The role of collagen in bone strength

Abstract

Bone is a complex tissue of which the principal function is to resist mechanical forces and fractures. Bone strength depends not only on the quantity of bone tissue but also on the quality, which is characterized by the geometry and the shape of bones, the microarchitecture of the trabecular bones, the turnover, the mineral, and the collagen. Different determinants of bone quality are interrelated, especially the mineral and collagen, and analysis of their specific roles in bone strength is difficult. This review describes the interactions of type I collagen with the mineral and the contribution of the orientations of the collagen fibers when the bone is submitted to mechanical forces. Different processes of maturation of collagen occur in bone, which can result either from enzymatic or nonenzymatic processes. The enzymatic process involves activation of lysyl oxidase, which leads to the formation of immature and mature crosslinks that stabilize the collagen fibrils. Two type of nonenzymatic process are described in type I collagen: the formation of advanced glycation end products due to the accumulation of reducible sugars in bone tissue, and the process of racemization and isomerization in the telopeptide of the collagen. These modifications of collagen are age-related and may impair the mechanical properties of bone. To illustrate the role of the crosslinking process of collagen in bone strength, clinical disorders associated with bone collagen abnormalities and bone fragility, such as osteogenesis imperfecta and osteoporosis, are described.

Full text

REVIEW
The role of collagen in bone strength
S. Viguet-Carrin Æ P. Garnero Æ P.D. Delmas
Received: 15 March 2005 / Accepted: 15 September 2005 / Published online: 9 December 2005
 International Osteoporosis Foundation and National Osteoporosis Foundation 2005
Abstract Bone is a complex tissue of which the principal
function is to resist mechanical forces and fractures.
Bone strength depends not only on the quantity of bone
tissue but also on the quality, which is characterized by
the geometry and the shape of bones, the microarchi-
tecture of the trabecular bones, the turnover, the min-
eral, and the collagen. Different determinants of bone
quality are interrelated, especially the mineral and col-
lagen, and analysis of their specific roles in bone strength
is difficult. This review describes the inte ractions of type
I collagen with the mineral and the contribution of the
orientations of the collagen fibers when the bone is
submitted to mechanical forces. Different processes of
maturation of collagen occur in bone, which can result
either from enzymatic or nonenzymatic processes. The
enzymatic process involves activation of lysyl oxidase,
which leads to the formation of immature and mature
crosslinks that stabilize the collagen fibrils. Two type of
nonenzymatic process are described in type I collagen:
the formation of advanced glycation end products due
to the accumulation of reducible sugars in bone tissue,
and the process of racemization and isomerization in the
telopeptide of the collagen. These modifications of col-
lagen are age-related and may impair the mechanical
properties of bone. To illustrate the role of the cross-
linking process of collagen in bone strength, clinical
disorders associated with bone collagen abnormalities
and bone fragility, such as osteogenesis imperfecta and
osteoporosis, are describe d.
Keywords Bone strength Æ Collagen Æ Crosslinks
(posttranslational modifications of collagen) Æ
Glycation (advanced glycation end products) Æ
Isomerization Æ Osteoporosis
Introduction
Bone is a highly anisotropic, viscoelastic material and
has the ability to continually adapt to changes in its
physiologic or mechanical environm ent. The capacity of
bone to resist mechanical forces and fractures depends
not only on the quantity of bone tissue but also on its
quality. The quantity of bone tissue is in part evaluated
by measuring bone mineral density (BMD) using dual
X-ray absorptiometry. However, determination of BMD
is not an accurate predictor of bone strength. BMD is
certainly a major parameter influencing bone strengt h,
but the three-dimen sional organization of the trabeculae
(microarchitecture), the shape and the geometry of
bones, the potential existence of microdamages, and the
intrinsic properties of the matrix (mineral and collagen)
also contribute to bone strength (Fig. 1). Bone matrix is
a two-phase system in which the mineral phase provides
the stiffness and the collagen fibers provide the ductility
and ability to absorb energy (i.e., the toughness).
Alterations of collagen properties can therefore affect
the mechanical properties of bone and increase fracture
susceptibility. Several studies suggest that part of the
large variation in bone strength may be related to dif-
ferences in the qual ity of the collagenous matrix,
including the nature and extent of its posttranslational
modifications. This review will analyze the contribution
of bone collagen properties to bone strength.
I. Synthesis and structure of bone type I collagen
Collagens constitute a family of proteins present in the
extracellular matrix of connective tissues. They are
Osteoporos Int (2006) 17: 319–336
DOI 10.1007/s00198-005-2035-9
S. Viguet-Carrin Æ P. Garnero Æ P.D. Delmas
INSERM Research Unit 403 and Claude Bernard University,
Lyon, France
P.D. Delmas (&)
INSERM Unit 403, Hoˆ pital E. Herriot, Pavillon F,
69437 Cedex 03 Lyon, France
E-mail: delmas@lyon.inserm.fr
Tel.: +33-4-72117484
Fax: +33-4-72117483

ubiquitous proteins responsible for maintaining the
structural integrity of vertebrates and many other mul-
ticellular organisms. Collagen is constituted by three
polypeptide chains (a-chains) that form a triple-helix
structure. More than 27 forms of collagen are present in
animal tissues. Some of them (types I, II, III, V, and XI)
are arranged in fibrils and are found in tissues that must
be ab le to resist tensile, shear, or compression forces,
including tendon, bone, cartilage, and skin.
Collagen fibrils are characterized by a 67-nm axial
periodicity; they also define the shape of the tissues in
which they occur. Fibrils are arranged in complex three-
dimensional arrays such as orthogonal lattices (e.g., in
the cornea), parall el bundles (e.g., in ligament and ten-
don), and concentric weaves (e.g., in bone). They are
characterized by three interwoven chains (a-chains) that
can be homotrimeric or heterotrimeric, depending on the
collagen type, each chain having a polyproline II-like
conformation [1, 2]. The tight triple helix configuration
is allowed by the repetitive Gly-X-Y triplet, where X can
be any other amino acid, but is usually a proline, and Y
is often a hydroxyproline [1]. Glycine is an absolute
requirement in every third position because it is the
smallest amino acid that can occupy the limited space in
the center of the triple helix.
The fibrils are composed of collagen molecules that
consist of a triple helix approximately 300 nm in length
and 1.5 nm in diameter, flanked by short terminal
globular domains known as the N- and C-propeptides
and that do not exhibit the Gly-X-Y repeat structure.
Proteolytic cleavage of the prop eptides results in triple
helical collagen molecules that have short telopeptides at
each end and can assemble into fibril s [3–5]. All fibrillar
collagens are synthesized and secreted into the extra-
cellular matrix in the form of soluble precursors called
procollagens. The biosynthesis of procollagen is a
complex process in which several enzymes and molecular
chaperones assist its folding and trimerization [6, 7].
Protein disulphide isomerase induces the formation
of inter- and intrachain disulphide bonds within the
C-propeptide, allowing the association between procol-
lagen chains [8, 9].
The C-propeptide ensures the association between
monomeric and heteromeric procollagen chains. Newly
synthesized procollagen chains are associated in trimers
through their C-propeptides, leading to nucleation and
folding in a C-to-N direction to form a triple helix
(Fig. 2). The biosynthesis of procollagen involves dif-
ferent posttranslational modifications that occur in the
endoplasmic reticulum: peptidylproline cis-trans isom-
erase is require d to convert the proline residues to the
trans form [10, 11], and prolyl 4-hydroxylase is required
to convert proline into hydroxyproline residues [12, 13].
The family of lysyl hydroxylase contributes to the for-
mation of hydroxylysine, whic h can be subsequently
further modified by specific enzymes. The collagen
chaperone heat shock protein (HSP) 47 is also required
for the folding of the collagen, but its specific function is
unknown [14, 15].
All of these enzymes responsible for these modifica-
tions act in a coordinated fashion to ensure the folding
and assembly of a correctly aligned and thermally stable
triple-helical molecule (Fig. 2). Once the polypeptide
chain is fully translocated into the lumen of the endo-
plasmic reticulum, the C-propeptide folds. Then, during
the secretion of these molecules into the extracellular
matrix, propeptides are removed by procollagen N and
C proteinases, thereby triggering spontaneous self-
assembly of collagen molecules into fibrils [16]. Thus, the
major extracellular function of the C-propeptides is to
keep the procollagen soluble during its trafficking
through the cell. The N-propeptides do not prevent fibril
formation, but they influence the fibril shape and
diameter. Finally, the triple-helical structure is stabilized
by several posttranslational modifications that allow
intermolecular and interfibrillar crosslinks to take place
between collagen fibrils as a result of the action of lysyl
oxidase (LOX), which catalyzes the crosslinking reaction
by activating lysine and hydroxylysine residues.
Type I collagen {[a1(I)]
2
a2(I)} is the most abundant
type of collagen and is widely distributed in almost all
connective tissues with the exc eption of hyaline cartilage.
It is the major protein in bone, skin, tendon, ligament,
sclera, cornea, and blood vessels. Type I collagen com-
prises approximately 95% of the entire collagen content of
bone and about 80% of the total proteins present in bone
[17]. Other types of collagen, such as types III and V, are
present at low levels in bone [18] and appear to modulate
the fibril diameter. Bone matrix, unlike other connective
tissues, has the unique ability to become calcified. Spindle-
or plate-shaped crystals of hydroxyapatite are found on
the collagen fibers and within or between them. They tend
to be oriented in the same direction as the collagen fibers.
The mechanical properties of bone reflect the inherent
material properties of its constit uents and the way in
which they are arranged and interact. In all connective
tissues, collagen has mechanical functions, providing
elasticity and structure for the component tissues. In
bone, a large body of evidence indicates that type I
collagen molecules are involved in mechanical properties
of bone. Several studies indicate that collagen plays a
substantial role in its toughness (capacity to absorb
energy), while the mineral content is mainly involved in
determining bone stiffness.
Fig. 1 Determinants of bone strength
320

