Vol. 166, No. 3, 1990 May 16, 1990
BIOCHEMICAL
Osteopetrotic
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1171-1176
(grey-lethal)
and TIUP in organ culture: Joan K. HeathI*, 1 Cell
Physiology
2 Connective Hospital
Received
March
John
J.
ReynoldsI,
bone produces collagenase regulation by vitamin A and
Murray
Department, Strangeways Research Cambridge CBI 4RN, England,
C. Meiklell*,+
Laboratory, UK
Worts
Causeway,
Tissue Research Unit, Department of Orthodontics, Eastman Dental and Institute of Dental Surgery, University of London, 256 Gray’s Inn Road, London WCIX 8LD, England, UK 12,
1990
Evidence has recently accumulated suggesting that osteoblasts play a direct role in bone resorption by producing collagenase. In this paper we describe studies carried out with explants of bone from osteopetrotic grey lethal (gl/gl) mice and show that despite the lack of osteoclastic activity the production of both active and latent collagenase and its specific inhibitor TIMP (tissue inhibitor of metalloproteinases) is similar to that of normal bones. Synthesis of collagenase was stimulated by the bone resorptive agent vitamin A (retinol); concomitantly, TIMP levels fell to zero and active enzyme was detected in the culture medium. This work supports the view that bone collagenase is produced by cells other than osteoclasts, since the response of the osteoblastic population to resorptive signals appears normal. 01990 Academic Press, Inc. Collagenase production by rodent osteoblasts in response to bone resorptive agents [l-43 led to the hypothesis that these bone cells play a major role in resorption, possibly by degrading the surface osteoid layer and thereby exposing the underlying mineralized matrix to osteoclastic action [for review see ref.51. More recent work suggests that osteoid removal is regulated in a complex manner which involves the stimulation of procollagenase production by osteoblasts, followed by activation of the enzyme extracellularly by proteolytic cleavage requiring plasmin [6,7]. Vitamin A was first shown to stimulate bone and cartilage resorption in organ cultures of embryonic limb bones by Fell and Mellanby C81, and recent evidence suggests that retinol and retinoic acid can stimulate bone resorption by a direct action on osteoclasts [9]. Vitamin A metabolites thus appear to be unique among bone-resorbing agents in being able to act directly on osteoclasts. Osteopetrotic animals (which lack active osteoclasts) provide * Present address: St.Vincent‘s Parade, Fitzroy 3065, Melbourne,
Institute Australia.
of Medical
Research,
41 Victoria
90 whom correspondence should be addressed at Strangeways Research Cambridge CBl 4RN, England, UK. Laboratory, Worts Causeway,
1171
0006-291x/90 $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
168, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
valuable experimental models for investigating the role of cells other than osteoclasts in bone resorption. