remodeling (Sloan et al., 1993).
From the above evidence it is clear that Sharpey’s fibers are uniquely placed and have the morphological complexity (Figure Figure33)
to mediate musculoskeletal cohesion and exchange. They are the only
continuous anatomical structure to (i) integrate directly with the
muscles, ligaments, and tendons, (ii) traverse the periosteum from which
they arise, and (iii) permeate the extracellular matrix multiaxially
and to varying degrees. Insight into their most basic structural
modulation may be found in tooth movement where orthodontic forces
strengthen the CIII periodontal attachment (Wang et al., 1980)
by increasing the diameter of the Sharpey’s fibers. Again, in calvarial
bone, the Sharpey’s fibers are organized relative to the pull of the
masticatory muscles (Simmons et al., 1993), while in spaceflight there is apparently disorganization of the subperiosteal collagen fibrils (Wronski and Morey, 1983; Vailas et al., 1988).
Diagram showing (A)
a stylized CIII/CVI-rich periosteal Sharpey’s fiber with adherent
beaded chains of tenascin and encircled by a coil of elastin, and (B) tracings of the same coarse fibers (about 15 μm diameter) in cross section showing ...
There
now follows seven reasons why the Sharpey’s fiber network may act as an
extracellular regulatory system in bone. Its candidature has been a
lengthy one. Though not assigned as such, elements of the trabecular
framework proposed below probably commenced in the seventeenth century
at the dawn of microscopy with descriptions by Clopton Havers (Dobson, 1952)
of penetrating “fibrillae,” thereby possibly preceding Sharpey himself.
The precise nature of the musculoskeletal exchange mechanism instigated
remains to be established, for example, a piezoelectric phenomenon (the
piezoelectric modulus of tendon is apparently 30-fold that of bone; Marino and Becker, 1971) or one involving stress-regulated excitatory amino acids analogous to neural pathways (Mason et al., 1997)
may be considered; there is also evidence that the ligaments with which
the Sharpey’s fibers integrate may function as proprioceptors (Johansson et al., 1991).
SHARPEY’S FIBERS IN FETAL BONE DEVELOPMENT (FIBRONECTIN AND TENASCIN FACTORS)
A
system of Sharpey’s fibers continuous with the ectodermal membrane is
present from an early embryonic stage. They appear as dorso-ventral
fibrillar bundles, about 1μm thick, containing also CI, fibronectin, and
tenascin. They occupy an area that becomes an intracortical CIII-rich
domain in the limb bud that is linked to tendon generation (Hurle et al., 1989)
and variations from the norm can have pathological consequences. This
is illustrated by comparing intramembranous bone development in the
normal human femoral anlagen with that of dysplastic lesions (Carter et al., 1991).
Key structural molecules in the genesis of new trabeculae are not only
collagen types III and VI, but also adherent are the glycoproteins
tenascin and fibronectin. Regarded as “biological organizer” molecules
they carry the adhesive RGD sequence, fibronectin apparently influencing
fibroblast migration. However, in relation to Sharpey’s fibers it is
tenascin that seems to have a special role, where it may mediate
attachment of osteoblasts by means of its cell recognition signal (Ruoslahti and Pierschbacher, 1986).
The occasional surface location of alkaline phosphatase on some fibers
may relate to this signal and may indicate the expansion of thinner
fibers with circumferential apposition in response to brief loading (Aaron, 1980b). Immunostaining for tenascin indicates that it adopts a highly characteristic beaded pattern (Figure Figure33) the linear alignment of which is critical for normal development, as follows.
Contiguous
with the periosteum surrounding developing intramembranous bone are
arrays of CIII-rich Sharpey’s fibers which apparently form a scaffold
upon which the new trabeculae are assembled and the bone modeling event
takes place. The framework is recognized by antibodies to CI and
fibronectin, but these affinities disappear as the Sharpey’s fibers
become surrounded by calcified bony tissue. Remaining in association,
however, is tenascin in a remarkable regular beaded arrangement. The
intramembranous bone formation can only apparently continue in an
orderly manner toward maturity on condition that tenascin is
specifically associated with the Sharpey’s fibers at this crucial stage.
In its absence the bony tissue is permanently destined to remain
disorganized and immature, as is the case in fibrous dysplasia (Sloan et al., 1989; Carter et al., 1991).
The
preliminary framework appears to persist to maturity (being absent from
endochondrally derived bone) as periosteal myotendinous insertions of
Sharpey’s fibers. By providing this continuous, elastic (Keene et al., 1991)
intermediary between the developing musculature and the developing bone
matrix the CIII fibers may enable the translation of stresses generated
by contractile tissues into compliant modeling and remodeling of the
contiguous trabecular architecture in the femoral anlagen (Pratt, 1959; Wong and Carter, 1990).
It may be envisaged that an understanding of such interactions between
organizing proteins (like tenascin and fibronectin) and extracellular
structures like CIII fibers which are fundamental to early trabecular
development in the first stages of life may direct novel strategies for
restitution of the atrophied skeleton in later life.
SHARPEY’S FIBERS IN MATURE BONE REPAIR FOLLOWING ABLATION (THE ENDOSTEUM FACTOR)
Just
as damage to the adult periosteum stimulates the polarized extension of
its Sharpey’s fibers to re-establish lost continuity, so also does the
endosteum appear to be similarly stimulated when damaged experimentally,
as in the course of tissue ablation of a cylindrical hole in the ovine
pelvic girdle caused by the removal of an 8 mm diameter trephine bone
biopsy (Aaron and Skerry, 1994).
Picking up the damaged threads, and considerably more numerous where
there are bone fragments (a likely source of local growth factors),
there arises from the excised surfaces marshaled arrays of uncalcified,
discrete coarse (5–25 μm) birefringent fibers, converging centripetally.
It is upon this assembly that the replacement primary trabeculae gain
support, and in regions where the scaffold is absent, so also absent is
trabecular genesis. This endosteally derived fibrous framework remains
unmineralized and therefore apparently protected from osteoclastic
resorption (Aaron, 1980a),
aided by other inhibitory intrinsic factors such as CVI. It apparently
survives, even when the thickening primary bars are significantly opened
up by resorption channels into a typical network of mature secondary
trabeculae. The outcome of this endosteal activity is the guaranteed
presence of a persistent fibrillar assembly that crosses domain
boundaries without interruption, bonding soft to hard tissues and new
bone to old, and which seems central to a self-repair process of
admirable efficacy. Thus, the subperiosteal trabecular generation of
embryonic skeletal development in Section “Sharpey’s Fibers in Fetal
Bone Development (Fibronectin and Tenascin Factors)” above is conserved
and recapitulated subendosteally in the adult in response to insult (and
possibly also sclerotic pathology such as Paget’s disease).
....-
an associated decline in muscle insertions. The apparent demise of the proximal domain in this way was accompanied by a significant reduction in both the length and the number of CIII/CVI fibers.
Diagram
of a mature rat femur showing the gross configuration of the expansive
CIII/CVI-rich proximal domain of Sharpey fiber bone (striped area),
terminating in CII/CVI-rich “anchors” (round dots), (A) in a normal control, (B) expanded ...
Inga kommentarer:
Skicka en kommentar