Specific abnormalities of collagen structure can be
induced by genetic mutations (osteogenesis imperfecta)
or pharmacologic agents (lathyric agents). In osteogen-
esis imperfecta, a disease characterized by decreased
material properties and bone fragility, some mutations
in the amino acid sequence of type I collagen can result
in the formation of branched fibers responsible for
brittle bone and abnormal mineralization [19]. When the
formation of crosslinks is inhibited by lathyric agents,
bone strength decreases despite normal mineralization.
Posttranslational modifications of collagen
Collagen biosynthesis involves an unusually large num-
ber of posttranslational events, many of which are spe-
cific to collagens and to some other proteins having
collagen-like structure.
Proline and lysine hydroxylation
In the endoplasmic reticulum, hydroxylation of the
proline and lysine residues occurs on the a-chain before
helix formation (Fig. 2). It is mediated by three specific
enzymes: the prolyl 4-hydroxylase, the prolyl 3-hydrox-
ylase, and the family of lysyl hydroxylases. These en-
zymes hydroxylate about 100 prol ine residues in the Yaa
position, some proline residues in the Xaa position, and
about 10 lysine residues in the Yaa position.
The proline 4-hydroxylase is required to convert
proline residues to hydroxyproline residues. This enzyme
is a tetramere with an a
2
b
2
subunit composition [20, 21]
that requires a-ketoglutarate, Fe
2+
, ascorbic acid, and
molecular oxygen for its activity. Vitamin C (ascorbic
acid) deficiency results in scurvy, characterized by fragile
blood vessels and skin lesions. Proline hydroxylation
contributes to collagen stability because it induces the
formation of hydrogen bonds med iated by bridging
water molecul es [22, 23]. The proline 3-hydroxylase is
less well characterized but has similarly been shown to
act on unfolded collagen [24].
Lysine residues are also hydroxylated by a family of
lysine hydroxylases [25, 26]. Hydroxylysine is necessary
for the intermolecular crosslinking of the molecules in
the fibers. Compared with the skin, type I collagen in
bone tissue has less hydroxyline residue in the helical
regions and more hydroxyline residue in the telopeptide
region. The amino acid sequences around the lysine in
the telopeptide and the triple helix are significantly dif-
ferent, suggesting that different lysyl hydroxylases are
involved. Thre e isoforms of lysyl hydroxylase have been
identified, one of which has been reported to be specific
for the telopeptide in bone collagen [27–30]. Lysine
hydroxylation varies with the rate of collagen turnover
[31]. Hydroxylysine content in normal adult and old
bone is about 25%, but hydroxylation may increase in
osteoporotic bone, suggesting that it may play a role in
bone strength [31].
Hydroxylysine O-glycosylation
Two collagen glycosyltransferases have been character-
ized that bind carbohydrates on hydroxylysine. These
two specific enzymes, called galactosyltransferase and
galactosylhydroxylysyl-glucosyltransferase, act on the
e-hydroxyl group of a hydroxylysine during the bio-
synthesis of procollagen [32, 33]. They result in the
formation of the a-1,2-glucosyl-galactosyl-hydroxylysine
(Glc-Gal-Hyl) and the b-1-galactosyl-hydroxylysine
(Gal-Hyl). As hydroxylases, these enzymes act as long as
the colla gen triple helix is not completely formed. The
degree of glycosylation changes according to the
collagen type and the tissue. For example, bone contains
essentially Gal-Hyl, whereas soft tissues such as
synovium [34] and skin contain mainly Glc-Gal-Hyl [35,
36].
Maturation and aging of bone colla gen
Collagen crosslinks
Once fibers have been formed in the extracellular envi-
ronment, they are further stabilized by the formation of
inter- and intramolecular crosslinks. This process occurs
through the action of LOX [37]. This enzyme acts only
on the extracellular aggregated molecules and recognizes
different binding sites that are similar in the type I, II,
and III collagens and are located in the telopeptides and
the triple-helix. Currently, several isoforms of LOX have
been identified [i.e., LOX-like proteins (LOXL) 1–4)]
[38, 39], but their function and their tissue specificity
remain unclear. Recently, Atsawasuman et al. [40] re-
ported that all isoforms were expressed in MC3T3-E1 in
osteoblastic cells, except for isoform 2. They described a
specific temporal expression pattern of LOX and the
isoforms LOXL 1, 3, and 4 during osteoblastic cell
differentiation.
The process of collagen crosslinking is initiated by the
conversion of telopeptidyl lysine and hydroxylysine
residue to aldehyde through the action of LOX (EC
1.4.3.13, protein-lysine 6-oxidases). LOX is a copper
metalloenzyme that requires pyridoxal phosphate and
tyrosyl-lysine quinone as cofactors [ 41]. Th is en zyme
catalyzes the oxidative deamination of the e-amino
group on lysyl or hydroxylysyl side chains of telopep-
tides, resulting in the formation of two aldehydes, ally-
sine and hydroxyallysine. The hydroxylysine–aldehyde
pathway predo minates in the telopeptides of bone, car-
tilage, ligament, tendon, and most major internal con-
nective tissues [42].
A series of spontaneous nonenzymatic reactions then
occurs in the crosslinking sites within these structural
proteins (Fig. 3). LOX realizes an oxidative deamination
of hydroxyline resid ues in position 9 N (N-terminal
telopeptide) and 16C (C-terminal telopeptide), trans-
forming them in hydroxyallysine. These chemical reac-
tions result in the formation of crosslinks by
321

Fig. 2 Schematic diagram showing the different posttranslational modifications and assembly of type I collagen into fibrils (adapted from
Myllyharju and Kivirikko [3] with permission)
322

condensation of a hydroxyallysine with a lysyl or hy-
droxylysyl side chain at the collagen triple helix in
positions 930 and 87 and create, respectively, two diva-
lent immature crosslinks: lysino-5-ketonorleucine and
the hydroxylysino-5-ketonorleucine [43]. Because these
compounds are unstable, they can be stabilized by
reduction with borohydride to form hydroxylysinonor-
leucine (HLNL) and dihydroxylysinonorleucine
(DHLNL), respectively, which correspond to the
immature crosslinks assayed in different tissues. These
Fig. 3 Pathways of collagen enzymatic crosslinks (adapted from Eyre [57] with permission)
323

divalent crosslinks, which stabilize the immature colla-
gen fibers, further react with another telopeptide alde-
hyde group to form trivalent pyridinium crosslinks.
The conversion of immature to mature crosslinks is a
continuous process independent of turnover rate.
However, the relative amounts of mature and immature
crosslinks are affected by the bone tissue turnover rate.
In bovine skin, there is an initial increase of immature
crosslinks during growth, and then their proportion
relative to the total collagen gradually declines to a low
level at maturity [42]. Bone appear s to be the only tissue
that contains a significant pool of immature crosslinks
[44].
The formation of trivalent mature crosslinks such as
pyridinoline (PYD) and deoxypyridinoline (DPD) from
the hydroxyallysine pathway is still poorly understood.
Some authors have hypothesized that they may result
either in the spontaneous condensation between two
divalent ketoamine crosslinks [45] or a ketoamine
crosslink reacting with a hydroxyallysine residue [46].
The latter model of pyridinoline formation is less limited
in molecular and fibrillar organization of the collagen
than the model proposed by Eyre and Oguchi [45].
Pyridinium crosslinks that are naturally fluorescent
are predominant in mature connective tissues except in
normal skin. They are sensitive to ultraviolet, ozone, and
c radiation [47, 48]. PYD and DPD are formed during
extracellular maturation of collagen fibrils, and they are
the predominant stabilizing bounds in mature tissues.
Although relatively ubiquitous, PYD dominates in
cartilage, whereas DPD is more specific for bone [43]. In
human bone, the content of PYD and DPD reaches a
maximum concentration between 15 and 40 years of age
[44, 49]. In adults, the PYD/DPD ratio remains constant
at about 3.5. Their concentrations are lower in trabec-
ular bone than in cortical bone [44].
Furthermore, collagen posttranslati onal modifica-
tions may differ between the distal and proximal me-
taphysis in the femur of rats [50], with low pyridinium
crosslink content and high glycosylated hydroxylysine
content in the distal metaphysis, contrasting with high
pyridinium crosslinks and low glycosylated hydroxyly-
sine content in the proximal metaphysis and in the
diaphysis. The concentration of crosslinks in the matrix
also differs according to organs and locations, a varia-
tion that may be related to their function. Indeed, hy-
droxypyridinium crosslinks are absent in skin, cornea,
and rat tail tendon, but they are present in load-bearing
tendons of mature bovine Achilles tendon and in liga-
ments [51, 52]. In bone, the level of immature crosslinks
is higher than that of mature PYD and DPD (one per
collagen molecule and one per five collagen molecules,
respectively), a finding that does not appear to
adequately account for the insolubility and strength of
bone collagen. This deficiency of the conversion of
immature to mature pyridinoline suggests the formation
of additional crosslinks from the immature crosslinks.
One of these additional crosslinks may be the trivalent
pyrrole crossl inks formed by the condensation of a ly-
sino-5-ketonorleucine or a hydroxylysino-5-ketonorleu-
cine crosslink with an allysine (Fig. 4).
Pyrrole is present in mineralizing tissues and tendons
[53–56], but it is absent in cartilage [57] and skin. Pyrrole
is often called ‘‘Ehrlich’s chromogen’’ because its
Fig. 4 Mechanism of formation
of the pyrrole and pyridinoline
crosslinks from the divalent
ketoamines and the telopeptide
lysyl aldehyde (adapted from
Bailey and Knott [56] with
permission)
324

