(Received for publication, June 11, 1996, and in revised form, October 11, 1996)
From the Departments of Biochemistry,
Operative Dentistry, and
¶ Prosthodontics, Resting chondrocytes do not usually undergo
differentiation to the hypertrophic stage and calcification. However,
incubating these cells with concanavalin A resulted in 10-100-fold
increases in alkaline phosphatase activity, binding of
1,25(OH)2-vitamin D3, type X collagen
synthesis, 45Ca incorporation into insoluble material, and
calcium content. On the other hand, other lectins tested (including
wheat germ agglutinin, lentil lectin, pea lectin, phytohemagglutinin-L,
and phytohemagglutinin-E) marginally affected alkaline phosphatase activity, although they activate lymphocytes. Methylmannoside reversed
the effect of concanavalin A on alkaline phosphatase within 48 h.
Concanavalin A did not increase alkaline phosphatase activity in
articular chondrocyte cultures. In resting chondrocyte cultures,
succinyl concanavalin A was as potent as concanavalin A in increasing
alkaline phosphatase activity, the incorporation of
[35S]sulfate,
D-[3H]glucosamine, and
[3H]serine into proteoglycans, and the incorporation of
[3H]serine into protein, although concanavalin A, but not
succinyl concanavalin A, induced a rapid change in the shape of the
cells from flat to spherical. These findings suggest that concanavalin A induces a switch from the resting, to the growth-plate stage, and
that this action of concanavalin A is not secondary to changes in the
cytoskeleton. Chondrocytes exposed to concanavalin A may be useful as a
novel model of endochondral bone formation.
Many plant lectins, including concanavalin A
(ConA)1 and phytohemagglutinin (PHA), with
various sugar-binding properties induce blastoid transformation of
lymphocytes and have been extensively applied in studies of
proliferation and differentiation (1). However, the mechanism by which
plant lectins activate lymphocytes is unknown because of the lack of
lectin specificity. ConA and wheat germ agglutinin (WGA) also have
insulin-like activity in adipocytes (2-4). Our group has shown that in
chondrocyte cultures, ConA, but not other lectins, stimulates the
synthesis of cartilage-matrix proteoglycan (aggrecan) by chondrocytes,
and that this effect is greater than that of growth factors and
hormones (5). Furthermore, ConA induces cartilage in amphibian early
gastrula ectoderm (6). These findings suggest that cell-surface
glycoproteins with N-linked sugar chains specific for ConA
play crucial roles in chondrocyte differentiation.
In growth plates and bone fracture callus, chondrocytes undergo a
sequence of cell changes including proliferation, cessation of cell
division, matrix synthesis, maturation (hypertrophy), and
calcification. The hypertrophic chondrocytes produce alkaline phosphatase (ALPase) (7), type X collagen (8), and
1,25-dihydroxyvitamin D3 (1,25(OH)2-vitamin D)
receptor (9), and they induce matrix calcification (10-13), which is
essential for elongation and repair of the skeleton.
Chondrocytes in the resting zone provide a reserve chondrocyte pool in
the growth plate, and articular chondrocytes contribute to resilience
of the tissue. Neither of these chondrocyte subtypes usually undergoes
endochondral ossification. Cells from the resting zone undergo
hypertrophy at slower rates than growth-plate chondrocytes when
maintained at high density in the presence of 10% serum (9). On the
other hand, articular chondrocytes rarely undergo hypertrophy even at
high density in the presence of 10% serum (14), although they become
hypertrophic in arthritic joints (15). The mechanism by which
hypertrophy is suppressed in permanent cartilages is not known.
However, lectins may alter the differentiation program in permanent
chondrocytes by cross-linking cell-surface glycoproteins, because some
surface glycoproteins serve as growth factor receptors, extracellular
matrix receptors, and modulators of signal transduction. To test this
hypothesis, we examined the effect of 11 lectins with different
sugar-binding properties on the expression of maturation-related phenotypes by resting and articular chondrocytes. The results showed
that among the tested lectins, ConA alone induces ALPase activity, type
X collagen synthesis, vitamin D receptor synthesis, and matrix
calcification in resting, but not in articular chondrocyte cultures.
Lectins, Chondrocytes were isolated from the
growth plate and resting cartilage of ribs of 4-week-old male Japanese
White rabbits, as described (14, 16). Unlike growth plates of long
bones, the rib growth plate does not contain resting chondrocytes. The resting cell population is not contaminated by growth-plate
chondrocytes, although the growth-plate cell population may be
contaminated by a few (<10%) resting chondrocytes (9). The majority
of "growth-plate chondrocytes" start to proliferate after seeding
and recapitulate the differentiation program after cell division stops,
even though they originate from different (proliferating,
matrix-forming, and hypertrophic) zones (10). Articular chondrocytes
were isolated from the surface (0.2 mm) of the articular cartilage from
the femur at knee joints of the same rabbits, since the deep zone contains maturing chondrocytes (14).
