By far the majority of calcium pyrophosphate dihydrate (CPPD) crystal deposition disease is an idiopathic/sporadic disease with onset in later life, but early-onset familial disease also occurs.
Familial CPPD crystal deposition disease is primarily an autosomal dominant disorder linked to a variety of mutations in ANKH , a chromosome 5 gene that encodes an inorganic pyrophosphate (PP i ) transporter.
Disordered PP i metabolism is central to the pathogenesis of CPPD crystal deposition disease, but dysregulated chondrocyte growth factor responsiveness and differentiation are also involved, as is the biology of aging.
The connective tissue matrix of fibrocartilaginous menisci, of articular hyaline cartilage, and of some ligaments and tendons are susceptible to pathologic calcification with CPPD crystals. Dehydration of the fibrocartilaginous menisci in aging may be a factor in particularly promoting PP i excess there.
Activation of the NLRP3 (cryopyrin) inflammasome, and consequent maturation and release of interleukin-1β, mediate CPPD crystal–induced inflammation.
This work was supported by the VA Research Service.
Genetics of Calcium Pyrophosphate Dihydrate Crystal Deposition Disease
The vast majority of calcium pyrophosphate dihydrate (CPPD) crystal deposition disease is an idiopathic/sporadic disorder related to aging, and some cases are related to joint trauma or certain metabolic diseases. However, some CPPD crystal deposition disease is familial, and typically these cases have early onset (defined as onset before age 55). Familial chondrocalcinosis is heterogeneous. For example, osteoarthritis (OA) with prominent CPPD and hydroxyapatite (HA) crystal deposits and cartilage and periarticular calcification were described in a kindred that has not yet been linked to a specific chromosomal locus. One chromosomal linkage of familial CPPD crystal deposition disease (with early-onset OA) is with 8q, and this was previously designated CCAL1. Chromosome 5p–linked chondrocalcinosis (termed CCAL2) is more common and is mediated by mutants of the gene ANKH (which encodes a transmembrane protein with PP i transport and other apparent functions discussed later); has been established in these studies. ANKH structure-function is reviewed in detail later and in a recent review.
A syndrome of spondyloepiphyseal dysplasia tarda, brachydactyly, precocious OA, and intraarticular calcifications with CPPD and/or HA crystals, as well as periarticular calcifications, was linked to mutation of the procollagen type II gene in indigenous natives of the Chiloe Island region of Chile. This population has a high prevalence of familial CPPD deposition disease. Families affected with diffuse idiopathic skeletal hyperostosis (DISH) and/or chondrocalcinosis have been identified in the Azores Islands, and this may reflect a shared pathogenesis, although the specific mechanism is not clear.
Pathogenesis of CPPD Crystal Deposition Disease
This discussion supplements and complements the review of this topic, with emphasis on pathology, by Kenneth P.H. Pritzker in Chapter 1 . The connective tissue matrix of fibrocartilaginous menisci, of articular hyaline cartilage, and of certain ligaments and tendons is particularly susceptible to calcification with CPPD (chemical formula Ca 2 P 2 O 7 •H 2 O, calcium:phosphate ratio 1.0). In addition, basic calcium phosphate (BCP) crystals can be deposited in articular cartilage, most commonly in OA. Unlike growth plate cartilage, joint cartilages are specialized to avoid developing matrix calcification, with lack of vascularity and abundant, intact proteoglycans among factors that limit access to phosphatases that liberate inorganic phosphate (P i ). However, altered matrix composition and hydration in aging and OA compromise these defense mechanisms.
CPPD crystals will precipitate where PP i concentration is highest. This is usually where matrix is most efficient at sequestering PP i and the furthest from pyrophosphatase activities. Phosphatases have broad substrate specificity, as exemplified by extracellular alkaline phosphatase, which has phosphatase, pyrophosphatase, and CPPD crystal dissolution activity, as discussed below. Classically, the locations for CPPD crystal deposition are in meniscal fibrocartilage and in middle zones of hyaline cartilage in synovial and symphyseal sites. However, CPPD sometimes forms in reparative fibrocartilage on the articular cartilage surface. Dehydration of the fibrocartilaginous menisci in aging may be a factor in particularly promoting PP i excess there.
It is rare for CPPD and BCP crystals to coexist at a single finite locus, since the physical formation conditions are mutually exclusive. However, this does occur at some sites, and, at these sites, the crystals likely have formed at different times; for example, HA might form when CPPD deposits dissolve. Usually, BCP in human cartilage forms near the surface of the articular cartilage. At times, driven by intraarticular steroids, BCP crystals form around the chondrons. Also, in advanced OA, BCP can form in the advancing calcification front of cartilage. Many, if not most cases, however, represent BCP debris from exposed or dislodged bone.