detection is based on a reaction with an Ehrlich’s reagent
(p-dimethylaminobenzaldehyde). Unfortunately, the
concentration of pyrrole is difficult to determine because
of its instability. Human bone appears to contain similar
amounts of pyrrole and pyridinoline crosslinks [55].
Pyridinium crosslinks are located at both N- and C-
terminal ends of collagen, whereas pyrrole seems to be
preferentially located at the N-terminal ends of the
molecule [58]. In addition, there are more DPD than
PYD at the N-terminal end of collagen. The difference in
location of the pyridinoline crosslinks and pyrrole sug-
gests different functions for each of these molecules. In
view of the correlation of pyrrole with bone strength,
Knott and Bailey and colleagues [59, 60 ] speculated that
pyrrole forms interfibrillar crosslinks, whereas the pyri-
dinoline would preferentially form intrafibrillar cross-
links.
Most of the ketoimine crosslinks in skin, bone, and
cartilage are glycosylated [61, 62]. Both lysino-5-keto-
norleucine and hydroxylysino-5-ketonorleucine cross-
links give rise to mono- and disaccharide derivatives.
The proportion of glycosylated crosslinks varies widely
between tissues; 50–80% of the crosslinks in rat bone,
cartilage, and skin are glycosylated, but these deriva-
tives are not detected in rat tail tendons [62]. The
3-hydroxypyridinium crosslinks in bone have been
shown to be glycosylated, although the degree of
glycosylation is difficult to measure due to the labil-
ity of the hydroxypyridinium ring during alkaline
hydrolysis.
Nonenzymatic modifications
Advanced glycation end product s
Nonenzymatic glycation is a common posttranslational
modification of proteins mediated by reducing sugars,
which plays an important role in the development of
some age-related diseases and diabetes complications.
Advanced glycation end prod ucts (AGEs) are formed in
vivo via the so-called Maillard reaction [63]: reactiv e
aldose or ketose sugars and metabolic intermediates
(3-deoxyglucosone, glyoxal, and methylglyoxal) react
with free amino groups in lysine, hydroxylysine, or
arginine residues, leading to the formation of a Schiff’s
base, which undergoes rearrangement to form a
relatively stable Amadori product (Fig. 5).
AGEs formation can also be initiated by metal-cat-
alyzed glucose auto-oxidation [64] as well as by lipid
peroxidation. Although there are multiple sources and
numerous pathways leading to AGEs formation in vivo,
most factors playing a role in AGEs formation are
presently unknown. Another consequence of these
highly diverse reaction pathways is that AGEs present a
wide variety of chemical structures that are not yet all
fully characterized [60]. Some AGEs are adducts to
proteins (carboxymethyllysine, carboxyethyllysine),
whereas others present protein-protein crosslinks
(pentosidine, glyoxal-derived lysine dimer, methylgly-
oxal-derived lysine dimer). Pentosidine is one of the best
studied AGEs. This protein crosslinking AGE is derived
from a spontaneous condensation of glucose and the e-
amino groups of lysine and arginine of a protein. Pent-
osidine is a naturally fluorescent AGE that is resistant to
acid hydrolysis and accumulates with aging in different
connective tissues, including bone [65].
AGEs in vivo contribute to some age-related dis-
eases and have been involved in long-term complica-
tions of diabetes [66] and Alzheimer’s disease [67]. The
extent of damages caused by glycation is amplified in
diabetic tissues because of the high serum level of
sugars, which contribute to the vascular complications
by reducing the elasticity and increasing the perme-
ability of blood vessels [ 68]. AGEs are present in tissues
that contain long-lived proteins—proteins with a slow
turnover, such as lens crystallin proteins, colla gen in
the extracellular matrix of connect ive tissues (skin,
tendon, cartilage, and bone), and amyloid plaques.
With aging, AGE modifications are characterized by an
accumulation of yellow and fluorescent material, a
resistance to proteol ytic degradation, and decreased
solubility. In cartilage collagen, the degradation of the
AGE-modified collagen by matrix metalloproteinases is
impaired compared with unmodified collagen [69].
AGE-modified proteins can affect some specific sig-
naling mechanisms such as cell proliferation and/or
differentiation and cell–matrix interactions in different
tissues [70 –72].
AGEs exert their effects not only by altering tissue
turnover but also by adversel y affecting the mechanical
properties of the matrix. The accumulation of AGE
crosslinks in cartilage results in increased stiffness and
brittleness of the tissue [73, 74 ]. Similarly, AGE accu-
mulation is correlated with increa sed tissue stiffness in
lens, arteries, skin, tendon, and bone [75–77], thereby
rendering it more prone to impaired mech anical
properties.
Isomerization and racemization
With aging, racemization and isomerization reactions
on aspartyl acid or asparagine resid ues occur in tissues
with a low metabolic turnover such as bone, lens, brain,
and cartilage. Racemization consists of spontaneous
conversion of the L-enantiomeric form to the biologi-
cally rare D-form, whereas the b-isomerization results
in the transfer of the peptide backbone from the as-
partyl residue a-carboxyl group to the sid e chain b-or
c-carboxyl group [78]. Such transformations occur
through an intermediat e unstable succinimide ring. The
hydrolysis of this intermediate generates four different
isoforms: the native peptide form (aL) and the three
age-related forms, the isomerized form (bL), the race-
mized form (aD), and the isomerized/racemized form
325

(bD).These reactions have been shown to occur during
aging of proteins in vivo and in vitro [79–84]. For
example, the racemization of Asp
23
of the b-amyloid
protein results in the aggregation of b-amyloid fibrils
and may be involved in the pathogenesis of Alzheimer’s
disease [85]. In bone, the a1-chain of the type I collagen
molecule is prone to racemization and isomerization in
vivo on the aspartyl acid residue D
1211
localized in the
CTX epitope from the C-telopeptide [81, 86]. This
spontaneous nonenzymatic reaction leads to the for-
mation of structurally altered forms of the collagen
molecule (Fig. 6).
In summary, isomerization is a slow process that in-
duces conformation modifications of proteins, disrupt-
ing protein regulation or function. In Alzheimer’s
disease, the isomerization of the aspartic residue of the
amyloid protein [80, 82, 87] can eventually lead to pro-
tein inactivation followed by a loss of biological activity
[85]. Thus, studying the maturation and aging process of
collagen in bone is prob ably relevant for understanding
bone strength.
Collagen properties and bone strength
Biomechanical assessment of bone strength
The ability of bone to resist fractures depends on the
amount of bone, its architecture, and its turnover, and
also on the intrinsic properties of the mineral and
organic matrix. Bone strength can be evalu ated by dif-
ferent mechanical tests that simulate in vitro the
mechanical load exe rted on bone. These comprise com-
pressive tests on the vertebra l body, three- or four-point
bending tests on long bones such as femur and humerus,
and shear tests on the femoral neck [88].
In biomechanical tests, a load-deflection curve is
obtained, which allows the determination of different
parameters such as maximum load and deflection and
stiffness of the sample, which corresponds to the slope
of the linear region of the curve (Fig. 7). These
parameters can be normalized after adjusting for the
sample size, allowing load conversion to stress and
deformation to strain, and obtaining a stress-strain
Fig. 5 Pathways of advanced
glycation end products
formation (adapted from
Lapolla et al. [63] with
permission)
326

curve. The slope of the curve within the elastic region is
called Young’s modulus and is an index of the stiffness
of the material tested. The linear portion of the stress-
strain curve is known as the elastic region. After the
yield point, the curve becomes nonlinear and corre-
sponds to the plastic region. In the elastic region there
is a proportional deformation with increasing load ex-
erted, and when the load is removed, bone return s to its
original shape. After the yield point, the stress causes
permanent damage to bone structure. The ultimate
stress corresponds to the maximum stress that bone can
sustain without breaking. The area under the curve is a
measure of the strain energy. The total strain energy
stored at the point of failure is known as the toughness
of bone sample, which represents the amount of energy
needed to cause a fracture.
Collagen organization and bone strength
Whole bone strength is determined by a number of
interrelated variables, which include bone mass and
bone quality (Fig. 1). The latter consists of bone archi-
tecture, the material properties of bone (microdamage
and its repair, the structure of bone mineral and matrix),
and bone turnover, which influences the two other
components. The importance of the components of bone
quality is evident from diseases such as osteogenesis
Fig. 7 Mechanical testing of bone
Fig. 6 Schematic presentation of the Asp-Gly site susceptible to
racemization/isomerization in the type I collagen monomer. The
attack by the peptide backbone nitrogen of the glycine G
1212
on the
side chain carbonyl group of the adjacent aspartic D
1211
residue
occurs through an intermediate unstable succinimide ring. The
spontaneous hydrolysis of the succinimide produces beta-Asp
peptides
327

imperfecta and osteoporosis. In postmenopausal and
idiopathic osteoporosis, there is increasing evidence that
a high remodeling with a negative bone balance induces
bone loss and modifica tion of the microarchitecture,
such as decreased trabecular thickness and loss of con-
nectivity, decreased cortical thickness, and increased
cortical porosity. The high remodeling rate is associated
with decreased bone mineralization that may reduce
bone stiffness and may be associated with a modification
of the content of collagen crosslinks.
Because determinants of bon e quality are intercorre-
lated, it is difficu lt to identify their individual specific
roles. The integrity of structural components of bone
material appears to be important, however, in the
overall mechanical competence of the skeleton.
Collagen and mineral
In bone tissue, collagen fibrils are stiffene d by integra-
tion of the mineral phase [89]. Collagen molecules are
precisely aligned within the fiber in a quarter-staggered
end-overlap fashion (Fig. 2). This arrangement provides
holes within the fiber for nucleation of the calcium
apatite crystals, and these crystals then grow parallel to
the colla gen fibrils. The structure and organization of
collagen fibrils limit the size of crystals and control their
orientation. In diseases such as osteogenesis imperfecta,
extrafibrillar mineral crystals tend to be larger than
crystals found in normal individuals, whereas collagen-
associated crystals are small er than those in age-matched
normal bone [90–92]. The presence of the organ ic phase
in the tissue increases the ultimate tensile strength of
fibrils by about two-fold and the Young modulus by
about ten-fold. In woven bone constituted of unorga-
nized collagen fibrils, the mechanical properties are
lower than in lamellar bone, although the mineral con-
tent is probably higher, demonstrating the importance of
the collagen fibers’ orientations in determining
mechanical properties [93].
Several studies demonstrate that bone strength is
mainly determined by tissue mass and stiffness, which is
determined by the mineral phase [94–96], whereas the
collagen matrix contributes mainly to bone toughness
[31, 97–101]. With age, the properties of the mineral and
the collagen change, altering the material properties of
bone tissue. The different role of collagen in the
mechanical properties of bone is exemplified by in vitro
studies showing that ionizing radiation that dam ages
bone collagen results in decreased bone toughness
without modification of the Young’s modulus [102, 103].
Another in vitro study demonstrated the important
role of the organic matrix for the mechanical properties
of bone. Fantner et al. [104] reported that heat degraded
the organic matrix and, consequently, changed the mi-
crofracture appearance and behavior of trabecular bone,
resulting in decreased elasticity and in toughness.
Degeneration of the organic matrix by heat is thus
responsible for the decreased bone strength. Further
studies are needed to delineate the respective roles of
mineral and collagen in bone strength.
Collagen and orientation loading
Bone strength is explained not only by the interaction
between the mineral phase and collagen but also by the
orientation of collagen fibers according to the direction
of the load. For example, the femur is capable of
resisting significant vertical compressive load with no
significant dam age. In contrast, the same load applied
transversally may cause fractures. Thus, the strength of
bone is higher in the direction of physiological loading
that corresponds to the orientation of osteons in the
cortical bone [105]. Martin et al. [106] reported that
longitudinal fibers are found in regions supporting ten-
sile loads, and transverse fibers correspond to regions
under compressive loading. Differences in the fibril ori-
entation probably account for mechanical properties of
bone [107]. For example, Puustjarvi et al. [108] found
that although long-term running significantly reduces
BMD in dog vertebrae, there was no change in the bone
mechanical prope rties. By using polarized light micros-
copy, they showed a reorganization of collagen fibrils
without modification of the collagen maturation
(crosslinks) or content (hydroxyproline per tissue), sug-
gesting that collagen reorganization during exercise may
contribute to the maintenance of strength despite de-
creased mineral density.
In conclusion, collagen properties intera ct with sev-
eral other determinants of bone strength, which all
contribute to the mechanical properties of bone. The
interactions between these different parameters limit the
analysis of the contribution of bone collagen and espe-
cially of biochemical modifications on bone strength.
These complex interrelationships create a challenge in
investigating the independent role of collagen properties
as a determinant of bone strength.
Posttranslational modifications of collagen and bone
strength
Although several studies have demonstrated the impor-
tance of collagen properties as a determinant of bone
strength, especially in postyield mechanical properties,
the molecular mechanisms involved in collagen
mechanical properties remain unclear. Some studies
suggest that the posttranslational modifications of col-
lagen may be involved. A study of avian cortical bones
suggested that the content of trivalent pyrrole crosslinks
is significantly associated with bending ultimate stress,
whereas no association was found with pyridinoline [53].
Animal models have been used to study the role of
collagen properties. The formation of pyridinoline
crosslinks can be inhibited in young individua ls by
b aminopropionitrile or vitamin B
6
- or copper-deficient
diets, which induce decreased LOX activity and result in
328