Cells were seeded at 104 or 3 × 103
cells/6-mm plastic microwell, and grown in 0.1 ml of Eagle's medium,
Chondrocytes were seeded
at 104 or 3 × 103 cells/6-mm well, and
maintained in Medium A. On day 5 or 7, they were transferred to
Cells were disrupted in a glass homogenizer in 1 ml of 0.9% NaCl,
0.2% Triton X-100 at 0-4 °C and centrifuged for 15 min at 12,000 × g. ALPase activity in the supernatant, which
accounted for 95% of the total, was assayed in 0.5 M
Tris-HCl buffer (pH 9.5) supplemented with 0.5 mM
p-nitrophenyl phosphate and 0.5 mM
MgCl2 (17).
The DNA
content was determined by means of a fluorometric procedure (18). Total
protein was determined by dye binding (19). Hexuronic acid was
determined as described by Bitter and Muir (20).
After incubation with or without ConA for
24 h on day 5, resting chondrocytes in 6-mm wells were incubated
for 10 days in 0.1 ml of Medium A in the presence of 100 µg/ml
ascorbic acid. Cell layers were washed five times with saline and twice
with distilled water at 4 °C. They were homogenized at 4 °C with
a solution containing 0.9% NaCl, 0.2% Triton X, and protease
inhibitors (1 mM phenylmethylsulfonyl fluoride, 25 mM EDTA, and 10 mM
N-ethylmaleimide). The suspension was centrifuged at
3,000 × g for 15 min at 4 °C, and the pellet was
incubated with pepsin at 0.4 mg/ml in 0.5 M acetic acid for
48 h at 4 °C. The pepsin-resistant material was solubilized
with Laemmli buffer, and proteins were separated by SDS-polyacrylamide
gel electrophoresis (SDS-PAGE) under nonreducing conditions, then
stained with silver nitrate.
Chondrocytes in
100-mm dishes were homogenized at 0 °C in 3 ml of buffer A (10 mM Tris-HCl, pH 7.4, 2 mM EDTA, 5 mM dithiothreitol) containing 0.3 M KCl. The
homogenate was centrifuged at 500 × g for 10 min at
4 °C, after which the supernatant was centrifuged at 4 °C at
226,000 × g for 60 min. Portions (0.5 mg of protein in
0.4 ml of buffer A) of cytosols were transferred to polypropylene tubes
containing
1,25-dihydroxy[26,27-methyl-3H]cholecalciferol
(20,000 dpm/26 pg/32 fmol). After an incubation at 20 °C for
1.5 h, 50 µl of buffer A containing 5% charcoal and 0.5%
dextran was added to the solution, and the suspension was agitated
every 5 min with a Vortex mixer. After 30 min at 4 °C, unbound
vitamin D metabolites were removed by centrifugation at 500 × g for 20 min to precipitate the charcoal, and the
radioactivity in the vitamin D receptor complex in the supernatant was
determined. Nonspecific binding was determined by incubating the cells
with a 200-fold excess of unlabeled 1,25(OH)2-vitamin
D.
Chondrocytes in 35-mm dishes were exposed to
45CaCl2 (20 µCi/culture) in 2 ml of
To determine the calcium content, the cell layers were washed five
times with saline, then heated at 800 °C for 10 h. The ash was
dissolved in 4 M HCl containing 1% lanthanum chloride, and
the calcium content was determined by atomic absorption spectrometry (model AA-640; Shimadzu, Kyoto, Japan). The cell-matrix layers were
stained with alizarin red-S as described (22).
Resting chondrocytes were seeded at a density of
106 cells/100-mm dish, and maintained in Medium A. On days
7, 12, and 17, the cells were exposed or not to 10 µg/ml ConA in the
absence of serum for 6 h. They were incubated in 10 ml of Medium A
with 10% serum from days 7 to 12 and from days 15 to 17, and in 10 ml
of Alternatively, proteins in the media were transferred
electrophoretically to a polyvinylidene difluoride membrane (23), and
stained with Coomassie Brilliant Blue. The N-terminal amino acids of a
15-kDa protein were determined using an automatic protein sequencer
(Applied Biosystems model 476A).
When resting chondrocyte cultures in 6-mm wells became
confluent, they were incubated in 0.1 ml of Dulbecco's modified
Eagle's medium containing 0.3% fetal bovine serum, 32 units/ml
penicillin, and 40 µg/ml streptomycin (Medium B). They were then
incubated in 0.1 ml of fresh Medium B in the presence of lectins and/or methyl- Plastic tissue culture dishes
(diameter, 35 mm) were incubated with 1 ml of 0.1 M
NaHCO3 containing type II collagen (10 µg/ml) or
fibronectin (10 µg/ml) at 37 °C for 3 h, then washed three times with saline.
Chondrocytes in primary cultures were
harvested with phosphate-buffered saline containing 0.01% trypsin and
0.01% EDTA, then seeded at a density of 2 × 105
cells/35-mm plastic, fibronectin- or collagen-coated well. They were
incubated with lectins at 37 °C in 2 ml of Dulbecco's modified Eagle's medium for 3 h before being photographed.
Resting chondrocytes in high or low density
cultures were exposed for 24 h to ConA on day 5 or 7, respectively, after the cultures reached confluence. Thereafter, they
were maintained in the presence of 10% serum for 10 days. ConA (10 µg/ml) was added to the cultures only once on day 5 or 7, because
ConA bound to the cells and extracellular matrix macromolecules within
24 h.2 The incubation with ConA caused a
>12-fold increase in ALPase activity in high and low density cultures
on days 15 and 17, respectively (Fig. 1).