Major factors in the deposition of CPPD in joint cartilages are schematized in Figure 20-1 . Alterations in many of the same factors alternatively can promote BCP crystal deposition; for example, decreased and increased PP i can promote BCP crystal deposition, with increased tissue nonspecific alkaline phosphatase (TNAP) activity a cofactor. CPPD deposition reflects a breakdown of checks and balances imposed heterogeneously by genetics, inflammation, dysregulated chondrocyte growth factor responsiveness and differentiation, ATP and PP i transport and metabolism, and extracellular matrix environment, especially with aging. Increase in the concentrations of PP i , and the solubility product of PP i and ionic calcium clearly are factors in promoting CPPD crystal formation. However, concentration of magnesium, P i , iron, and cartilage extracellular matrix content (including high density of negative charges in intact proteoglycans) regulates the dynamics of CPPD crystal formation and helps to determine whether monoclinic and/or triclinic CPPD crystals are formed. In this context, monoclinic CPPD crystals appear more inflammatory than triclinic CPPD crystals.
The influence of the extracellular matrix on CPPD crystal formation (reviewed in part in Chapter 1 ) has typically been analyzed in model gel systems. Such studies, in particular using type I collagen as a variable in the gel system to promote CPPD deposition, have revealed stimulation of CPPD formation by ATP, osteopontin (a sialoprotein increased with chondrocyte hypertrophic differentiation and in OA cartilage), and addition of corticosteroids; in contrast, type II collagen and intact proteoglycans can suppress ATP-induced CPPD crystal formation in vitro.
Some experimental systems to analyze CPPD (and BCP) crystal deposition also have used matrix vesicles isolated from chondrocytes. Matrix vesicles are small, membrane-limited bodies released from chondrocytes (and other calcifying cells such as osteoblasts) that are enriched in constituents that regulate and can promote calcification. Matrix vesicles initially have intracellular [Mg 2+ ], [Ca 2+ ], and pro-calcifying protein molecules in their interior and TNAP on outside. As the vesicle “deflates,” Ca 2+ diffuses in and [Mg 2+ ] diffuses out. Toward the end of this process, the vesicle has an extracellular ion environment, but the remnant has protein and, in particular, lipids that bind calcium and promote calcification of the BCP crystal type, which may be an amorphous calcium phosphate before it becomes BCP crystals.
Matrix vesicles are clearly involved in cartilage growth plate calcification with BCP. However, it is not yet clear whether BCP crystal formation in articular cartilages is initiated more by matrix vesicles or by nucleation of crystals in the extracellular matrix. Moreover, areas in which CPPD crystals are deposited clearly include areas removed from collagen and matrix vesicles (and from pyrophosphatases) in cartilages affected by CPPD deposition disease (see Chapter 1 ). Matrix vesicles can provide phospholipids, proteinases, enzymes that regulate PP i metabolism, and other regulators of articular cartilage calcification. However, CPPD crystal deposition is likely to be initiated in the extracellular matrix and unlikely to be initiated within matrix vesicles, due to the very large size of CPPD (micron size, unlike submicroscopic BCP) crystals relative to matrix vesicles, and the substantial content of TNAP on the exterior of the vesicles, and magnesium and P i in the interior of matrix vesicles.
Loci of pericellular concentration of PP i may be needed to drive CPPD crystal formation at low micromolar PP i concentrations developing in cartilages with chondrocalcinosis Moreover, it is not clear what the effects on CPPD deposition are of apoptotic bodies, which have an inside-out vesicle orientation (i.e., where calcification-regulating mediators, such as the PP i -generating enzyme ENPP1, may be functionally misplaced on the surface of the structure).
There are unequivocal physical effects of calcium, P i , and PP i on crystal nucleation and propagation. These solutes also regulate mineralization by effects on gene expression, differentiation, and viability in chondrocytes, mediated partly by calcium-sensing receptors and sodium-dependent P i cotransport in chondrocytes. Excess PP i on chondrocytes also appears to be sensed (by unclear mechanisms) in chondrocytes, as evidenced by deleterious induction of matrix metalloproteinase-13 (MMP-13) expression, suppression of chondrogenesis, and promotion of apoptosis. These observations support the long-used clinical term “pyrophosphate arthropathy” as an umbrella term for the phenotype of chronic cartilage degeneration seen in CPPD crystal deposition disease.