a disorder similar to the disease known as lathyrism,
which is characterized by reduced bone strength. In rats
treated with b aminopropionitrile, the bone concentra-
tion of the pyridinoline crosslinks was decreased by 45%
[109], and the deflection capacity at fracture, bending
strength, and stiffness were decreased by 20–30%. In
addition to altering the collagen matrix, b aminopropi-
onitrile decreases bone mineralization in vitro and
in vivo.
Lees et al. [110] reported a decrease in torsional
stiffness and strength of rat femora having abnormal
bone matrix and mineralization. In long bones of rats, a
vitamin B
6
-deficient diet has been shown to result in
decreased torsional stiffness and strength [111]. Simil ar
results have been reported in copper-deficient cockerels,
with a decrease of ultimate torsional strength and a lack
of plastic deformation [112].
Therefore, the covalent LOX-mediated collagen
crosslinks seem to be an important determinant of bone
strength and stiffness. However, because these animal
models are characterized not only by collagen cross-
linking alterations but also by mineral modifications,
they may not allow analysis of the independent contri-
butions of collagen defects to impaired mechanical
properties.
More recently, Banse et al. [113] investigated the
association between the structural organization of tra-
beculae in the human vertebral body and the type of
posttranslational modifications of collagen. They found
that bones with a high pyrrole or low PYD content were
characterized by a thick and apparently disconnected
trabecular structure, whereas those with a low pyrrole or
high PYD content appeared to have a thinner and more
connected structure. The relative concentrations of
pyridinoline and pyrrole crossl inks may reflect the
structural organization of the trabeculae.
The same investigators in the same human vertebra l
model [114] reported that the compressive biomechani-
cal ultimate strength was correlat ed—independently of
BMD—with the ratio PYD/DPD, but not with PYD,
DPD, or pyrrole separately. The interindividual varia-
tion of the PYD/DPD ratio may be explained by the
different rate of lysine hydroxylation. Indeed, crosslinks
formation depends on the extent of hydroxylation of
the specific lysine residues involved in crosslinking,
with pyridinoline formation requiring a high level of
telopeptide hydroxylation. A change in the level of lysine
hydroxylation may therefore result in a change of the
type of crosslinks subsequently formed.
Recently, Uzawa et al. [115] reported that the over-
expression of lysyl hydroxylase-2b (LH-2b) in osteo-
blastic cells was associated with an increase in lysine
hydroxylation of the nonhelical telopeptide domain of
type I collagen and with an increase of the immature and
mature crosslinks [116]. Moreover, this ectopic expres-
sion of LH-2b gene expression induced the formation of
smaller collagen fibrils and led to defective matrix min-
eralization [117]. These experiments underline the rela-
tionships between posttranslational modi fications of
collagen, collagen fibrillogenesis, and mineral deposi-
tion. Altered lysyl hydroxylase activity may change the
ratio between immature and mature crosslinks and
consequently may modify the mechanical properties of
bone.
Wang et al. [76] have recently shown that the
mechanical integrity of collagen fibers from human
cortical bones, assessed after demineralization, deterio-
rates with increasing age and is associated with a
30–50% decrease of the work to fracture, especially its
postyield portion. Thus, the collagen network may be-
come weaker with age, leading to decreased toughness.
The authors reported no significant change with age in
the concentrations of mature enzymatic crosslinks but
reported an increase in pentosidine concentrations,
which was significantly associated with the decreased
mechanical properties of bone (Table 1). Pentosidine is
considered a useful marker of glycation because it is one
of the few AGEs that forms covalent crosslinks between
adjacent molecules and that can be measured accurately.
However, it is not yet known whether a change in the
concentration of pentosidine quantitatively reflects levels
of other, more abundant glycation products that might
affect the material properties of bone.
Vashishth et al. [118] also reported that AGEs may
contribute to age-related loss of bone toughness. Using
an in vitro model of bovine cortical bone incubation
with ribose, Vashishth et al. [77] found that accumula-
tion of AGEs was associated with stiffening of the col-
lagen network with no modification in the tensile and
compressive moduli. In this model, the AGEs induced
increased yield stress and yield strain but did not sig-
nificantly change the postyield properties of cortical
Table 1 Effect of collagen crosslinking on mechanical properties of
bone as function of age. Data were obtained on demineralized
bone. Reproduced from Wang et al. [76] with permission. (E
c
elastic modulus, W
fc
work to fracture of the collagen network, PEN
pentosidine, DPD deoxypyridinoline, PYD pyridinoline, ANOVA
analysis of variance)
Age group E
c
W
fc
PEN DPD PYD
(Mpa) (N.mm) (mmol/mol coll) (mol/mol coll) (mol/mol coll)
Young (<50 years) 20±36.1 63.3±24.2 0.44±0.21 0.188±0.058 0.376±0.076
Middle-aged 207±36.4 48.3±29.1 0.90±0.23
a
0.177±0.041 0.376±0.77
Old (>70 years) 143±33.1
a,b
35.8±17.1
a
1.39±0.29
a, b
0.196±0.061 0.395±0.106
ANOVA p<0.05 p<0.05 p<0.05 p>0.05 p>0.05
a
Statistically significant difference vs. young group (p<0.05)
b
Statistically significant difference vs. middle-aged group (p<0.05)
329

bone. They also conducted some multicyc lic creep tests
on control and in vitro glycated specimens of mineral-
ized and demineralized human cortical bone to identify
the effect of AGE s on bone toughening. They found that
levels of AGEs increase stiffness (29%) and reduce the
creep rate (71%) and maximum strain to failure (47%)
of bone collagen [119, 120]. In vivo and in vitro results
show that modification of collagen by AGEs may alter
the deformation and the damaged behavior of bone, and
thus that collagen may contribute to mechanical prop-
erties of bone, particularly its toughness.
To better understand the independent contribution of
collagen’s mechanical properties to bone biomechanics,
we developed a model based on the incubation of fetal
cortical bone at 37C for up to 3 months, resulting in an
increase of nonenzymatic and enzymatic posttransla-
tional modifications, keeping constant the bone geome-
try, structure, and BMD. There was a two- to three-fold
increase of PYD and DPD content, a fifty-fold increase
of the AGE pentosidine, and a five-fold increase of type I
collagen b-isomerization. These changes were associated
with changes in yield stress and postyield energy
absorption, but did not influence bone stiffness [121, 122].
These in vitro studies indicate that collagen posttransla-
tional modifications may play an important role in bone
strength independently of other determinants, although
it remains to be clarified which crosslinks among the
various ones are most important for bone strength.
In summary, changes in posttranslational modifica-
tions of collagen (occurring during maturation and
aging of bone co llagen) are associated with mechanical
properties of bone. It seems that an increase in enzy-
matic colla gen crosslinks (PYD, DPD, an d pyrrole) may
be associated with increased ultimate stress and stiffness,
whereas the content of AGEs due to the age-related
nature of this modification is associated with postyield
energy absorption. Together, these findings support the
notion that the extent and nature of collagen crosslink-
ing are important contributors to bone quality.
Clinical disorders associated with bone collagen
abnormalities
Osteogenesis imperfecta
Type I collagen mutations observed in osteogenesis im-
perfecta are useful for understanding the contribution of
this protein to bone strength. Osteogenesis imperfecta is
a heritable brittle-bone disease resulting from mutations
in the COL1A1 and COL1A2 genes, which encode for
the a1- and a2-chains of type I collagen, respectively.
Over 100 different mutations have been identified so far,
but the most frequent is a point mutation that affec ts the
conserved glycine residue adjacent to a bulky side chain
amino acid in either COL1A1 or COL1A2. The severity
of the disease depends on which amino acid is substi-
tuted for the glycine, which of the two a-chains is af-
fected, and which position the mutation takes in the
triple helix. As a result, a mixture of normal and
abnormal collagen fibrils are produced in connective
tissues [123]. The collagen fibrils are abnormally thin,
resulting in excessive brittleness of bone. The formation
of thin collagen fibrils in some forms of osteogenesis
imperfecta results from increased lysine hydroxylation
due to a slower triple helix formation [124].
The mouse model of osteogenesis imperfecta, with a
phenotype similar to severe human osteogenesis imper-
fecta, has also been used to analyze the contribution of
the intrinsic properties of bone in its biomechanical
properties. These mice, which suffer from fragility frac-
tures, present with abnormal orientation of collagen fi-
bers an d altered size of the mineral crystal [125]. It is
interesting to note that the most affected mechanical
property is the postyield deformation, which is de-
creased by 60% [126, 127].
Paget’s disease of bone
In Paget’s disease of bone, the normal lamellar bone is
replaced by a woven structure with an irregular
arrangement of collagen fibers. Although pagetic bon es
are usually larger, denser, and more mineralized than
normal bones, they are more fragile, leading to an in-
creased risk of fracture. We have shown that the woven
bone was characterized by a lower degree of type I
collagen isomerization by immunocytochemistry on pa-
getic bone matrix by using monoclonal antibodies spe-
cific for the native (a) and isomerized (b) C-telopeptide
(CTX). Thus, urinary excretion of a CTX is greater than
that of b CTX, resulting in a significant increase of the
a/b ratio in urine of untreated pagetic patients [128]. A
single injection of zoledronate could normalize urinary
excretion of type I collagen isomerized/nonisomerized
CTX, reflecting the progressive replacement of woven
bone by a lamellar bone with a normal degree of isom-
erization under bisphosphonate [129].
Lathyrism
The importance of the formation of crosslinks in the
mechanical functions of collagen is demonstrated in
lathyrism, which occurs in humans and in some animals
after ingesting the seeds of sweet peas, Lathyrus odora-
tus, resulting in severe abnormalities of bones, joints,
and blood vessels due to increased fragility of collagen
fibrils. The factor responsible for lathyrism is the am-
inopropionitrile contained in the seeds, which irrevers-
ibly inhibits LOX activity by binding covalently to its
active site [130, 131].
Osteoporosis
Type I collagen gene polymorphism
Although a familial contribution to osteoporosis and
fracture risk has been clearly established, it is not yet
330