Among growth factors and hormones, BMP, thyroid hormone, calcitonin,
and retinoic acid increase ALPase and/or type X collagen synthesis by
chondrocytes (25-30). These compounds were added to ConA-free cultures
every other day, and their effect on ALPase was compared with that of
ConA. Incubating resting chondrocytes with BMP-2 from day 5 or 7 for 10 days consistently increased ALPase activity in high, but not low
density cultures (Fig. 1; data not shown). T3 also
increased ALPase activity in high and low density cultures, whereas
IGF-I induced a low level of ALPase in low, but not high density
cultures. In all studies, the effect of ConA on ALPase was greater than
that of BMP-2, IGF-I, or T3 (Fig. 1; data not shown). The
effect of retinoic acid and calcitonin on ALPase was much lower than
that of ConA.2
PTH and TGF- In articular chondrocyte cultures, ConA (10 µg/ml) had little effect
on ALPase activity, although BMP and T3 markedly increased ALPase activity (Fig. 2). In MC3T3-E1 osteoblast
cultures, ConA (10 µg/ml) decreased ALPase activity at 48-72 h,
whereas BMP-2 increased this activity 2-20-fold at 24-72
h.2 These findings suggested that ConA and BMP-2 modulate
ALPase synthesis by different mechanisms. The effect of ConA on ALPase in resting chondrocyte cultures was detectable at 1 µg/ml and maximal
at 5-10 µg/ml (Fig. 3A). ConA at 10 µg/ml increased ALPase activity in resting chondrocytes 50-fold (Fig.
3A), whereas it increased ALPase activity in growth-plate
chondrocytes only 1.3-fold because of a high basal level (Fig.
3B).
The lectin specificity of the ConA effect on ALPase induction was
examined by exposing resting chondrocytes to the various lectins that
activate lymphocytes. A divalent derivative of ConA (succinyl ConA) was
as potent as the native tetravalent ConA in stimulating ALPase activity
(Fig. 3, A and B). Lentil lectin, garden pea
lectin (pea), and Phaseolus vulgaris agglutinin-E (PHA-E) induced very low levels of ALPase activity (Fig. 3A).
Maclura promifera, lima bean lectin, WGA, Scotch broom
lectin, Ulex europeaus agglutinin (I + II), lotus lectin,
and P. vulgaris agglutinin-L (PHA-L) did not affect ALPase
activity (Fig. 3A).
Fig. 4 shows the time courses
of the increases in DNA, uronic acid, and ALPase contents in cultures
exposed or not to ConA. Resting chondrocytes seeded at 104
cells/6-mm well proliferated for two generations, yielding a confluent
monolayer on day 5, then underwent one or two rounds of cell division
to form a multi-cell layer by day 10. The DNA content reached a plateau
on day 16 (Fig. 4A). Exposing chondrocytes to ConA for
24 h on day 5 suppressed DNA synthesis on days 6, 8, and 10 (Fig.
4A). The cells exposed to ConA started to proliferate from
day 14, then the DNA content reached the level in cultures not exposed
to ConA by day 24 (Fig. 4A).
Most of the macromolecules containing uronic acid synthesized by
chondrocytes exposed or not to ConA are large, chondroitin-sulfate proteoglycans (aggrecan) (5). ConA selectively stimulates aggrecan synthesis without increasing the synthesis of hyaluronic acid and small
proteoglycans (5). In ConA-free cultures, the uronic acid level
increased from day 10, and reached a plateau on day 18 (Fig.
4B). In cultures exposed to ConA, the level promptly increased from day 5-6, and reached a maximum on day 8. Thereafter, it
decreased gradually, and returned to the level of ConA-free cultures on
day 20. On day 8, the uronic acid level in the exposed cultures was
4-fold higher than that in ConA-free cultures.
In ConA-free chondrocyte cultures, the level of ALPase activity was
very low throughout a 24-day culture period (Fig. 4C). However, upon exposure to ConA, it markedly increased from day 8, and
reached a maximum on day 14 (Fig. 4C). The high ALPase level
was sustained at least until day 24. Some cultures were exposed twice
to ConA on days 5 and 14. The second exposure on day 14 increased
ALPase activity 1.4-fold within 48 h,2 even though the
uronic acid content reached a plateau on day 8. These findings
suggested that the ConA induction of ALPase is not secondary to the
accumulation of proteoglycan.
Incubation of chondrocytes with ConA for 6 h had the same effects
on uronic acid and ALPase levels as that for 24 h,2
perhaps because the lectin binds to the cell-matrix layer within 6 h.
The addition of MeMan on day 5 for 24 h suppressed
the effect of ConA on ALPase activity dose-dependently with
an ED50 of 10 mM (Fig.
5A). MeMan alone had little effect on ALPase
activity. Furthermore, the addition of MeMan on day 12 reversed the
effect of ConA on ALPase activity within 48 h, suggesting that
prolonged action of the lectin is required for ALPase synthesis after
the accumulation of proteoglycan. In contrast, the addition of MeMan on
day 16 had little effect on ALPase activity (Fig. 5B),
perhaps because the ConA-activated cells had already differentiated to the irreversible hypertrophic stage.
