Clin Biochem, Vol. 25, pp. 175-180, 1992 Printed in the USA. All righLs reserved.
0009-9120/92 $5.00 + .00 Copyright c 1992 The Canadian Society of Clinical Chemists.
Serum Keratan Sulfate A Marker of Predisposition to Polyarticular Osteoarthritis EUGENE J-M.A. THONAR 1'2 and TIBOR GLANT 1'2'3 Departments of 1Biochemistry, 2Internal Medicine (Section of Rheumatology), and 3Orthopedic Surgery, Rush Medical College, Rush-Presbyterian-St. Luke's Medical Center, 1653 W. Congress Parkway, Chicago, IL 60612-3864, USA We have used an ELISA to quantify a highly sulfated epitope present on keratan sulfate, a carbohydrate chain found principally in cartilage proteoglycans. The serum level of the epitope provides an indirect measure of the rate of degradation of cartilage proteoglycans during normal turnover and can be used to diagnose specific abnormalities in keratan sulfate metabolism. Serum levels of the epitope are elevated in a high percentage of patients with osteoarthritis and correlate with the number of joints involved. The elevated rate of proteoglycan turnover in these patients appears to be systemic, affecting not only the degenerating articular surfaces but apparently normal articular cartilages as well. We have postulated that this acceleration in the rate of proteoglycan turnover precedes clinical evidence of degenerative changes; and we discuss the rationale for the contention that this elevation may predispose adult humans to polyarticular osteoarthritis.
KEY WORDS: cartilage; ELISA; keratan sulfate, osteoarthritis; proteoglycan; proteoglycan catabolism; serum marker; aggrecan. Metabolism of cartilage aggrecan he extracellular matrix of cartilage contains sev-
eral populations of proteoglycans (PGs) that enT able the tissue to undergo rapid and reversible deformation. Aggrecan is, by far, the most abundant type of PG found in cartilage (1). It consists of a core protein to which two t y p e s of very negativelycharged g l y c o s a m i n o g l y c a n side chains are covalently attached: chondroitin sulfate, a chain that is prominent in the PGs of most tissues, and keratan sulfate (KS), a chain found almost exclusively (>95%) in cartilage aggrecan (1). These glycosaminoglycans are present at very high concentrations in the tissue; they give aggrecan its ability to undergo changes in shape and the tissue its viscoelasticity. Newly-synthesized aggrecan molecules interact extracellularly with hyaluronan (HA), an unsulfated long glycosaminoglycan chain, and link proteins to
Correspondence: Eugene J-M.A. Thonar, Ph.D., Department of Biochemistry, Rush Medical College, RushPresbyterian-St. Luke's Medical Center, 1653 W. Congress Parkway, Chicago, IL 60612-3864, USA. Manuscript received July 31, 1991; revised November 27, 1991; accepted December 5, 1991. CLINICAL BIOCHEMISTRY, VOLUME 25, J U N E 1992
form aggregates of very large size ( M r = > 1 0 7 ) . These aggregates are entrapped very effectively within the abundant fibrous network comprised of several types of collagens (2). In normal h u m a n adult cartilage, aggregates that are degraded during normal turnover are replaced by newly-synthesized molecules, thereby maintaining a constant PG content. The rate of degradation of the PGs is regulated by the cells that secrete metalloproteinases, such as stromelysin, into the extracellular milieu in latent form as well as inhibitors of these enzymes. Most glycosaminoglycan-bearing fragments that are generated via the action of these metalloproteinases diffuse rapidly out of the tissue (1); less than 1% appears to be taken up by the cells and degraded intracellularly. The fate of these fragments, once they appear in body fluids, is unclear at the present time: a large proportion of the fragments are thought to be eliminated by the lymphatic system but some eventually reach the circulation before being cleared by the liver and kidneys (3). Measurement of keratan sulfate in body fluids by an ELISA Measurements of aggrecan-derived fragments in blood and synovial fluid can provide important information about the degradation of aggregating PGs in cartilage (1,3). The fragments in blood are present in low levels but can be quantified by an ELISA that measures a highly-sulfated epitope present on the KS chains. The level of this epitope in blood is difficult to interpret since the aggrecan fragments present in this body fluid are derived from the many different cartilages, that is, hyaline, elastic, and fibrous cartilages, which are all rich in aggrecan (3). The KS epitope is present at a much higher concentration and on larger fragments in synovial fluid than in blood; in the former, its concentration is relatively high and provides a useful measure of the degradation of aggrecan molecules within the cartilages in that joint. In body fluids KS is usually quantified by a competitive indirect ELISA that uses the wellcharacterized 1/20/5-D-4 anti-KS monoclonal anti175
THONAR AND GLANT
body (4). This antibody recognizes a highly sulfated sequence of several repeats of the disaccharide (6sulfated N-acetyl glucosamine linked [31-3 to 6-sulfated galactose) present in long KS chains distributed over the entire length of the polysaccharideattachment region (1). The shortest chains do not contain an epitope while the longest KS chains may contain an uninterrupted sequence of disulfated disaccharides that can bind to at least two antibody molecules (5) as shown in Figure 1. K e r a t a n sulfate related epitopes are easier to quantify accurately than protein-related epitopes. For example, the HA-binding region, a globular domain present at the N-terminal end of the core protein of aggrecan, must be denatured prior to quantification of the epitopes present within it to prevent binding of this region to HA, link protein, or other HA-binding region molecules; this is necessary as such interactions have been shown to result in the masking or unmasking of epitopes (6). While denaturation is easy to achieve, the need to remove the denaturing agent, which can inactivate the antibody or suppress a n t i g e n - a n t i b o d y interactions, is a complicating factor. The ELISA we developed to quantify the sulfated KS epitope (3) was modified recently after discovering that the anti-KS antibody bound to antigenic KS chains with greater avidity at pH 5.3 than at pH 7.0 (7). This modification yields s t e e p e r inhibition curves for both standards and unknowns and thus increases the ability to discriminate between concentrations of antigenic KS that are not markedly different (1). The ELISA gives a good measure of the epitope irrespective of the size of the molecule on which it is present (1). For example, pretreatment of aggrecan with keratanase I (Pseudomonas sp.) and endo-[3-galactosidase (Escherichia freundii), enzymes that cleave KS chains in non-antigenic regions, causes no detectable loss of KS epitope in the
C
o R E
' . 1 ,
P R O T
I
i
i
i
i
T
t
_- e ~ , - G -
i
"
'/
KS epitopes
E I N • Galactooe • N-acetylglucosamine 0
:~
~lactosamlne Neuramlnic acid
Figure 1 -- Schematic representation of the tentative structure of antigenic and nonantigenic KS chains. The longest KS chain is shown here as containing 2 contiguous epitopes on the basis of its demonstrated ability to bind to at least 2 anti-keratan sulfate (1/20/5-D-4) antibody molecules. The negatively charged sulfate groups on galactose and N-acetylglucosamine are represented as short vertical lines. Reproduced from Ref. (5) with the permission of Raven Press, New York. 176
ELISA. On the other hand, similar t r e a t m e n t with keratanase II, a new KS-degrading enzyme (Seikagaku Company, Japan) that cleaves the [31-3 linkage between 6-sulfated N - a c e t y g l u c o s a m i n e and 6-sulfated galactose (8), causes total loss of antigenicity. It is important to note that not all anti-KS monoclonal antibodies can be used to quantify the KS epitope in serum by the competitive indirect ELISA; some of the antibodies bind with much lower affinity to KS epitopes on soluble PG fragments than to the same fragments insolubilized on the plastic (1). The ELISA is very sensitive. Quantification in duplicate requires less than 50 ~L of serum or plasma. The epitope is extremely stable in serum; samples stored at - 6 0 °C for several years or frozen and thawed several times show no detectable loss in antigenicity (1). Care should be taken to avoid hemolysis of red blood cells that may cause some loss in antigenicity when the hemolysis is very pronounced. Since the KS epitope recognized by the 1/20/5-D-4 antibody is not available in large quantities, the results in all of our studies are reported in terms of equivalents of an international standard of KS purified from h u m a n costal cartilage. One should avoid using undefined commercial standards of KS; those we tested were not pure and apparently had lost much of their antigenicity during purification. Human and animal studies of keratan sulfate in serum
Keratan sulfate bearing PG fragments injected intravenously in rabbits are eliminated from the blood at different rates depending upon their size and composition with half-lives of 6 - 5 0 min (1). Exposure of terminal galactose residues on KS chains and oligosaccharides causes a marked increase in the rate of elimination of fragments of all sizes, suggesting that binding of the fragments to galactose receptors on liver cells is an important if not major p a t h w a y of elimination from the blood (1). As KS-bearing molecules isolated from h u m a n blood had the longest half-life (approximately 50 min), it was suggested that they may not provide a true reflection of the structure and composition of the KS-bearing fragments that enter