clear which genes are involved. One of the gene poly-
morphisms that has been postulated to play a role in
osteoporosis is the polymorphism of the Sp1 binding site
in the COL1A1 gene [132, 133]. This (G fi T) Sp1
polymorphism may affect COL1A1 gene transcription
and result in alteration of the nature of the collagen
produced. Mann et al. [134] have analyzed the molecular
mechanism underlying the association between the
polymorphism Sp1 binding site in the COL1A1 gene and
osteoporosis. They found that the binding affinity of the
polymorphic ‘‘s’’ allele for the transcription factor Sp1
was increased, resulting in an increased amount of col-
lagen a1 (I)-chain relative to a2 (I). Indeed, the ratio of
collagen was 2.3 in ‘‘Ss’’ heterozygotic patients com-
pared with the expected value of 2 in ‘‘SS’’ homozygotes.
This abnormal ratio of collagen in ‘‘Ss’’ patients re-
ported here is potentially relevant for the pathogenesis
of osteoporotic fractures. Although some studies have
shown a slight association between this ‘‘s’’ allele and
lower BMD, consistent and stronger relationships with
fracture have been reported, suggesting that changes in
the collagen structure induced by this polymorphism are
partly associated with bone strength independen tly of
BMD. Another study demonstrated that it is more rel-
evant to correlate the increased incidence of vertebral
fractures with the ‘‘s’’ allele than with BMD alone [135].
Thus, the COL1A1 Sp1 polymorphism may be asso ci-
ated with both BMD and bone strength.
Osteoporosis and diabetes mellitus
Some studies suggest that diabetes mellitus is associated
with an increased risk of fracture of the hip, foot, or
proximal humerus [136–138]. The altered biomechanical
integrity of the diabetic bones may relate to a number of
factors involving hormonal and vascular factors, bone
structure, bone density, and the interaction between
matrix components. A decreased bone strength that
contributes to fracture risk has been reported in human
and animal studies. In rats with streptozotocin-induced
diabetes, the mechanical integrity of bone tissues has
been studied by a three-point bending test. Reddy et al.
[139] found decreased ultimate strength and toughness
but increased stiffness of long bones compared with
controls. Verhaeghe et al. [140] reported that femoral
bone in spontaneously diabetic rat had decreased tor-
sional strength, angular deformation, and energy
absorption, although the BMD and the bone mineral
content were not significantly decreased.
Type 1 diabetes is associated with a slight reduction
in BMD, whereas type 2 diabetes is often characterized
by elevated BMD. The increased BMD is probably re-
lated to weight-bearing and hormonal factors such as
insulin, estrogens, and leptin. The increased fractur e risk
in type 2 diabetes may be explained by poorer bone
quality. The relationship between diminished biome-
chanical properties of bones and matrix metabolism in
diabetes is not clearly understood. One possible expla-
nation for decreased bone strength in diabetes is the
accumulation of AGEs in bone collagen due to hyper-
glycemia [141].
Role of posttranslational modifications of collagen
in postmenopausal osteoporosis
Several studies have demonstrated that as bone loss in-
creases in postmenopausal osteoporosis, the turnover of
the organic matrix also increases, with an imbalance in
favor of resorption over formation. The increased rate
of colla gen synthesis may also lead to a change in the
quality of collagen fibers, which weakens them. After
menopause, hormonal and systemic factors may also
directly or indirectly modify type I collagen properties.
For example, estrogen deficiency has been suggested to
affect the stability of collagen by decreasing LOX
activity [142]. In healthy persons, TGF-b increases LOX
production but decreases the activity of lysyl hy droxy-
lase [143, 144].
Several biochemical studies performed on bone
specimens taken from patients with osteoporosis have
shown abnormalities in posttranslational modifications
of collagen. An overhydroxylation of lysine residues and
an overglycosylation of hydroxylysine have been re-
ported [145–147], resulting in the formatio n of fibrils of
small diameter [148] that may affect the collagen fiber’s
ability to mineralize normally, leading to decreased bone
strength.
The reduced concentration of reducible crosslinks
may also be associated with reduced bone strength [149].
These authors reported no age-related change of PYD
and DPD concentrations but did observe a decrease in
the concentration of the divalent reducible collagen
crosslinks (DHLNL )30% and HLNL )24%) in oste-
oporotic vertebrae compared with gender- and age-
matched controls. These findings are in agreement with
those of Bailey et al. [31], who found no alteration in the
pyridinoline concentration but a de creased concentra-
tion of immature crosslinks in cancellous bone of fem-
oral head and neck from osteoporotic patients.
Consequently, overhydroxylation and poor cross-
linking of bone collagen may play a role in bone fragility
in osteoporosis. More recently, fourier transform infra-
red imaging analysis was done on forming trabecular
surfaces of human iliac crest biopsy obtained from pa-
tients with high- or low-turnover osteoporosis and in
spontaneously fracturing premenopausal women [150].
The study revealed differences in the spatial distribution
of the pyridinium/reducible collagen crosslinks ratio in
these patients compared with healthy controls. In both
high- and low-turnover osteoporosis, the values of the
mature/immature crosslinks ratio were higher than those
obtained in the normal samples. Premenopausal women
with fractures presented with a collagen crosslinks pro-
file identical to that of the low-turnover osteoporosis
patient group. The difference in the pattern of crosslinks
collagen in patients with fragile bone compared with
331

normal bone suggests a role of collagen crosslinks in
bone strength [150].
In the OFELY population-based cohort of post-
menopausal women, we reported that the urinary
excretion of the aL/bL CTX ratio and the aL/aD ratio
was increased in 65 women with incident fractures
compared with 343 women without fractures [151]. The
ratio between native and age-related forms of urinary
CTX was associated with a twofold increased risk of
fracture independently of femoral neck BMD or bone
turnover (Table 2). These results suggest that decreased
racemization/isomerization of type I colla gen reflected
by an increased aL/bLandaL/aD ratio could be asso-
ciated with alterations in the structure or maturation of
collagen fibers, leading to increased skeletal fragility
independently of BMD.
In summary, in postmenopausal osteoporosis there is
growing evidence that at the material level, the volume
fraction of mineral and the relative amounts of mature
and immature collagen crosslinks are affected by the
tissue turnover rate, thus contributing to bone fragility.
Because the extent and the nature of collagen cross-
linking are important contributors to bone matrix
quality, it is important to perform additional studies to
delineate their independent contributions to collagen
and, ultimately, bone strength.
Conclusion
Bone is a sophisticated composite material with complex
relationships between mineral and collagen that influ-
ence bone strength. A change in collagen prop erties may
alter the amount and disposition of the mineral, which
would by itself affect bone mechanics. Knowledge is
lacking concerning the precise mechanisms leading to
the binding between mineral and collagen and their
contribution to mechanical properties. The rate of
turnover of bone collagen may be important in deter-
mining bone strength because it influences the pattern of
mature/immature crosslinking in bone.
Future studies should be performed to investigate the
influence of collagen posttranslational modifications and
bone strength, taking into account the contribution of
tissue turnover as well as other determi nants of bone
strength such as mineral composition and distribution,
microfracture accumulation and repair, and osteocyte
viability. Finally, the development of noninvasive bio-
chemical markers reflecting collagen ‘‘quality’’ (such as
crosslinking and isomerization pattern) would be useful
for studying in vivo the abnormalities of the collagen
matrix in osteoporosis and their role in skeletal fragility.
Acknowledgement We thank D. Herbage for thoroughly reviewing
this manuscript.
References
1. Van der Rest M, Garrone R (1991) Collagen family proteins.
FASEB J 5:2814–2823
2. Brodsky B, Shah NK (1995) The triple helix motif in proteins.
FASEB J 9:1537–1546
3. Myllyharju J, Kivirikko KI (2004) Collagens, modifying en-
zymes and their mutations in humans, flies and worms.
Trends Genet 20:33–43
4. Dominguez LJ, Barbagallo M, Moro L (2005) Collagen
overglycosylation: a biochemical feature that may contribute
to bone quality. Biochem Biophys Res Commun 330:1–4
5. Kuznetsova N and Leikin S (1999) Does the triple helical
domain of type I collagen encode molecular recognition
and fiber assembly while telopeptides serve as catalytic
domains? Effect of proteolytic cleavage on fibrillogenesis
and on collagen-collagen interaction in fibers. J Biol Chem
274:36083–36088
6. Lamande SR, Bateman JF (1999) Procollagen folding and
assembly: the role of endoplasmic reticulum enzymes and
molecular chaperones. Semin Cell Dev Biol 10:455–464
7. Hendershot LM, Bulleid NJ (2000) Protein-specific chaper-
ones: the role of HSP47 begins to gel. Curr Biol 10:912–915
8. Koivu J 1987 Identification of disulphides bonds in carboxy-
terminal propeptides of human type I procollagen. FEBS Lett
212:229–232
9. Bulleid NJ, Dalley JA, Lees JF 1997 The C-propeptide do-
main of procollagen can be replaced with a transmembrane
domain without affecting trimer formation or collagen triple
helix folding during biosynthesis. EMBO J 16:6694–6701
10. Bachinger HP (1987) The influence of peptidyl-prolyl cis-trans
isomerase on the in vitro folding of type III collagen. J Biol
Chem 262:17144–17148
11. Galat A, Metcalfe SM (1995) Peptidylproline cis/trans
isomerases. Prog Biophys Mol Biol 63:67–118
12. Kivirikko KI, Myllyharju J (1998) Prolyl 4-hydroxylases and
their protein disulfide isomerase subunit. Matrix Biol 16:357–
368
13. Myllyharju J (2003) Prolyl 4-hydroxylases, the key enzymes of
collagen biosynthesis. Matrix Biol 22:15–24
14. Nagata K (1998) Expression and function of heat shock
protein 47: a collagen-specific molecular chaperone in the
endoplasmic reticulum. Matrix Biol 16:379–386
15. Nagata K (2003) HSP47 as a collagen-specific molecular
chaperone: function and expression in normal mouse devel-
opment. Semin Cell Dev Biol 14:275–14282
16. Engel J, Prockop DJ (1991) The zipper-like folding of colla-
gen triple helices and the effects of mutations that disrupt the
zipper. Annu Rev Biophys Biophys Chem 20:137–152
Table 2 Urinary CTX ratio of women with incident osteoporotic
fractures and controls in the OFELY postmenopausal women.
Reproduced from Garnero et al. [128] with permission. Data are
shown as mean±SD (CTX C-telopeptide)
Urinary
CTX ratio
Women with fracture Controls p-value
(n=65) (n=343)
Urinary CTX level (lmmol/mmol Cr)
a-L 5.00±2.89 4.12±2.39 0.0053
b-L 9.76±3.60 9. 22±4.09 0.14
a-D 2.56±1.15 2.54±1.09 0.76
b-D 5.68±2.03 5.44±2.17 0.39
Total
a
22.8±8.0 21.2±8.5 0.10
Urinary CTX ratio
a-L/b-L 0.53±0.30 0.45±0.20 0.0055
a-L/a-D 2.10±1.15 1.67±0.80 0.0014
a-L/b-D 0.91±0.44 0.76±0.31 0.0018
a
Total CTX level was obtained by making the sum of the a-L-CTX,
b-L-CTX, a-D-CTXand b-D-CTX.
332