Because ConA was
removed gradually from the cell-matrix layer (Fig. 4), chondrocytes
were incubated three times with ConA on days 7, 12, and 17. Resting
chondrocytes exposed to ConA (Fig. 6B), but
not the untreated cells (Fig. 6A), became hypertrophic by
day 20. On the other hand, articular chondrocytes became fibroblastic in the absence of ConA by day 20 (Fig. 6C), because of the
low seeding cell density and the low serum concentration used in this study. Nevertheless, the articular chondrocytes exposed to ConA maintained a spherical phenotype, although they did not become hypertrophic (Fig. 6D). When articular chondrocytes were
seeded at high density (104 cells/6-mm plastic microwell)
and maintained in the presence of 10% serum, they assumed a spherical
configuration in the absence of ConA. However, they did not become
hypertrophic, even in the presence of 10 µg/ml ConA.2 The
resting chondrocytes exposed to ConA (B, 1279 ± 591 µm2) were 2.5- and 3-fold larger than those that were not
(A, 506 ± 176 µm2) and articular
chondrocytes (D, 385 ± 85 µm2),
respectively. The diameter of resting chondrocytes exposed to ConA
(B, 39.0 ± 10.2 µm) was similar to that (30-40
µm) of hypertrophic chondrocytes in the growth plate in
vivo.2
Resting chondrocytes maintained without ConA
synthesized type II collagen but not type X collagen, a marker of
hypertrophic chondrocytes. However, ConA-exposed chondrocytes
synthesized both type II and X collagens on day 15 (Fig.
7).
Next we examined whether ConA
stimulated the expression of the other maturation-related phenotypes in
resting chondrocyte cultures. Resting chondrocytes maintained without
ConA showed only a low level of the 1,25(OH)2-vitamin D
binding capacity on days 14 and 20 (Fig. 8). However,
exposure to ConA on day 7 resulted in 30- and 5-fold increases in the
binding of 1,25(OH)2-vitamin D to chondrocytes on days 14 and 20, respectively.
No calcification was evident in ConA-free
cultures of resting chondrocytes. However, exposure to ConA for 24 h on day 7 resulted in 10-20-fold increases in 45Ca
incorporation into insoluble material and the calcium content on day 35 (Fig. 9, A and B). The cell-matrix
layer of chondrocytes exposed to ConA was intensely stained with
alizarin red (Fig. 9C).
ConA had little effect on the
SDS-PAGE profiles of secreted proteins in the media conditioned by
resting chondrocytes, except that it markedly increased the level of
15-kDa protein (Fig. 10). N-terminal analysis of the
15-kDa protein revealed a 28-amino acid sequence of
LLGGLEDVDAQEKDVQRALGFAESSYNK, which was homologous to that of rat,
bovine, mouse, and human cystatin C (61, 57, 57, and 50% identity,
respectively) and human cystatins D, SA, S, and SN (54, 54, 58, and
54% identity, respectively). The molecular mass of the 15-kDa protein
was also similar to that of cystatin. It is unlikely however, that the
cystatin-like protein mediates the action of ConA on chondrocytes,
because this protein partially purified from the conditioned medium had
little effect on ALPase activity.3 The
media conditioned by ConA-treated chondrocytes did not increase ALPase
activity in resting chondrocyte cultures,3 suggesting that
the ConA stimulation of ALPase is not mediated by secreted
factors.
Native, tetravalent ConA stimulates lymphocyte
proliferation at lower concentrations than the divalent ConA
derivative, succinyl ConA (33). However, in chondrocyte cultures,
succinyl ConA was as potent as ConA in stimulating the incorporation of
[35S]sulfate (Fig. 11A),
[3H]glucosamine (Fig. 11B), and
[3H]serine (Fig. 11C) into proteoglycans and
the incorporation of [3H]serine into total protein (Fig.
11D). On the other hand, succinyl ConA suppressed DNA
synthesis to a lesser extent than ConA (Fig. 11E). The
inhibition of DNA synthesis by these lectins was
dose-dependently reversed by MeMan (Fig. 11E).
The effect of ConA on proteoglycan synthesis was also reversed by MeMan
(5).
Succinyl ConA and ConA differed with respect to their actions on
chondrocyte morphology. Resting chondrocytes assumed a fibroblastic configuration in monolayer cultures at a low serum concentration (0.3%). ConA at 10 µg/ml remarkably changed the cells from
fibroblastic to spherical within 24 h (5), whereas succinyl ConA
induced a moderate change from fibroblastic to polygonal at 48-96
h,2 perhaps because of the increase in the proteoglycan
content. ConA, but not succinyl ConA, may directly alter the
architecture of the cytoskeleton.
To test this notion, chondrocytes were seeded on fibronectin-coated
dishes and incubated for 3 h in the absence or presence of ConA.
ConA-free cells fully spread on the substrate within 3 h (Fig.
12A). ConA inhibited the formation of long
and/or thin cellular projections at 1-3 µg/ml (Fig. 12, B
and C), and abolished cell spreading at 10 µg/ml (Fig.