the circulation (1). F u t u r e studies should help determine if these KS-bearing fragments that predominate in blood are a major or, more likely, only a minor subpopulation of the KSbearing molecules that enter the circulation. The serum level of KS is surprisingly constant in normal adults (9). It shows no apparent variation during the day or from day-to-day. This strongly suggests that the rate of turnover of aggrecan molecules is under very tight regulation and is not readily modified by m a n y of the factors which in vitro have been shown to affect the metabolism of cartilage PGs. It does not change significantly in inexperienced runners following a 3-mile run or in experienced runners after a marathon (1). Bed rest is not accompanied by a detectable change in the CLINICAL BIOCHEMISTRY, VOLUME 25, JUNE 1992
SERUM KS: A MARKER OF PREDISPOSITION TO OA
level of serum KS either. However, if bed rest is maintained for several months, the level eventually shows a significant decrease (10). Interestingly, the serum level returns to the pre-inactivity level following remobilization. These results suggest that measurement at a single time point provides a good indirect estimate of the rate of PG turnover in any individual. In adult life, levels show a tendency to increase with age (3,11). Similar age-related changes have been detected in many different m a m m a l i a n species (1). Within the adult population, levels vary, sometimes markedly, from individual to individual (1). The individual variation in the level of serum KS epitope within the adult population is thought to reflect differences in the rate of cartilage aggrecan catabolism/turnover in different individuals (1). The contention that an adult who has an abnormally high level of serum KS epitope has a higher rate of cartilage PG catabolism/turnover is based on several assumptions. First, one cannot rule out at this stage the possibility that such a difference in the serum level does not correspond, at least in part, to a difference in either the proportion of cartilagederived K S - b e a r i n g f r a g m e n t s t h a t reaches the blood or in the rate of clearance of the fragments from the blood. Second, it is possible that adults with elevated levels of serum KS epitope have KS chains that are much more antigenic. A recent observation (12) that serum levels of the KS epitope show a good correlation w i t h the s e r u m level of an epitope present on the core protein of aggrecan suggests the level of the KS epitope probably offers a good measure of total KS present. Studies are in progress to measure total KS (antigenic + nonantigenic) in serum to determine if, as hypothesized, the level of antigenic KS in serum provides an accurate measure of total KS present. The serum level of the KS epitope has been shown to provide important information about abnormalities or changes in the metabolism of KS-bearing PGs in a wide variety of conditions [see (1) for review]. Absence of the KS epitope in serum has proved useful in diagnosing type I macular corneal dystrophy, a blinding disease caused by an abnormality in sulfation of the keratan chain on corneal PGs (13-15). Interestingly, the KS chains in cartilage aggrecan are similarly unsulfated but thus far, this has not been found to result in any abnormalities in cartilagenous structures. M e a s u r e m e n t of the serum level of the KS epitope has also proved useful in studying abnormalities in the metabolism of polylactosamines that, as KS, contain long segments of the repeat disaccharide N-acetylglucosamine-galactose (16), and in assessing tumor grade in patients with chondrosarcomas. Levels of the s e r u m KS epitope undergo marked age-related changes during growth (17) and show a positive correlation with percentile height, supporting the contention that the rate of growth is a function of the activity of the chondrocytes (17). Children with constitutional delay of m a t u r i t y (below the fifth percentile for height CLINICAL BIOCHEMISTRY, VOLUME 25, JUNE 1992
but not growth hormone deficient) and children who are growth hormone deficient have abnormally low levels (18). Administration of growth hormone to growth hormone-deficient children resulted in a significant rise in the level of serum KS together with expected increases in annualized growth rate velocity and plasma levels of IGF-I (18). In normal adults, the ratio of cartilage mass/blood volume, the concentration of KS in cartilage, and the size of the KS chains show only m o d e r a t e changes with increasing age (1). The observation that the serum level of the KS epitope in adults is extremely constant with time is therefore not surprising. The level of serum KS epitope does not change significantly following t r e a t m e n t with the non-steroidal anti-inflammatory drugs piroxicam or naproxen, suggesting