Altered PP i Metabolism in CPPD Deposition Disease
PP i is a potent, physiologic inhibitor of the nucleation and propagation of BCP crystals, and this has been well illuminated in mouse models of pathologic soft tissue calcification linked with deficient PP i generation and transport. Chondrocytes and osteoblasts are unique in robustly producing extracellular PP i . Depending on the ambient levels of cartilage ATP and PP i and the level of activity of P i -generating ATPases and TNAP, and the PP i -degrading effects of TNAP, formation of CPPD and HA crystals may be promoted in the same cartilages, an event that can occur in OA. However, the physical-chemical conditions favoring CPPD and BCP crystal formation are largely mutually exclusive. Where CPPD and BCP are found in adjacent domains, such as occasionally seen in OA, the crystal types formed at different times; in some cases secondary to partial dissolution of preexisting CPPD crystals.
Role of ENPP1 in PP i Metabolism in Chondrocalcinosis
Sporadic/idopathic CPPD crystal deposition disease associated with aging is consistently linked with an excess chondrocyte PP i -generating nucleotide pyrophosphatase phosphodiesterase (NPP) activity and increased PP i generation by chondrocytes. The NPP family isoenzymes ENPP1 (formerly known as NPP1 and plasma cell membrane glycoprotein-1 [PC-1]) and ENPP3 (formerly known as B10) actively generate PP i via hydrolysis of nucleoside triphosphates, principally ATP. Notably, some of the ATP used by chondrocytes to generate extracellular PP i is extracellular, and some is generated by the mitochondria.
ENPP1 plays a core role in driving extracellular PP i in chondrocytes (see Figure 20-1 ), and in some other cell types. Increased ENPP1 also is associated with apoptosis in vitro and in degenerative human cartilages. ENPP1 deficiency states in vivo and in vitro are linked with up to a 50% decrease in plasma and extracellular PP i . Contrastingly, in sporadic/idiopathic chondrocalcinosis of aging, cartilage NPP activity and PP i levels have been reported to average approximately double those of normal subjects. This PP i concentration is insufficient to cause CPPD deposition. Therefore, sequestration of PP i in pericellullar matrix is thought to be necessary to raise PP i levels sufficiently to achieve CPPD crystal deposition).
ANKH in the Molecular Genetics and Pathogenesis of CPPD Crystal Deposition Disease
ANKH encodes a multiple-pass transmembrane protein that functions in PP i channelling ( Figure 20-2 ) and also appears to directly or indirectly promote ATP release. ANKH also appears to regulate P i metabolism and uptake of P i by the type III sodium-dependent P i cotransporter Pit-1. ANKH also promotes bidirectional movement of PP i at the plasma membrane, but the gradient for ANKH-stimulated PP i movement in chondrocytes is from cytosol to the extracellular space. Chondrocytes produce abundant PP i in the cell both by ENPP1 and as a byproduct of matrix biosynthesis and other biochemical (e.g., intramitochondrial) reactions, and ANKH transport of PP i that is generated intracellularly by ENPP1, is likely the fundamental means by which chondrocytes regulate extracellular PP i concentrations. Molecular models of ANKH have limitations, but the PP i channelling function of ANKH may be via 10 or 12 membrane-spanning domains in the molecule that have an alternating inside-out orientation and provide a central channel for movement of PP i (see Figure 20-2 ).
ANKH is unequivocally involved in pathogeneses of both familial and idiopathic/sporadic chondrocalcinosis. Importantly, ANKH expression is regulated, and ANKH is increased in OA and chondrocalcinotic cartilages, and increased chondrocyte ANKH expression may drive secondary chondrocalcinosis in OA. Hypoxia inhibits ANKH expression, via the transcription factor hypoxia inducible actor-1α; it is possible that that increased permeability to oxygen via fissures and fibrillation in OA cartilage promotes increased ANKH expression. Conjoint effects of ANKH and signaling by extracellular P i promote chondrocyte maturation to the hypertrophic differentiation state that promotes calcification.
Figure 20-1 summarized a paradigm in which alterations in chondrocyte expression of both ANKH and ENPP1 drive PP i supersaturation in cartilage in idiopathic/sporadic and OA-associated CPPD crystal deposition arthropathy. Alternatively, mutations at different locations in ANKH can affect postnatal skeletal development, inducing autosomal dominant chondrocalcinosis ( Table 20-1 ) or a variety of other phenotypes, such as murine progressive ankylosis ( ank/ank mouse). In humans, craniometaphyseal dysplasia (CMD) is another phenotype of multiple mutations of ANKH (nine to date), most of which are in predicted cytosolic regions of central exons in the molecule. This phenotype has been linked with decreased transport of PP i within bone, which modulates bone resorption and remodeling likely directly and indirectly on the function of osteoclasts, as elucidated in Ank -deficient mice, and by a knock-in mouse model homozygous for the phenylalanine 377 deletion.