12D). In contrast, succinyl ConA had little effect on cell
spreading at 30 µg/ml (Fig. 12E). Similar results were
obtained with chondrocytes seeded on plastic tissue culture dishes and
type II collagen-coated dishes.2
Exposure to ConA markedly increased ALPase activity and the size
of resting chondrocytes. ALPase synthesis was stimulated more by ConA
than by hormones and growth factors that supposedly enhance ALPase
synthesis by chondrocytes. The specificity of the ConA action on ALPase
was indicated by the following observations. (i) ConA did not stimulate
ALPase synthesis and hypertrophy in articular chondrocyte cultures. The
marginal effect of ConA on ALPase activity in growth-plate chondrocyte
cultures could have been due to contamination of the growth-plate cell
population by resting chondrocytes. (ii) Lentil and garden pea lectins
had little effect on ALPase activity, although their
monosaccharide-binding properties (mannose/glucose) are similar to
those of ConA. The crystal structure of pea lectin is also similar to
that of ConA (34). However, the oligosaccharide-binding properties of
lentil and garden pea lectins differ from those of ConA (1). (iii) Of
the lectins examined, only ConA stimulated chondrocyte hypertrophy and
calcification. This is important, because many lectins with different
sugar-binding properties activate lymphocytes. (iv) ALPase induction by
ConA was suppressed by methylmannoside.
Our findings indicated that ConA does not randomly activate gene
expression by chondrocytes, but that it initiates the maturation program in resting chondrocytes. Exposing resting chondrocytes to ConA
for 6-24 h suppressed cell division, then increased the proteoglycan
content, the cell size, ALPase activity, the vitamin D receptor level,
the incorporation of 45Ca into insoluble material, and the
calcium content. These cell-matrix changes proceeded in the same order
as that seen in growth plates and bone fracture callus in
vivo. The cell size as well as the ALPase and the vitamin D
receptor levels in the ConA-exposed chondrocyte cultures were
comparable to those in the growth plate in vivo (14).
Chondrocytes in rib resting cartilage start to mature only near the
growth plate in vivo, although cells far from the growth plate can also become hypertrophic when cultured at high density in the
absence of a tissue culture surface (9). Changes in cell-matrix
interactions and/or growth factor levels may stimulate the maturation
of resting chondrocytes in vivo. We postulated that specific
lectins induce the maturation of "permanent chondrocytes" in
vitro by cross-linking or activating growth factor receptors, extracellular matrix receptors, and/or modulators of signal
transduction, because some of the cell-surface proteins are uniquely
glycosylated. The finding that ConA, but not other lectins, induces
maturation of resting chondrocytes suggests that ConA-binding surface
glycoprotein(s) plays a crucial role in switching from the immature, to
the mature stage.
Why only resting, but not articular, chondrocytes respond to ConA is
unknown. Resting cartilage provides the reserved chondrocyte pool for
the growth plate while remaining flexible, whereas articular cartilage
supports high compressive loads in the body while remaining flexible. A
fraction of resting chondrocytes may undergo hypertrophy even in normal
cartilage, whereas articular chondrocytes do not undergo hypertrophy
unless they become arthritic. The functional differences between these
cartilages suggest that the factors regulating maturation in resting
and articular cartilage should differ. In fact, PTH-related protein is
involved in inhibiting hypertrophy in resting, but not articular
cartilage in vivo (32), although BMP-2 and thyroid hormone
induce hypertrophy in both resting and articular chondrocyte
cultures.
Although ConA and WGA activate insulin receptors in adipocytes (2-4),
insulin induced only a very low level of ALPase activity in
chondrocytes.2 Thus, the stimulation of ALPase is not
mediated by the binding of ConA to insulin receptors. In addition, ConA
did not down-regulate PTH, TGF- ConA enhanced the synthesis of 15-kDa cystatin-like protein by resting
chondrocytes. Since ConA also enhanced this synthesis by articular
chondrocytes,3 the cystatin-like protease inhibitor is
unlikely to mediate the action of ConA on chondrocyte hypertrophy.
Previous studies have shown that ConA stimulates and inhibits the
synthesis of matrix-metalloproteinases and TIMP inhibitor,
respectively, by fibroblasts (35). These findings suggest that some
ConA-binding surface glycoproteins are involved in the control of
matrix breakdown in connective tissues.
Interactions of chondrocytes with artificial substrate impair their
phenotypic expression. Thus ConA may enhance proteoglycan and ALPase
synthesis via inhibition of their spreading on tissue culture surface.
We tested this notion by comparing the action of ConA with that of
succinyl ConA. Tetravalent ConA induced extensive cross-linking of
membrane glycoproteins, whereas the divalent ConA derivative did not.
This accounted for the differences between the actions of ConA and
succinyl ConA on lymphocytes (36). In this study, ConA, but not
succinyl ConA, inhibited the spreading of chondrocytes, although these
lectins exerted the same effects on ALPase activity, proteoglycan
synthesis, and total protein synthesis. Thus, extensive cross-linking
of cell-surface glycoproteins is required for the rapid cell-shape
change, but not for chondrocyte differentiation.
The tertiary structures of ConA and the ConA-sugar complex, as well as
its sugar-binding site, have been characterized (37). In addition, the
oligosaccharide structure required for ConA binding has been determined
(38), and a family of peptides that mimic the binding of MeMan to ConA
have been identified from screening a hexapeptide epitope library (39).