that these two NSAIDs have little if any effect on the rate of cartilage PG catabolism (19). In contrast, prednisone ( 4 0 - 6 0 mg daily), given orally for 3 days to patients attending an allergy clinic, causes a rapid decrease in the level of serum KS epitope that is already evident after 24 h and reaches a maximum on day 4 with a mean KS level at 61% of the pretreatment level (19). Importantly, the level of serum KS epitope was still only 69% of the pretreatment level 32 days after treatment was stopped. A single intraarticular injection of corticosteroids in the knee joint of patients with osteoarthritis (OA) caused a similar decrease in the level of the KS epitope in serum (20). Since KS present in serum is derived from all cartilagenous structures in the body, a slight change in the rate of PG catabolism in a single synovial joint is unlikely to give rise to a measurable increase in the level of KS epitope in serum. However, recent studies have shown that massive and rapid degradation of the cartilage PGs in a single synovial joint or intervertebral disc causes a large transient rise in the level of serum KS (7,9,21-24). Physicians injecting chymopapain into intervertebral discs for the treatment of a herniated disc have taken advantage of this approach to obtain evidence that the enzyme was injected in the correct place and that the PGs in the tissue have been degraded.
Keratan sulfate: A marker of predisposition to generalized osteoarthritis An elevation in the rate of degradation of PGs contributes to the progressive degeneration of articular cartilage in OA (1). This rise in the rate of PG catabolism appears to begin while the condition is clinically silent. In most cases, it is counterbalanced in its early stages by a corresponding increase in PG synthesis (25). Studies of levels of serum KS epitope suggest that this state of hypermetabolism develops during middle age in some individuals without any evidence of joint disease. We have postulated that it may predispose to OA (1,26,27). This contention is based on recent evidence that a proportion of the HA-binding regions (also called G1 regions or domains), which are generated during PG turnover by 177
THONARAND GLANT
the action of stromelysin on the core protein of the PGs, may remain attached to HA, thereby reducing the number of sites on HA molecules that are accessible to newly-synthesized PGs (1,28,29) as shown in Figure 2. These HA-binding regions that have been shown to accumulate in articular cartilage with age (28,29) contain no or few glycosaminoglycan chains and therefore contribute little to the functional properties of the tissue. With aging, this probably leads to depletion of glycosaminoglycans in some regions of the articular surface, predisposing collagen fibrils that are no longer supported by viscoelastic glycosaminoglycans to fracture, especially in maximally loaded regions. It is likely that in individuals who have high levels of the serum KS epitope and who are turning over their cartilage PGs at accelerated rates, the nonfunctional HA-binding regions accumulate at a faster rate, thereby predisposing tissues to failure and degeneration in regions that are maximally loaded (1). Recent studies have shown that serum levels of the KS epitope are higher in patients with polyarticular OA than in age-matched individuals without signs of joint disease (26,27,30,31) as shown in Figure 3. These results strongly support the widely held belief that aggrecan molecules in OA cartilages are being degraded at abnormally fast rates. It is worth noting that a proportion of patients with OA do not have elevated levels of the KS epitope while a small percentage of adults without any clinical signs of OA do (26). Consequently, a single m e a s u r e m e n t of the serum level of the KS epitope is not a useful marker of the articular cartilage destruction that is
/•A
Collagen
Figure 2 - - Hypothesized effect of age-related accumulation ofhyaluronate-binding region fragments in cartilage. Some of the HA-binding regions (G1 regions or domains) generated by the degradation of aggrecan by stromelysin remain attached to HA; these nonfunctional fragments occupy sites that are normally reserved for glycosaminoglycan-rich PG molecules. Note the difference in the hydrodynamic domains occupied by PGs tightly packed within the collagenous network (left hand side) and by PGs whose swelling capacity is no longer restricted (right hand side). With aging, the accumulation of HA-binding regions may lead to depletion of PGs within some regions of the matrix; as a consequence collagen fibrils that are not supported by the viscoelastic PGs are more likely to fracture (see the text for additional discussion). 178
OA
N
1000
80O
E i,i
Z
6OO
400
nil,
200
p
0009-9120/92 $5.00 + .00 Copyright c 1992 The Canadian Society of Clinical Chemists.