17. Niyibizi C, Eyre DR (1994) Structural characteristics of cross-
linking sites in type V collagen of bone chain specificities and
heterotypic links to type I collagen. Eur J Biochem 224:943–950
18. Niyibizi C, Eyre DR (1989) Bone type V collagen: chain
composition and location of a trypsin cleavage site. Connect
Tissue Res 20:247–250
19. Baum J and Brodsky B (1999) Folding of peptide models of
collagen and misfolding in disease. Curr Opin Struct Biol
9:122–128
20. Koivu J, Myllyla R, Helaakoski T, Pihlajaniemi T, Tasanen
K, Kivirikko KI (1987) A single polypeptide acts both as the
beta subunit of prolyl 4-hydroxylase and as a protein disul-
fide-isomerase. J Biol Chem 262:6447–9
21. John DC, Grant ME, Bulleid NJ (1993) Cell-free synthesis
and assembly of prolyl 4-hydroxylase: the role of the beta-
subunit (PDI) in preventing misfolding and aggregation of the
alpha-subunit. EMBO J 12:1587–1595
22. Privalov PL (1982) Stability of proteins. Proteins which do
not present a single cooperative system. Adv Protein Chem
35:1–104
23. Bella J, Brodsky B, Berman HM (1995) Hydration structure
of a collagen peptide. Structure 3:893–906
24. Vranka JA, Sakai LY, Bachinger HP (2004) Prolyl 3-
hydroxylase 1, enzyme characterization and identification of a
novel family of enzymes. J Biol Chem. 279:23615–21
25. Valtavaara M, Szpirer C, Szpirer J, Myllyla R (1998) Primary
structure, tissue distribution, and chromosomal localization
of a novel isoform of lysyl hydroxylase (lysyl hydroxylase 3).
J Biol Chem 273:12881–12886
26. Wang C, Luosujarvi H, Heikkinen J, Risteli M, Uitto L,
Myllyla R (2002) The third activity for lysyl hydroxylase 3:
galactosylation of hydroxylysyl residues in collagens in vitro.
Matrix Biol 21:559–566
27. Passoja K, Rautavuoma K, Ala-Kokko L, Kosonen T, Ki-
virikko KI (1998) Cloning and characterization of a third
lysyl hydroxylase isoform. Proc Natl Acad Sci USA
95:10482–10486
28. Mercer DK, Nicol PF, Kimbembe C, Robins SP (2003)
Identification, expression and tissue distribution of three rat
lysyl hydroxylase isoforms. Biochem Biophys Res Commun
307:803–809
29. Bank RA, Robins SP, Wijmenga C, Breslau-Siderius LJ,
Bardoel AF, van der Sluijs HA, Pruijs HE, TeKoppele JM
(1999) Defective collagen cross-linking in bone, but not in
ligament or cartilage, in Bruck syndrome: indication for a
bone specific telopeptide lysyl hydroxylase on chromosome
17. Proc Natl Acad Sci USA 96:1054–1058
30. Ha-Vinh R, Alanay Y, Bank RA, Campos-Xavier AB, Zankl
A, Superti-Furga A, Bonafe L (2004) Phenotypic and molecu-
lar characterization of Bruck syndrome (osteogenesis imper-
fecta with contractures of the large joints) caused by a recessive
mutation in PLOD2. Am J Med Genet A 131A:115–120
31. Bailey AJ, Wotton SF, Sims TJ, Thompson PW (1992) Post
translational modifications in the collagen of human osteopo-
rotic femoralhead. Biochem BiophysRes Commun 185:801–805
32. Harwood R, Grant ME, Jackson DS (1974) Collagen bio-
synthesis. Characterization of subcellular fractions from em-
byonic chick fibroblasts and the intracellular localization of
protocollagen prolyl and protocollagen lysyl hydroxylases.
Biochem J 144:123–130
33. Harwood R, Grant ME, Jackson DS (1975) Studies on the
glycosylation of hydroxylysine residues during collagen bio-
synthesis and the subcellular localization of collagen galac-
tosyltransferase and collagen glucosyltransferase in tendon
and cartilage cells. Biochem J 152:291–302
34. Gineyts E, Garnero P, Delmas PD (2001) Urinary excretion
of glucosyl-galactosyl pyridinoline: a specific biochemical
marker of synovium degradation. Rheumatology (Oxford)
40:315–323
35. Segrest JP, CunninghamLW (1970) Variations in human
urinary O-hydroxylysyl glycoside levels and their relationship
to collagen metabolism. J Clin Invest 49:1497–1509
36. Pinnell SR, Fox R, Krane SM (1971) Human collagens: dif-
ferences in glycosylated hydroxylysines in skin and bone.
Biochim Biophys Acta 229:119–122
37. Smith-Mungo L, Kagan HM (1998) Lysyl oxidase: properties,
regulation and multiple functions in biology. Matrix Biol
16:387–398
38. Maki JM, Tikkanen H, Kivirikko KI (2001) Cloning and char-
acterization of the fifth human lysyl oxidase enzyme: the third
member of the lysyl oxidase related subfamily with four scav-
enger repector cystein-rich domains. Matrix Biol 20:493–496
39. Csiszar K (2001) Lysyl oxidases: a novel multifunctional amine
oxidase family. Prog Nucleic Acid Res Mol Biol 70:1–32
40. Atsawasuwan P, Mochida Y, Parisuthiman D, Yamauchi M
(2005) Expression of lysyl oxidase isoforms in MC3T3-E1
osteoblastic cells. Biochem Biophys Res Commun 327:1042–
1046
41. Wang SX, Mure M, Medzihradszky KF, Burlingame AL,
Brown DE, Dooley DM, Smith AJ, Kagan HM, Klinman JP
(1996) A crosslinked cofactor in lysyl oxidase: redox function
for amino acid side chains Science 273:1078–1084
42. Robins SP, Shimokomaki M, Bailey AJ (1973) The chemistry
of the collagen crosslinks. Age-related changes in the reduc-
ible components of intact bovine collagen fibres. Biochem
J 131:771–780
43. Eyre DR, Paz MA, Gallop PM (1984) Cross-linking in col-
lagen and elastin. Annu. Rev Biochem 53:717–748
44. Eyre DR, Dickson IR, Van Ness K (1988) Collagen cross-
linking in human bone and articular cartilage. Age-related
changes in the content of mature hydroxypyridinium residues.
Biochem J 252:495–500
45. Eyre DR, Oguchi H (1980) The hydroxypyridinium crosslinks
of skeletal collagen: their measurement, properties and a
proposed pathway of formation. Biochem Biophys Res
Commun 92:403–410
46. Robins SP, Ducan A (1983) Cross-linking of the collagen.
Location of pyridinoline in bovine articular cartilage at two
sites of the molecules. Biochem J 215:167–173
47. Fujimori E (1985) Ultraviolet light- and ozone-induced
changes in pyridinoline, a trisubstituted 3-hydroxypyridinium
crosslink of collagen. Biochim Biophys Acta 828:104–106
48. Colwell A, Hamer A, Blumsohn A, Eastell R (1996) To
determine the effects of ultraviolet light, natural light and
ionizing radiation on pyridinium cross-links in bone and urine
using high-performance liquid chromatography. Eur J Clin
Invest 26:1107–1114
49. Saito M, Marumo K, Fujii K, Ishioka N (1997) Single-col-
umn high-performance liquid chromatographic-fluorescence
detection of immature, mature, and senescent cross-links of
collagen. Anal Biochem 253:26–32
50. Moro L, Romanello M, Favia A, Lamanna MP, Lozupone E
(2000) Posttranslational modifications of bone collagen type I
are related to the function of rat femoral regions. Calcif
Tissue Int 66:151–156
51. Bailey AJ, Peach CM (1968) Isolation and structural identi-
fication of a labile intermolecular crosslink in collagen Bio-
chem Biophys Res Commun 33:812–819
52. Barnes MJ, Constable BJ, Morton LF, Kodicek E (1971)
Hydroxylysine in the N-terminal regions of the 1- and 2-
chains of various collagens. Biochem J 125:433–437
53. Knott L, Whitehead CC, Fleming RH, Bailey AJ (1995)
Biochemical changes in the collagenous matrix of osteopo-
rotic avian bone. Biochem J 310:1045–1051
54. Kuypers R, Tyler M, Kurth LB, Jenkins ID, Horgan DJ
(1992) Identification of the loci of the collagen-associated
Ehrlich chromogen in type I collagen confirms its role as a
trivalent cross-link. Biochem J 283:129–136
55. Ristelli J, Eriksen H, Risteli L, Mansell JP, Bayley AJ (1994)
Pyrrolic crosslinks are as abundant in human bone type I col-
lagen as pyridinolines. J Bone Miner Res 9(Suppl 1):S186
56. Bailey AJ, Knott L (1999) Molecular changes in bone colla-
gen in osteoporosis and osteoarthritis in the elderly. Exp
Gerontol 1999 34:337–351
333