This information will be useful for future studies on the action of
ConA on animal cells.
In conclusion, the present study showed that ConA selectively acts on
resting chondrocytes to initiate the maturation program, and that the
ConA stimulation of chondrocyte differentiation is not directly linked
with lectin-induced changes in the cytoskeleton. Resting chondrocytes
exposed to ConA may be useful as a novel model for studies on
endochondral bone formation.
We gratefully acknowledge Dr. J. M. Wozney
(Genetics Institute, Cambridge, MA) and Dr. K. Takahashi (Yamanouchi
Pharmaceutical Co., Tokyo) for gifts of BMP-2. We thank the Research
Center for Molecular Medicine, Hiroshima University School of Medicine,
for the use of their facilities.
Biochemistry and ** Radiology,
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES
Materials
-methylmannoside, collagenase (Type
IA), insulin, retinoic acid, and triiodothyronine (T3) were
purchased from Sigma; Eagle's medium,
-modification was from Flow Laboratories (McLean, VA); fetal bovine
serum was from Life Technologies, Inc.; and fibronectin and type II
collagen were from Koken Co. (Osaka, Japan). Human bone morphogenetic
protein-2 (BMP-2, recombinant) was a gift of Dr. J. M. Wozney (Genetics
Institute, Cambridge, MA) and Dr. K. Takahashi (Yamanouchi
Pharmaceutical Co., Tokyo). Human transforming growth factor
-1
(TGF-
1, recombinant) and human insulin-like growth factor-I (IGF-I,
recombinant) were purchased from Wako Pure Chemical (Osaka, Japan).
Human calcitonin was purchased from Peninsula Laboratories Inc.
(Belmont, CA). 1,25(OH)2-Vitamin D,
1,25-dihydroxy[26,27-methyl-3H]cholecalciferol
(180 Ci/mmol), and parathyroid hormone (PTH-(1-34), human) were
supplied by Dr. K. Sato (Chugai Pharmaceutical Co., Tokyo).
45CaCl2 (37 mCi/mg) and
[35S]sulfate (carrier-free) were obtained from DuPont NEN
and the Japan Atomic Energy Research Institute (Tokyo), respectively. D-[6-3H]Glucosamine (27 Ci/mmol),
L-[3-3H]serine (28 Ci/mmol), and
[6-3H]thymidine (2 Ci/mmol) were purchased from Amersham
(UK).
-modification, supplemented with 10% fetal bovine serum, 50 µg/ml
ascorbic acid, 32 units/ml penicillin, and 40 µg/ml streptomycin
(Medium A). Alternatively, cells were seeded at densities of 2 × 105 cells/35-mm dish and 106 cells/100-mm dish,
and grown in 2 and 10 ml of Medium A, respectively. The cultures were
supplied with fresh Medium A every other day. After reaching confluence
in 6-, 35-, and 100-mm dishes (day 5 or 7), the cells were transferred
to Eagle's medium,
-modification in the presence of various lectins
and 0-2% fetal bovine serum for 6 or 24 h, after which the cell
layers were washed three times with Medium A. Because lectins bind to
serum glycoproteins, the serum concentration was reduced during the
incubation with lectins for 6-24 h. Thereafter, the cells were
maintained for 0-28 days in Medium A, unless otherwise specified.
-modified Eagle's medium supplemented with 2% fetal bovine serum
in the presence or absence of lectins or growth factors for 24 h.
Thereafter, the cells were maintained for 0-28 days in Medium A in the
absence of lectins and in the presence or absence of growth factors.
The media were replaced every other day. Unless otherwise specified,
lectins were added to the cultures only once on day 5 or 7, whereas
growth factors were added to the cultures every other day from day 5 or
7.
-modified Eagle's medium 3 h before the end of the incubation.
Cell-matrix layers were homogenized in 0.9% NaCl, 0.2% Triton X at
0 °C, then centrifuged for 15 min at 12,000 × g.
The precipitate was washed with 0.1 M CaCl2 in
0.005 M Tris-HCl, pH 7.4, at 20 °C for 30 min to remove
exchangeable 45Ca, then solubilized by incubation in 0.5 M HCl for 3 h at 20 °C (21). The levels of
45Ca were determined in aliquots.
-modified Eagle's medium without serum from days 12 to 15 and 17 to 20. Proteins in the conditioned media on days 12-15 and 17-20 were
separated by SDS-PAGE, then stained with silver nitrate.
-D-mannopyranoside for 24 h. The cells were
exposed to [35S]sulfate (0.5 µCi/culture) for 17 h, D-[3H]glucosamine (3 µCi/culture) for
6 h, L-[3H]serine (3 µCi/culture) for
6 h, or [3H]thymidine (1 µCi/culture) for 3 h
before the end of the incubation. We estimated the level of
proteoglycan synthesis by measuring the incorporation of
[35S]sulfate,
D-[3H]glucosamine, and
[3H]serine into material precipitated with
cetylpyridinium chloride after digestion with Pronase E (24). Total
protein and DNA synthesis were determined by measuring the
incorporation of [3H]serine and
[3H]thymidine into 5% trichloroacetic acid-insoluble
cell precipitates, respectively (24).