Serum Keratan Sulfate A Marker of Predisposition to Polyarticular Osteoarthritis EUGENE J-M.A. THONAR 1'2 and TIBOR GLANT 1'2'3 Departments of 1Biochemistry, 2Internal Medicine (Section of Rheumatology), and 3Orthopedic Surgery, Rush Medical College, Rush-Presbyterian-St. Luke's Medical Center, 1653 W. Congress Parkway, Chicago, IL 60612-3864, USA We have used an ELISA to quantify a highly sulfated epitope present on keratan sulfate, a carbohydrate chain found principally in cartilage proteoglycans. The serum level of the epitope provides an indirect measure of the rate of degradation of cartilage proteoglycans during normal turnover and can be used to diagnose specific abnormalities in keratan sulfate metabolism. Serum levels of the epitope are elevated in a high percentage of patients with osteoarthritis and correlate with the number of joints involved. The elevated rate of proteoglycan turnover in these patients appears to be systemic, affecting not only the degenerating articular surfaces but apparently normal articular cartilages as well. We have postulated that this acceleration in the rate of proteoglycan turnover precedes clinical evidence of degenerative changes; and we discuss the rationale for the contention that this elevation may predispose adult humans to polyarticular osteoarthritis.
KEY WORDS: cartilage; ELISA; keratan sulfate, osteoarthritis; proteoglycan; proteoglycan catabolism; serum marker; aggrecan. Metabolism of cartilage aggrecan he extracellular matrix of cartilage contains sev-
eral populations of proteoglycans (PGs) that enT able the tissue to undergo rapid and reversible deformation. Aggrecan is, by far, the most abundant type of PG found in cartilage (1). It consists of a core protein to which two t y p e s of very negativelycharged g l y c o s a m i n o g l y c a n side chains are covalently attached: chondroitin sulfate, a chain that is prominent in the PGs of most tissues, and keratan sulfate (KS), a chain found almost exclusively (>95%) in cartilage aggrecan (1). These glycosaminoglycans are present at very high concentrations in the tissue; they give aggrecan its ability to undergo changes in shape and the tissue its viscoelasticity. Newly-synthesized aggrecan molecules interact extracellularly with hyaluronan (HA), an unsulfated long glycosaminoglycan chain, and link proteins to
Correspondence: Eugene J-M.A. Thonar, Ph.D., Department of Biochemistry, Rush Medical College, RushPresbyterian-St. Luke's Medical Center, 1653 W. Congress Parkway, Chicago, IL 60612-3864, USA. Manuscript received July 31, 1991; revised November 27, 1991; accepted December 5, 1991. CLINICAL BIOCHEMISTRY, VOLUME 25, J U N E 1992
form aggregates of very large size ( M r = > 1 0 7 ) . These aggregates are entrapped very effectively within the abundant fibrous network comprised of several types of collagens (2). In normal h u m a n adult cartilage, aggregates that are degraded during normal turnover are replaced by newly-synthesized molecules, thereby maintaining a constant PG content. The rate of degradation of the PGs is regulated by the cells that secrete metalloproteinases, such as stromelysin, into the extracellular milieu in latent form as well as inhibitors of these enzymes. Most glycosaminoglycan-bearing fragments that are generated via the action of these metalloproteinases diffuse rapidly out of the tissue (1); less than 1% appears to be taken up by the cells and degraded intracellularly. The fate of these fragments, once they appear in body fluids, is unclear at the present time: a large proportion of the fragments are thought to be eliminated by the lymphatic system but some eventually reach the circulation before being cleared by the liver and kidneys (3). Measurement of keratan sulfate in body fluids by an ELISA Measurements of aggrecan-derived fragments in blood and synovial fluid can provide important information about the degradation of aggregating PGs in cartilage (1,3). The fragments in blood are present in low levels but can be quantified by an ELISA that measures a highly-sulfated epitope present on the KS chains. The level of this epitope in blood is difficult to interpret since the aggrecan fragments present in this body fluid are derived from the many different cartilages, that is, hyaline, elastic, and fibrous cartilages, which are all rich in aggrecan (3). The KS epitope is present at a much higher concentration and on larger fragments in synovial fluid than in blood; in the former, its concentration is relatively high and provides a useful measure of the degradation of aggrecan molecules within the cartilages in that joint. In body fluids KS is usually quantified by a competitive indirect ELISA that uses the wellcharacterized 1/20/5-D-4 anti-KS monoclonal anti175
THONAR AND GLANT