57. Eyre DR (1987) Collagen cross-linking amino acids. Methods
Enzymol 144:115–139
58. Hanson DA, Eyre DR (1996) Molecular site specificity of
pyridinoline and pyrrole crosslinks in type I collagen of hu-
man bone. J Biol Chem 271:26508–26516
59. Knott L, Bailey AJ (1998) Collagen cross-links in mineralising
tissues : a review of their chemistry, function, and clinical
relevance. Bone 22:181–187
60. Bailey AJ, Paul RG, Knott L (1998) Mechanisms of matu-
ration and ageing of collagen. Mech Ageing Dev 106:1–56
61. Eyre DR, Glimcher MJ (1973) Analysis of a crosslinked
peptide from calf bone collagen: evidence that hydroxylysyl
glycoside participates in the crosslink. Biochem Biophys Res
Commun 52:663–671
62. Robins SP, Bailey AJ (1974) Isolation and characterization of
glycosyl derivatives of the reducible cross-links in collagens.
FEBS Lett 38:334–336
63. Lapolla A, Traldi P, Fedele D (2005) Importance of mea-
suring products of non-enzymatic glycation of proteins. Clin
Biochem. 2005 38:103–115
64. Wolff SP,Dean RT (1987) Glucose autoxidation and protein
modification. The potential role of ‘‘autoxidative glycosyla-
tion’’ in diabetes. Biochem J 245:243–50
65. Sell DR, Monnier VM (1989) Structure elucidation of a
senescence crosslink from human extracellular matrix: impli-
cation of pentoses in the aging process. J Biol Chem
264:21597–21602
66. Monnier VM, Cerami A (1983) Detection of nonenzymatic
browning products in the human lens. Biochim Biophys Acta
760:97–103
67. Markesbery WR (1997) Oxidative stress hypothesis in Alz-
heimer’s disease. Free Ra Biol Med 23:134–147
68. Brownlee M, Cerami A, Vlassara H (1988) Advanced glyco-
sysalation end products in tissue and the biochemical basis of
diabetic complications. N Engl J Med 318:1315–1321
69. DeGroot J, Verzijl N, Wenting-Van Wijk MJ, Bank RA,
Lafeber FP, Bijlsma JW, TeKoppele JM (2001) Age-related
decrease in susceptibility of human articular cartilage to
matrix metalloproteinase-mediated degradation: the role of
advanced glycation end products. Arthritis Rheum 44:2562–
2571
70. Brownlee M (1994) Lilly Lecture 1993. Glycation and diabetic
complications. Diabetes 43:836–841
71. Katayama Y, Akatsu T, Yamamoto M, Kugai N, Nagata N
(1996) Role of nonenzymatic glycosylation of type I collagen
in diabetic osteopenia. J Bone Miner Res 11:931–937
72. Miyata T, Kawai R, Taketomi S, Sprague SM (1996) Possible
involvement of advanced glycation end-products in bone
resorption. Nephrol Dial Transplant 11:54S-57S
73. Verzijl N, DeGroot J, Ben ZC, Brau-Benjamin O, Maroudas
A, Bank RA, Mizrahi J, Schalkwijk CG, Thorpe SR, Baynes
JW, Bijlsma JW, Lafeber FP, TeKoppele JM (2002) Cross-
linking by advanced glycation end products increases the
stiffness of the collagen network in human articular cartilage:
a possible mechanism through which age is a risk factor for
osteoarthritis. Arthritis Rheum 46:114–123
74. Chen AC, Temple MM, Ng DM, Verzijl N, DeGroot J,
TeKoppele JM, Sah RL (2002) Induction of advanced gly-
cation end products and alterations of the tensile properties of
articular cartilage. Arthritis Rheum 46:3212–3217
75. Tomasek JJ, Meyers SW, Basinger JB, Green DT, Shew RL
(1994) Diabetic and age-related enhancement of collagen-
linked fluorescence in cortical bone of rats. Life Sci 55:855–861
76. Wang X, Shen X, Li X, Agrawal CM (2002) Age-related
changes in the collagen network and toughness of bone. Bone
31:1–7
77. Vashishth D, Gibson GJ, Khoury JI, Schaffler MB, Kimura J,
Fyhrie DP (2001) Influence of nonenzymatic glycation on
biomechanical properties of cortical bone. Bone 28:195–201
78. Clarke S (1987) Propensity for spontaneous succinimide for-
mation from aspartyl and asparaginyl residues in cellular
proteins. Int J Pept Protein Res 30:808–821
79. Maroudas A, Bayliss MT, Uchitel-Kaushansky N, Schnei-
derman R, Gilav E (1998) Aggrecan turnover in human
articular cartilage: use of aspartic acid racemization as a
marker of molecular age. Arch Biochem Biophys 350:61–71
80. Saido TC, Iwatsubo T, Mann DM, Shimada H, Ihara Y,
Kawashima S. (1995) Dominant and differential deposition of
distinct beta-amyloid peptide species, A beta N3(pE), in senile
plaques. Neuron 14:457–466
81. Fledelius C, Johnsen AH, Cloos PA, Bonde M, Qvist P (1997)
Characterisation of urinary degradation products derived
from type I collagen. Identification of a beta-isomerized Asp-
Gly sequence within the C-terminal telopeptide (alpha1) re-
gion. J Biol Chem 272:9755–9763
82. Roher AE, Lowenson JD, Clarke S, Wolkow C, Wang R,
Cotter RJ, Reardon IM, Zurcher-Neely HA, Heinrikson RL,
Ball MJ (1993) Structural alterations in the peptide backbone
of beta-amyloid core protein may account for its deposition
and stability in Alzheimer’s disease. J Biol Chem 268:3072–
3083
83. Geiger T and Clarke S (1987) Deamination, isomerization,
and racemization at asparaginyl and aspartyl residues in
peptides. Succinimide-linked reactions that contribute to
protein degradation. J Biol Chem 262:785–794
84. Gineyts E, Cloos PA, Borel O, Grimaud L, Delmas PD,
Garnero P (2000) Racemization and isomerisation of type I
collagen C-telopeptides in human bone and soft tissues :
assessment of tissue turnover. Biochem J 345:481–485
85. Tomiyama T, Asano S, Furiya Y, Shirasawa T, Endo N,
Mori H (1994) Racemization of Asp
23
residue affects the
aggregation of Alzheimer amyloid beta protein analogues.
J Biol Chem 269:10205–10208
86. Cloos PA, Fledelius C (2000) Collagen fragments in urine
derived from bone resorption are highly racemized and
isomerised: a biological clock of protein aging with clinical
potential. Biochem J 345:473–480
87. Shimizu T, Fukuda H, Murayama S, Izumiyama N, Shiras-
awa T. (2002) Isoaspartate formation at position 23 of amy-
loid beta peptide enhanced fibril formation and deposited
onto senile plaques and vascular amyloids in Alzheimer’s
disease. J Neurosci Res 70:451–461
88. Turner CH, Burr DB (1993) Basic biomechanical measure-
ments of bone: a tutorial. Bone 14:595–608
89. Ottani V, Raspanti M, Ruggeri A (2001) Collagen structure
and functional implications. Micron 32:251–260
90. Fratzl P, Paris O, Klaushofer K, Landis WJ (1996) Bone
mineralization in an osteogenesis imperfecta mouse model
studied by small-angle x-ray scattering. J Clin Invest 97:396–
402
91. Vetter U, Eanes ED, Kopp JB, Termine JD, Robey PG (1991)
Changes in apatite crystal size in bones of patients with
osteogenesis imperfecta. Calcif Tissue Int 49:248–250
92. Traub W, Arad T, Vetter U, Weiner S (1994) Ultrastructural
studies of bones from patients with osteogenesis imperfecta.
Matrix Biol 14:337–345
93. Marotti G, Muglia MA, Palumbo C (1994) Structure and
function of lamellar bone. Clin Rheumatol 13 Suppl 1:63–68
94. Currey JD (1979) Change in impact energy absorption of
bone with age. J Biomech 12:459–469
95. Currey JD (1988) The effect of porosity and mineral content
on the Young’s modulus of elasticity of compact bone.
J Biochem 21:131–139
96. Currey JD, Brear K, Zioupos P (1996) The effects of aging
and changes in mineral content in degrading the toughness of
human femora. J Biomech 29:257–260
97. Boskey AL, Wright TM, Blank RD (1999) Collagen and bone
strength. J Bone Miner Res 14:330–335
98. Wang X, Bank RA, TeKoppele JM, Agrawal CM (2001) The
role of collagen in determining bone mechanical properties.
J Orthop Res 19:1021–1026
99. Zioupos P, Currey JD, Hamer AJ (1999) The role of collagen
in the declining mechanical properties of aging human corti-
cal bone. J Biomed Mater Res 45:108–116
334