Effect of Lectins, Hormones, and Growth Factors on ALPase Activity
in Chondrocytes
Fig. 1.
Comparison of the effect of ConA on ALPase
with that of hormones and growth factors in resting chondrocyte
cultures. Rabbit resting chondrocytes were seeded at
104 (A) or 3 × 103 cells
(B) per 6-mm well, and maintained in Medium A. On day 5 (A) or 7 (B), the cells were exposed or not to
ConA, T3, BMP-2, IGF-I, PTH, or TGF-1 at the indicated
concentrations in the presence of 2% fetal bovine serum for 24 h.
They were then incubated in Medium A in the presence or absence of
growth factors or hormones. Hormones and growth factors were added to
the cultures every other day until the end of the incubation, whereas
ConA was added only once on day 5 or 7. ALPase activity was determined
on day 15 (A) or 17 (B). The values are
averages ± S.D. of four cultures.
[View Larger Version of this Image (39K GIF file)]
1 decreased ALPase activity in ConA-free resting
chondrocyte cultures (Fig. 1). These peptides also suppressed the ConA
stimulation of ALPase synthesis by resting chondrocytes,2
as predicted from published studies (10, 27, 31, 32). Basic fibroblast
growth factor, a potent inhibitor of hypertrophy (14), also decreased
ALPase activity in the absence and presence of ConA.2 Thus,
ConA did not down-regulate receptors for these inhibitors of
hypertrophy.
Fig. 2.
Effects of ConA, T3, and BMP-2 on
ALPase activity in high density cultures of articular
chondrocytes. Rabbit articular chondrocytes were seeded at
104 cells/6-mm well, and maintained in Medium A. On day 5, the cells were exposed to ConA at 10 µg/ml, T3 at
107 M, or BMP-2 at 100 ng/ml in the presence
of 2% fetal bovine serum for 24 h. Thereafter, they were
incubated in ConA-free medium containing 10% fetal bovine serum and
the growth factors. T3 and BMP-2 were added to the cultures
every other day until the end of the incubation, whereas ConA was added
only once on day 5. ALPase activity was determined on days 14 and 18. The values are averages ± S.D. of four cultures.
[View Larger Version of this Image (39K GIF file)]
Fig. 3.
Effect of increasing concentrations of
various lectins on ALPase activity in cultures of resting chondrocytes
(RC, A) and growth-plate chondrocytes
(GC, B). Cells were seeded at
104 cells/6-mm well and maintained in Medium A. On day 5, they were exposed to ConA (closed circles), succinyl ConA
(open squares), PHA-E, lentil lectin, lima bean lectin,
garden pea lectin, Scotch broom lectin, M. promifera
(MPA), lotus lectin, WGA, PHA-L, or U. europeaus
agglutinin (I + II) at the indicated concentrations in the presence of
2% fetal bovine serum for 24 h, then incubated in medium
containing 10% fetal bovine serum. ALPase activity was determined on
day 15. The values are averages ± S.D. of triplicate cultures.
[View Larger Version of this Image (35K GIF file)]
Fig. 4.
Effects of ConA on the levels of DNA and
uronic acid, and ALPase activity in resting chondrocyte cultures.
Resting chondrocytes were seeded at 104 cells/6-mm well,
and maintained in Medium A. On day 5, they were exposed (closed
circles) or not to 10 µg/ml ConA (open circles) in
the presence of 2% fetal bovine serum for 24 h. They were then incubated in the medium containing 10% fetal bovine serum. The DNA
(A) and uronic acid levels (B) and ALPase
activity (C) were determined on the indicated days. The
values are averages ± S.D. of four cultures.
[View Larger Version of this Image (22K GIF file)]
-mannopyranoside (MeMan) on ALPase
Activity
Fig. 5.
Effect of MeMan on ALPase activity in the
presence or absence of ConA. Resting chondrocytes were seeded at
104 cells/6-mm well and maintained in Medium A. A, on day 5, the cells were not exposed (open
circles) or exposed to 10 µg/ml ConA (closed circles)
in the presence of various concentrations of MeMan for 24 h. They
were then incubated in the medium supplemented with 10% fetal bovine
serum. ALPase activity was determined on day 15. B, On day 5, the cells
were not exposed (open circles) or exposed to ConA at 10 µg/ml for 24 h (closed circles). MeMan at 40 mM was added to the cultures on the indicated days
(open triangles). The values are averages ± S.D. of
four cultures.
[View Larger Version of this Image (21K GIF file)]
Fig. 6.
Morphology of resting (A and
B) and articular (C and D)
chondrocytes maintained with (B and D) or
without ConA (A and C). Resting
chondrocytes were seeded at 2 × 105 cells/35-mm dish
and maintained in Medium A. On day 7, they were exposed or not to 10 µg/ml ConA in the absence of serum for 24 h, then incubated in
-modified Eagle's medium containing 5% fetal bovine serum, 50 µg/ml ascorbic acid, and antibiotics. They were further exposed to 10 µg/ml ConA in the absence of serum for 24 h on days 12 and 17, and photographed on day 20. The size and diameter of 50 cells/population were measured using a computer-assisted video camera
and the image-processing software, Image 1 (Universal Imaging Co., West
Chester, PA). Bar, 37 µm.