body (4). This antibody recognizes a highly sulfated sequence of several repeats of the disaccharide (6sulfated N-acetyl glucosamine linked [31-3 to 6-sulfated galactose) present in long KS chains distributed over the entire length of the polysaccharideattachment region (1). The shortest chains do not contain an epitope while the longest KS chains may contain an uninterrupted sequence of disulfated disaccharides that can bind to at least two antibody molecules (5) as shown in Figure 1. K e r a t a n sulfate related epitopes are easier to quantify accurately than protein-related epitopes. For example, the HA-binding region, a globular domain present at the N-terminal end of the core protein of aggrecan, must be denatured prior to quantification of the epitopes present within it to prevent binding of this region to HA, link protein, or other HA-binding region molecules; this is necessary as such interactions have been shown to result in the masking or unmasking of epitopes (6). While denaturation is easy to achieve, the need to remove the denaturing agent, which can inactivate the antibody or suppress a n t i g e n - a n t i b o d y interactions, is a complicating factor. The ELISA we developed to quantify the sulfated KS epitope (3) was modified recently after discovering that the anti-KS antibody bound to antigenic KS chains with greater avidity at pH 5.3 than at pH 7.0 (7). This modification yields s t e e p e r inhibition curves for both standards and unknowns and thus increases the ability to discriminate between concentrations of antigenic KS that are not markedly different (1). The ELISA gives a good measure of the epitope irrespective of the size of the molecule on which it is present (1). For example, pretreatment of aggrecan with keratanase I (Pseudomonas sp.) and endo-[3-galactosidase (Escherichia freundii), enzymes that cleave KS chains in non-antigenic regions, causes no detectable loss of KS epitope in the
C
o R E
' . 1 ,
P R O T
I
i
i
i
i
T
t
_- e ~ , - G -
i
"
'/
KS epitopes
E I N • Galactooe • N-acetylglucosamine 0
:~
~lactosamlne Neuramlnic acid
Figure 1 -- Schematic representation of the tentative structure of antigenic and nonantigenic KS chains. The longest KS chain is shown here as containing 2 contiguous epitopes on the basis of its demonstrated ability to bind to at least 2 anti-keratan sulfate (1/20/5-D-4) antibody molecules. The negatively charged sulfate groups on galactose and N-acetylglucosamine are represented as short vertical lines. Reproduced from Ref. (5) with the permission of Raven Press, New York. 176
ELISA. On the other hand, similar t r e a t m e n t with keratanase II, a new KS-degrading enzyme (Seikagaku Company, Japan) that cleaves the [31-3 linkage between 6-sulfated N - a c e t y g l u c o s a m i n e and 6-sulfated galactose (8), causes total loss of antigenicity. It is important to note that not all anti-KS monoclonal antibodies can be used to quantify the KS epitope in serum by the competitive indirect ELISA; some of the antibodies bind with much lower affinity to KS epitopes on soluble PG fragments than to the same fragments insolubilized on the plastic (1). The ELISA is very sensitive. Quantification in duplicate requires less than 50 ~L of serum or plasma. The epitope is extremely stable in serum; samples stored at - 6 0 °C for several years or frozen and thawed several times show no detectable loss in antigenicity (1). Care should be taken to avoid hemolysis of red blood cells that may cause some loss in antigenicity when the hemolysis is very pronounced. Since the KS epitope recognized by the 1/20/5-D-4 antibody is not available in large quantities, the results in all of our studies are reported in terms of equivalents of an international standard of KS purified from h u m a n costal cartilage. One should avoid using undefined commercial standards of KS; those we tested were not pure and apparently had lost much of their antigenicity during purification. Human and animal studies of keratan sulfate in serum
Keratan sulfate bearing PG fragments injected intravenously in rabbits are eliminated from the blood at different rates depending upon their size and composition with half-lives of 6 - 5 0 min (1). Exposure of terminal galactose residues on KS chains and oligosaccharides causes a marked increase in the rate of elimination of fragments of all sizes, suggesting that binding of the fragments to galactose receptors on liver cells is an important if not major p a t h w a y of elimination from the blood (1). As KS-bearing molecules isolated from h u m a n blood had the longest half-life (approximately 50 min), it was suggested that they may not provide a true reflection of the structure and composition of the KS-bearing fragments that enter the circulation (1). F u t u r e studies