100. Thompson JB, Kindt JH, Drake B, Hansma HG, Morse DE,
Hansma PK (2001) Bone indentation recovery time correlates
with bond reforming time. Nature 414:773–776
101. Zioupos P (2001) human bone: factors affecting its biome-
chanical properties and the role of collagen. J Biomater Appl
15:187–229
102. Currey JD, Foreman J, Laketic I, Mitchell J, Pegg DE, Reilly
GC (1997) Effects of ionizing radiation on the mechanical
properties of human bone. J Orthop Res 15:111–117
103. Currey JD (2003) Role of collagen and other organics in the
mechanical properties of bone. Osteoporos Int 14:S29-S36
104. Fantner GE, Birkedal H, Kindt JH, Hassenkam T, Wea-
ver JC, Cutroni JA, Bosma BL, Bawazer L, Finch MM,
Cidade GA, Morse DE, Stucky GD, Hansma PK (2004)
Influence of the degradation of the organic matrix on the
microscopic fracture behavior of trabecular bone. Bone
35:1013–1022
105. Hert J, Fiala P, Petrtyl M (1994) Osteon orientation of the
diaphysis of the long bones in man. Bone 15:269–277
106. Martin RB, Burr DB, Sharkey NA (1998) Skeletal tissue
mechanics. Springer, Berlin Heidelberg New York
107. Martin RB and Boardman DL (1993) The effects of collagen
fiber orientation, porosity, density, and mineralization on
bovine cortical bone bending properties. J Biomech 26:1047–
1054
108. Puustjarvi K, Nieminen J, Rasanen T, Hyttinen M, Helminen
HJ, Kroger H, Huuskonen J, Alhava E, Kovanen V (2003)
Do more highly organized collagen fibrils increase bone
mechanical strength in loss of mineral density after one-year
running training? J Bone Miner Res 14:321–329
109. Oxlund H, Barckman M, Ortoft G, Andreassen TT (1995)
Reduced concentrations of collagen cross-links are associated
with reduced strength of bone. Bone 17:365S–371S
110. Lees S, Hanson D, Page E, Mook HA (1994) Comparison of
dosage-dependent effects of beta-aminopropionitrile, sodium
fluoride, and hydrocortisone on selected physical properties of
cortical bone. J Bone Miner Res 9:1377–1389
111. Masse PG, Rimnac CM, Yamauchi M, Coburn SP, Rucker
RB, Howell DS, Boskey AL (1996) Pyridoxine deficiency af-
fects biomechanical properties of chick tibial bone. Bone
18:567–574
112. Opsahl W, Zeronian H, Ellison M, Lewis D, Rucker RB,
Riggins RS (1982) Role of copper in collagen cross-linking
and its influence on selected mechanical properties on selected
mechanical properties of chick bone and tendon. J Nutr
112:708–716
113. Banse X, Devogelaer JP, Lafosse A, Sims TJ, Grynpas M,
Bailey AJ (2002) Cross-link profile of bone collagen correlates
with structural organization of trabeculae. Bone 31:70–76
114. Banse X, Sims TJ, Bailey AJ (2002) Mechanical properties of
adult vertebral cancellous bone: correlation with collagen
intermolecular cross-links. J Bone Miner Res 17:1621–1628
115. Uzawa K, Grzesik WJ, Nishiura T, Kuznetsov SA, Robey
PG, Brenner DA, Yamauchi M (1999) Differential expression
of human lysyl hydroxylase genes, lysine hydroxylation, and
cross-linking of type I collagen during osteoblastic differen-
tiation in vitro. J Bone Miner Res 14:1272–1280
116. Pornprasertsuk S, Duarte WR, Mochida Y, Yamauchi M
(2004) Lysyl hydroxylase-2b directs collagen cross-linking
pathways in MC3T3-E1 cells. J Bone Miner Res 19:1349–1355
117. Pornprasertsuk S, Duarte WR, Mochida Y, Yamauchi M
(2005) Overexpression of lysyl hydroxylase-2b leads to
defective collagen fibrillogenesis and matrix mineralization.
J Bone Miner Res 20:81–87
118. Vashishth D, Wu P, Gibson GJ (2003) Age-related loss in
bone toughness is explained by non-enzymatic glycation of
collagen. 49th annual meeting of the Orthopaedic Research
Society, New Orleans
119. Boxberger J, Vashishth D (2003) Nonenzymatic glycation
affects bone fracture by modifying creep and inelastic prop-
erties of collagen. 49th annual meeting of the Orthopaedic
Research Society, New Orleans
120. Wu P, Koharski C, Nonnemann H, Vashishth D (2003)
Loading of non-enzymatically glycated and damaged bone
results in an instantaneous fracture. 49th annual meeting of
the Orthopaedic Research Society, New Orleans
121. Garnero P, Proust Y, Borel O, Gineyts E, Duboeuf F, Solberg
H, Delmas P (2003) Collagen crosslinking modifies the
mechanical properties of cortical bone. J Bone Miner Res
18(Suppl 2):S196
122. Garnero P, Borel O, Gineyts E, Duboeuf F, Christiansen C,
Delmas P (2004) The degree of posttranslational modifica-
tions of collagen is an important determinant of the toughness
of cortical bone. J Bone Miner Res 19(Suppl 1):S220
123. Rauch F, Glorieux F (2004) Osteogenesis imperfecta. Lancet
363:1377–1385
124. Kirsch E, Krieg T, Remberger K, Fendel H, Bruckner P,
Muller PK (1981) Disorder of collagen metabolism in a pa-
tient with osteogenesis imperfecta (lethal type): increased de-
gree of hydroxylation of lysine in collagen type I and III. Eur
J Clin Invest 11:39–47
125. Grabner B, Landis WJ, Roschger P, Rinnerthaler S, Peterlik
H, Klaushofer K, Fratzl P (2001) Age- and genotype-depen-
dence of bone material properties in the osteogenesis imper-
fecta murine model (oim). Bone 29:453–457
126. Jepsen KJ, Schaffler MB, Kuhn JL, Goulet RW, Bonadio J,
Goldstein SA (1997) Type I collagen mutation alters the
strength and fatigue behavior of Mov 13 cortical tissue.
J Biomech 30:1141–1147
127. Pereira RF, Hume EL, Halford KW, Prockop DJ (1995)
Bone fragility in transgenic mice expressing a mutated gene
for type I procollagen (COL1A1) parallels the age-dependent
phenotype of human osteogenesis imperfecta. J Bone Miner
Res 10:1837–1843
128. Garnero P, Fledelius C, Gineyts E, Serre CM, Vignot E,
Delmas PD (1997) Decreased beta-isomerization of the C-
terminal telopeptide of type I collagen alpha 1 chain in Paget’s
disease of bone. J Bone Miner Res 12:1407–1415
129. Garnero P, Gineyts E, Schaffer AV, Seaman J, Delmas PD
(1998) Measurement of urinary excretion of nonisomerized an
beta-isomerized forms of type I collagen breackdown prod-
ucts to monitor the effects of the bisphosphonate zoledronate
in Paget’s disease. Arthritis Rheum 41:354–360
130. Mikkonen L, Tuominen T and Kulonen E (1960) Collagen
fractions in lathyritic rats. Bioch Pharma 3:181–183
131. Lees S, Barnard S, Mook H (1987) Neutron studies of col-
lagen in lathyritic bone. Int J Biol Macromol 9:32–38
132. Grant SF, Reid DM, Blake G, Herd R, Fogelman I, Ralston
SH (1996) Reduced bone density and osteoporosis associated
with a polymorphic Sp1 binding site in the collagen type I
alpha 1 gene. Nature Genet 14:40–45
133. Mann V and Ralston SH (2003) Meta-analysis of COL1A1
Sp1 polymorphism in relation to bone mineral density and
osteoporotic fracture. Bone 32:711–717
134. Mann V, Hobson EE, Li B, Stewart TL, Grant SF, Robins
SP, Aspden RM, Ralston SH (2001) A COL1A Sp1 bind-
ing site polymorphism predisposes to osteoporotic fracture
by affecting bone density and quality. J Clin Invest
107:899–907
135. Uitterlinden AG, Burger H, Huang Q, Yue F, McGuigan FE,
Grant SF, Hofman A, van Leeuwen JP, Pols HA, Ralston SH
(1998) Relation of alleles of the collagen type I alpha1 gene to
bone density and the risk of osteoporotic fractures in post-
menopausal women. N Engl J Med 338:1016–1021
136. Schwartz AV, Sellmeyer DE, Ensrud KE, Cauley JA, Tabor
HK, Schreiner PJ, Jamal SA, Black DM, Cummings SR (2001)
Older women with diabetes have an increased risk of fractures:
a prospective study. J Clin Endocrinol Metab 86:32–38
137. Nicodemus KK, Folsom AR (2001) Type 1 and type 2 dia-
betes and incident hip fractures in postmenopausal women.
Diabetes Care 24:1192–1197
138. Keegan TH, Kelsey JL, Sidney S, Quesenberry CP Jr (2002)
Foot problems as risk factors of fractures. Am J Epidemiol
155:926–931
335

139. Reddy KG, Stehno-bittel L, Hamade S, Enwemeka CS (2001)
The biomechanical integrity of bone in experimemtal diabe-
tes. Diabetes Res Cli Pract 54:1–8
140. Verhaeghe J, Suiker AM, Einhorn TA, Geusens P, Visser WJ,
Van Herck E, Van Bree R, Magitsky S, Bouillon R (1994)
Brittle bones in spontaneously diabetic female rats cannot be
predicted by bone mineral measurements: studies in diabetic
and ovariectomized rats. J Bone Miner Res 9:1657–1667
141. Paul RG, Bailey AJ (1996) Glycation of collagen: the basis of
its central role in the late complications of ageing and dia-
betes. Int J Biochem Cell Biol 28:1297–1310
142. Sanada H, Shikata J, Hamamoto H, Ueba Y, Yamamuro T,
Takeda T (1978) Changes in collagen cross-linking and lysyl
oxidase by estrogen. Biochim Biophys Acta 541:408–413
143. Feres-Filho EJ, Choi YJ, Han X, Takala TE, Trackman PC
(1995) Pre-translational and post-translational regulation of
lysyl oxidase by transforming growth-factor-beta-1 in osteo-
blastic MC3T3-E1 cells. J Biol Chem 270:30797–30803
144. Seitzer U, Batge B, Acil Y, Muller PK (1995) Transforming
growth-factor beta 1 influences lysyl hydroxylation of colla-
gen I and reduces steady-state level of lysyl hydroxylase
mRNA in human osteoblast-like cells. Eur J Clin Invest
25:959–966
145. Kowitz J, Knippel M, Schuhr T, Mach J (1997) Alteration in
the extent of collagen I hydroxylation, isolated from femoral
heads of women with a femoral neck fracture caused by
osteoporosis. Calcif Tissue Int 60:501–505
146. Batge B, Diebold J, Stein H, Bodo M, Muller PK (1992)
Compositionnal analysis of the collagenous bone matrix. A
study on adult normal and osteopenic bone tissue. Eur J Clin
Invest 22:805–812
147. Bailey AJ, Wotton SF, Sims TJ, Thompson PW (1993) Bio-
chemical changes in the collagen of human osteoporotic bone
matrix. Connect Tissue Res 29:119–132
148. Torre-Blanco A, Adachi E, Hojima Y, Wootton JA, Minor
RR, Prockop DJ (1992) Temperature induced post-transla-
tional over-modification of type I collagen. Effects of
over-modification of the protein on the rate of cleavage by
procollagen N-proteinase and on self-assembly of collagen
into fibrils. J Biol Chem 267:2650–2655
149. Oxlund H, Mosekilde L, Ortoft G (1996) Reduced concen-
tration of collagen reducible crosslinks in human trabecular
bone with respect to age and osteoporosis. Bone 19:479–484
150. Paschalis EP, Shane E, Lyritis G, Skarantavos G, Mendel-
sohn R, Boskey AL (2004) Bone fragility and collagen cross-
links. J Bone Miner Res 19:2000–2004
151. Garnero P, Cloos P, Sornay-Rendu E, Qvist P, Delmas PD
(2002) Type I collagen racemization and isomerisation and
the risk of fracture in postmenopausal women: the OFELY
prospective study. J Bone Miner Res 17:826–833
336