|Location (cDNA Position ∗ )||Nucleotide Variation (Amino Acid Change)||Role in Sporadic CPPD Deposition||Change in ANKH Expression|
|5′-UTR (–11 bp)||C>T (+4 amino acids † )||No role||Unknown|
|5′-UTR ‡ (–4 bp)||G>A (NA)||Yes||↑|
|Exon 1 (+13 bp)||C>A (Pro→Thr)||Unknown||↔|
|Exon 1 (+14 bp)||C>T (Pro→Leu)||No role||↑ or ↔|
|Exon 2 (+143 bp)||T>C (Met→Thr)||No role||↔|
|Exon 12 § (+490 bp)||GAGdel (Glu deletion)||No role||↑|
Recently, a consanguineous family was defined in which homozygous ANK missense mutation L244S was detected in all those with a novel phenotype of mental retardation, in addition to deafness (a finding in some CMD kindreds), and joint ankylosis, and with skeletal features such as painful small joint soft tissue calcifications, progressive spondyloarthropathy, osteopenia, and mild hypophosphatemia. In this kindred, the mutant ANKH was transcribed and synthesized and moved into the plasma membrane. However, fibrosis and mineralization of the articular soft tissues developed in homozygotes. Significantly, heterozygous carriers of this L244S mutation developed mild osteoarthritis, without altered serum phosphate or the bone changes of the homozygotes. Such findings suggest a fundamental homeostatic role of PP i metabolism in maintenance of articular cartilage, likely through maintenance of physiologic articular chondrocyte differentiation, but other mechanisms may be at play, since ANKH affects bone geometry at the joint, and affects differentiation of bone marrow cells (of the erythroid lineage).
Chondrocalcinosis associated with ANKH mutations demonstrates clinical and mechanistic heterogeneity (see Table 20-1 ). This is consistent with the concept of differing functional effects of ANKH mediated by specific regions of the molecule that cause either chondrocalcinosis (largely the N-terminal ANKH domain) or CMD (certain cytosolic regions). Most N-terminal ANKH mutations identified, to date, in association with familial chondrocalcinosis (see Table 20-1 ) appear to elevate PP i transport, but some of ANKH mutations have differing effects on chondrocyte differentiation. Moreover, the M48T ANKH mutant, originally characterized in a French kindred, appears to have unique effects; first, it is linked with increased intracellular PP i , and, second, it interrupts ANKH interaction of with the sodium-dependent P i cotransporter Pit-1. The significance of such an effect may be because elevated P i increases both ANKH and Pit-1 expression; in addition, ANKH and Pit-1 colocalize in the plasma membrane in chondroctyres. Moreover, P i and P i uptake modulate chondrocyte differentiation and promote chondrocyte hypertrophy. The findings of a single case with sporadic CPPD deposition disease also are instructive, via linkage with the ANKH mutation ΔE590 ; ANKHΔE590 appears to indirectly suppress PP i catabolism by association with impairment of TNAP expression. This suggests an alternative mechanism of disrupting PP i metabolism by mutant ANKH.
Collectively, the capacity of ANKH to promote chronic, low-grade chondrocyte “PP i leakiness” clearly can promote CPPD crystal deposition via extracellular matrix supersaturation with PP i . In this context, homozygosity for a single nucleotide substitution (–4 G to A) in the ANKH 5′-untranslated region (see Figure 20-2 ) that promotes increased ANKH mRNA expression was present in about 4% of British subjects identified as having idiopathic/sporadic chondrocalcinosis of aging. These findings indicate that a small but significant subset of CPPD deposition disease in late middle to later life likely has a slow onset familial component.
TNAP and CPPD Crystal Dissolution and CPPD Deposition Disease, Including in Hypophosphatasia
TNAP is a major physiologic antagonist of ENPP1-mediated elevation of extracellular PP i . Conversely, physiologic ENPP1-induced PP i generation antagonizes the essential pro-mineralizing effects of TNAP mediated by P i generation, and presumed PP i excess in the joint space promotes a scenario to drive chondrocalcinosis in young adults in hypophosphatasia. The rate-limiting factor for PP i concentration in extracellular fluid is TNAP activity. TNAP hydrolyzes PP i to more soluble P i . TNAP chondrocyte alkaline phosphatase has been shown to dissolve the normally very insoluble CPPD crystals at physiologic pH. To do so, TNAP must be attached to the surface of the CPPD crystal, indicating that that TNAP must act on soluble PP i that is in equilibrium with solid state PP i on the crystal surface. This provides a formidable challenge to crystal dissolution strategies that are less efficient than TNAP. Further, TNAP inhibition, by decreasing PP i hydrolysis, inhibits CPPD crystal dissolution. Endogenous small molecules such as cysteine and mercaptopyruvate can inhibit TNAP and CPPD crystal dissolution in vitro. Interestingly, relative elevation of cysteine and mercaptopyruvate substances may occur in vivo in hypoxic states.