should help determine if these KS-bearing fragments that predominate in blood are a major or, more likely, only a minor subpopulation of the KSbearing molecules that enter the circulation. The serum level of KS is surprisingly constant in normal adults (9). It shows no apparent variation during the day or from day-to-day. This strongly suggests that the rate of turnover of aggrecan molecules is under very tight regulation and is not readily modified by m a n y of the factors which in vitro have been shown to affect the metabolism of cartilage PGs. It does not change significantly in inexperienced runners following a 3-mile run or in experienced runners after a marathon (1). Bed rest is not accompanied by a detectable change in the CLINICAL BIOCHEMISTRY, VOLUME 25, JUNE 1992
SERUM KS: A MARKER OF PREDISPOSITION TO OA
level of serum KS either. However, if bed rest is maintained for several months, the level eventually shows a significant decrease (10). Interestingly, the serum level returns to the pre-inactivity level following remobilization. These results suggest that measurement at a single time point provides a good indirect estimate of the rate of PG turnover in any individual. In adult life, levels show a tendency to increase with age (3,11). Similar age-related changes have been detected in many different m a m m a l i a n species (1). Within the adult population, levels vary, sometimes markedly, from individual to individual (1). The individual variation in the level of serum KS epitope within the adult population is thought to reflect differences in the rate of cartilage aggrecan catabolism/turnover in different individuals (1). The contention that an adult who has an abnormally high level of serum KS epitope has a higher rate of cartilage PG catabolism/turnover is based on several assumptions. First, one cannot rule out at this stage the possibility that such a difference in the serum level does not correspond, at least in part, to a difference in either the proportion of cartilagederived K S - b e a r i n g f r a g m e n t s t h a t reaches the blood or in the rate of clearance of the fragments from the blood. Second, it is possible that adults with elevated levels of serum KS epitope have KS chains that are much more antigenic. A recent observation (12) that serum levels of the KS epitope show a good correlation w i t h the s e r u m level of an epitope present on the core protein of aggrecan suggests the level of the KS epitope probably offers a good measure of total KS present. Studies are in progress to measure total KS (antigenic + nonantigenic) in serum to determine if, as hypothesized, the level of antigenic KS in serum provides an accurate measure of total KS present. The serum level of the KS epitope has been shown to provide important information about abnormalities or changes in the metabolism of KS-bearing PGs in a wide variety of conditions [see (1) for review]. Absence of the KS epitope in serum has proved useful in diagnosing type I macular corneal dystrophy, a blinding disease caused by an abnormality in sulfation of the keratan chain on corneal PGs (13-15). Interestingly, the KS chains in cartilage aggrecan are similarly unsulfated but thus far, this has not been found to result in any abnormalities in cartilagenous structures. M e a s u r e m e n t of the serum level of the KS epitope has also proved useful in studying abnormalities in the metabolism of polylactosamines that, as KS, contain long segments of the repeat disaccharide N-acetylglucosamine-galactose (16), and in assessing tumor grade in patients with chondrosarcomas. Levels of the s e r u m KS epitope undergo marked age-related changes during growth (17) and show a positive correlation with percentile height, supporting the contention that the rate of growth is a function of the activity of the chondrocytes (17). Children with constitutional delay of m a t u r i t y (below the fifth percentile for height CLINICAL BIOCHEMISTRY, VOLUME 25, JUNE 1992
but not growth hormone deficient) and children who are growth hormone deficient have abnormally low levels (18). Administration of growth hormone to growth hormone-deficient children resulted in a significant rise in the level of serum KS together with expected increases in annualized growth rate velocity and plasma levels of IGF-I (18). In normal adults, the ratio of cartilage mass/blood volume, the concentration of KS in cartilage, and the size of the KS chains show only m o d e r a t e changes with increasing age (1). The observation that the serum level of the KS epitope in adults is extremely constant with time is therefore not surprising. The level of serum KS epitope does not change significantly following t r e a t m e n t with the non-steroidal anti-inflammatory drugs piroxicam