[View Larger Version of this Image (163K GIF file)]
Fig. 7.
Effects of ConA on type II and X collagen
contents of chondrocyte cultures. Resting chondrocytes were seeded
at 104 cells/6-mm well and maintained in Medium A. ConA (10 g/ml) was added or not on day 5 for 24 h. Collagen contents in day
15 cultures were determined as described under "Experimental
Procedures."
[View Larger Version of this Image (41K GIF file)]
Fig. 8.
Effects of ConA on the binding of
1,25(OH)2-vitamin D to chondrocytes. Resting
chondrocytes were seeded at 2 × 105 cells/35-mm dish,
and maintained in Medium A. On day 7, they were exposed or not to 10 µg/ml ConA in the absence of serum for 24 h, then incubated in
Medium A. On days 14 and 20, the cells were homogenized and the
cytosols were incubated with 3H-labeled
1,25(OH)2-vitamin D in the presence or absence of excess nonradioactive 1,25(OH)2-vitamin D at 20 °C for 1.5 h. The values are averages ± S.D. of four cultures.
[View Larger Version of this Image (17K GIF file)]
Fig. 9.
Effects of ConA on the incorporation of
45Ca into insoluble material and the calcium content in the
cell layer of resting chondrocytes. Resting chondrocytes were
seeded, maintained, and exposed (closed bars) or not to 10 µg/ml ConA (open bars) as described in the legend to Fig.
6. On day 35, the incorporation of 45Ca into insoluble
material (A) and the calcium level (B) were determined, and the cell layers in the ConA-free (a) and
ConA-exposed (b) cultures were stained with alizarin red
(C). The values are averages ± S.D. of four
cultures.
[View Larger Version of this Image (60K GIF file)]
Fig. 10.
SDS-PAGE profiles of secreted proteins in
the media conditioned by chondrocytes treated with or without
ConA. Resting chondrocytes were seeded at 106
cells/100-mm dish and maintained in Medium A. On days 7, 12, and 17, the cells were exposed (B and D) or not to 10 µg/ml ConA (A and C) in the absence of serum
for 6 h. They were incubated in Medium A from days 7 to 12 and
from days 15 to 17. These cells were incubated in -modified Eagle's
medium in the absence of serum from days 12 to 15 and from days 17 to
20. Proteins in the conditioned media on days 12-15 (A and
B) and days 17-20 (C and D) were
concentrated 20-fold using a Centriprep (Amicon,
Mr 10,000 cut-off). The proteins in 10 µl of
the solution were separated by SDS-PAGE and then stained with silver
nitrate.
[View Larger Version of this Image (52K GIF file)]
Fig. 11.
Effects of ConA and succinyl ConA on
proteoglycan synthesis, total protein synthesis, and DNA
synthesis. When resting chondrocyte cultures in 6-mm wells reached
confluence, the cells were incubated in 0.1 ml of medium containing
0.3% fetal bovine serum (Medium B). They were then incubated in 0.1 ml
of fresh Medium B containing various concentrations of lectins for
24 h (A-D). Alternatively, they were incubated in 0.1 ml of fresh Medium B containing various concentrations of MeMan, 10 µg/ml lectin, or both for 24 h (E). These cells were
exposed to [35S]sulfate (A) for 17 h,
D-[3H]glucosamine (B) for 6 h, L-[3H]serine (C and
D) for 6 h, or [3H]thymidine
(E) for 3 h before the end of incubation. Proteoglycan synthesis was estimated by measuring the incorporation of
[35S]sulfate (A),
D-[3H]glucosamine (B), or
[3H]serine (C) into material precipitated with
cetylpyridinium chloride after digestion with Pronase E. Total protein
(D) and DNA syntheses (E) were determined by
measuring the incorporation of [3H]serine and
[3H]thymidine into 5% trichloroacetic acid-insoluble
cell precipitate, respectively. The values are averages ± S.D. of
four cultures.
[View Larger Version of this Image (36K GIF file)]
Fig. 12.
Effects of ConA and succinyl ConA on
chondrocyte spreading. Resting chondrocytes were seeded onto
fibronectin-coated dishes and incubated in the presence of 0 (A), 1 (B), 3 (C), and 10 µg/ml
(D) ConA or 30 µg/ml succinyl ConA (E) for
3 h.
[View Larger Version of this Image (127K GIF file)]
1, and fibroblast growth factor
receptors in chondrocytes.
*
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
§§
To whom correspondence should be addressed: Dept. of
Biochemistry, School of Dentistry, Hiroshima University, 1-2-3, Kasumi, Minami-ku, Hiroshima-city 734, Japan. Tel./Fax: 82-257-5629.
1
The abbreviations used are: ConA, concanavalin
A; PHA, phytohemagglutinin; WGA, wheat germ agglutinin; T3,
triiodothyronine; PTH, parathyroid hormone; PAGE, polyacrylamide gel
electrophoresis; MeMan, methylmannoside; TGF-1, transforming growth
factor
-1; IGF-I, insulin-like growth factor I; ALPase, alkaline
phosphatase.
3
M. Nishimura, unpublished data.
2
W. Yan, unpublished data.
©1997 by The American Society for Biochemistry and Molecular Biology, Inc.