Hypophosphatasia is due to deficient activity of TNAP, consequently with effects including limitation of hydrolysis PP i to generate P i . Generalized PP i excess in hypophosphatasia is evidenced by increased PP i excretion in urine. Enpp1 knockout mice and mice homozygous for the ENPP1 truncation mutant ttw demonstrate marked articular cartilage calcification with HA and OA, as well as ankylosing spinal ligament hyperostosis and synovial joint ossific fusion; extracellular PP i levels and mineralization disturbances in soft tissues (but not long bones) of Enpp1 knockout and TNAP-deficient mice are mutually corrected by crossbreeding.
Imbalance of Chondrocyte Growth Factor Responses on PP i Metabolism in CPPD Deposition Disease
Insulin-like growth factor (IGF)-I and transforming growth actor (TGFβ) are major chondrocyte anabolic growth factors, but they have antagonistic effects on chondrocyte PPi metabolism. Specifically, TGFβ stimulates ENPP1 expression and ENPP1 movement to the plasma membrane, and ATP release by chondrocytes, which stimulate increased extracellular PP i . TGFβ also stimulates ANKH expression, mediated in part by calcium entry into the cell through voltage-operated channels, and associated calcium-mediated signal transduction. The capacity of TGFβ to raise chondrocyte PP i rises with aging in humans, and TGFβ-stimulated NPP activity also does so. Growth-promoting effects in articular chondrocytes of TGFβ, on the other hand, decrease with aging. Moreover, TGFβ suppresses TNAP in chondrocytes, an effect mediated by P i .
In contrast to TGFβ, IGF-I physiologically suppresses extracellular PP i (as well as ATP release), , in chondrocytes (see Figure 20-1 ). Importantly, chondrocyte IGF-I resistance develops in aging and osteoarthritic cartilages (see Figure 20-1 ). Significantly, IGF-I induces expression of cartilage intermediate layer protein (CILP) (see Figure 20-1 ), and this large cartilage interterritorial matrix protein rises in expression in cartilage in aging and OA. CILP is most prevalent in the middle zone of articular cartilage, and it is this cartilage zone where CPPD crystal deposition is most abundant. The CILP-1 isoform, but not CILP-2, promotes increased extracellular PP i in chondrocytes in an indirect manner by inhibiting IGF-I signaling at the receptor level.
It should be noted that the inflammatory cytokine IL-1β, whose expression rises in OA, and which promotes OA progression, also suppresses both ENPP1 expression and extracellular PP i in chondrocytes. IL-1β also inhibits the effects of TGFβ on PP i .
Inflammation, Hypertrophic Chondrocyte Differentiation, and Transglutaminase 2 in Joint Cartilage Calcification
Changes in chondrocyte differentiation and viability appear to promote joint cartilage calcification, not simply OA. Multifocal development of chondrocyte maturation to hypertrophy is found in OA, and also chondrocyte hypertrophy is seen adjacent to CPPD crystal deposits. Chapter 1 presents several arguments against chondrocyte hypertrophy being central in CPPD crystal deposition disease, including the fact that TNAP, which promotes PP i hydrolysis and dissolution of CPPD crystals, is actually upregulated in hypertrophic chondrocytes. On the other hand, articular chondrocyte hypertrophy ( Figure 20-3 ) is associated with ANKH expression as a marker (at least in growth plate cartilage), increased PP i elaboration, and increased release of matrix vesicles, and certain other calcification-promoting changes such as increased expression of certain transglutaminases (transglutaminase 2 [TG2] and FXIIIA) and of osteopontin (which promote CPPD crystal formation) and loss of physiologic extracellular matrix composition that suppresses calcification. Significantly, changes in TGFβ signal transduction in aging (and OA) likely are involved in promoting chondrocyte maturation to hypertrophy. One can strongly argue that the key studies linking chondrocyte hypertrophy to CPPD deposition have not yet been done, since the classic hypertrophy marker type X collagen has not been analyzed in situ. Moreover, it is possible that CPPD crystals deposited in chondrons of hypertrophic chondrocytes eventually kill such chondrocytes, making the chondrocyte characterization problematic.