or naproxen, suggesting that these two NSAIDs have little if any effect on the rate of cartilage PG catabolism (19). In contrast, prednisone ( 4 0 - 6 0 mg daily), given orally for 3 days to patients attending an allergy clinic, causes a rapid decrease in the level of serum KS epitope that is already evident after 24 h and reaches a maximum on day 4 with a mean KS level at 61% of the pretreatment level (19). Importantly, the level of serum KS epitope was still only 69% of the pretreatment level 32 days after treatment was stopped. A single intraarticular injection of corticosteroids in the knee joint of patients with osteoarthritis (OA) caused a similar decrease in the level of the KS epitope in serum (20). Since KS present in serum is derived from all cartilagenous structures in the body, a slight change in the rate of PG catabolism in a single synovial joint is unlikely to give rise to a measurable increase in the level of KS epitope in serum. However, recent studies have shown that massive and rapid degradation of the cartilage PGs in a single synovial joint or intervertebral disc causes a large transient rise in the level of serum KS (7,9,21-24). Physicians injecting chymopapain into intervertebral discs for the treatment of a herniated disc have taken advantage of this approach to obtain evidence that the enzyme was injected in the correct place and that the PGs in the tissue have been degraded.
Keratan sulfate: A marker of predisposition to generalized osteoarthritis An elevation in the rate of degradation of PGs contributes to the progressive degeneration of articular cartilage in OA (1). This rise in the rate of PG catabolism appears to begin while the condition is clinically silent. In most cases, it is counterbalanced in its early stages by a corresponding increase in PG synthesis (25). Studies of levels of serum KS epitope suggest that this state of hypermetabolism develops during middle age in some individuals without any evidence of joint disease. We have postulated that it may predispose to OA (1,26,27). This contention is based on recent evidence that a proportion of the HA-binding regions (also called G1 regions or domains), which are generated during PG turnover by 177
THONARAND GLANT
the action of stromelysin on the core protein of the PGs, may remain attached to HA, thereby reducing the number of sites on HA molecules that are accessible to newly-synthesized PGs (1,28,29) as shown in Figure 2. These HA-binding regions that have been shown to accumulate in articular cartilage with age (28,29) contain no or few glycosaminoglycan chains and therefore contribute little to the functional properties of the tissue. With aging, this probably leads to depletion of glycosaminoglycans in some regions of the articular surface, predisposing collagen fibrils that are no longer supported by viscoelastic glycosaminoglycans to fracture, especially in maximally loaded regions. It is likely that in individuals who have high levels of the serum KS epitope and who are turning over their cartilage PGs at accelerated rates, the nonfunctional HA-binding regions accumulate at a faster rate, thereby predisposing tissues to failure and degeneration in regions that are maximally loaded (1). Recent studies have shown that serum levels of the KS epitope are higher in patients with polyarticular OA than in age-matched individuals without signs of joint disease (26,27,30,31) as shown in Figure 3. These results strongly support the widely held belief that aggrecan molecules in OA cartilages are being degraded at abnormally fast rates. It is worth noting that a proportion of patients with OA do not have elevated levels of the KS epitope while a small percentage of adults without any clinical signs of OA do (26). Consequently, a single m e a s u r e m e n t of the serum level of the KS epitope is not a useful marker of the articular cartilage destruction that is
/•A
Collagen
Figure 2 - - Hypothesized effect of age-related accumulation ofhyaluronate-binding region fragments in cartilage. Some of the HA-binding regions (G1 regions or domains) generated by the degradation of aggrecan by stromelysin remain attached to HA; these nonfunctional fragments occupy sites that are normally reserved for glycosaminoglycan-rich PG molecules. Note the difference in the hydrodynamic domains occupied by PGs tightly packed within the collagenous network (left hand side) and by PGs whose swelling capacity is no longer restricted (right hand side). With aging, the accumulation of HA-binding regions may lead to depletion of PGs within some regions of the matrix; as a consequence collagen fibrils that are not supported by the viscoelastic PGs are more likely to fracture (see the text for additional discussion). 178
OA
N
1000
80O
E i,i
Z
6OO
400
nil,
200
p
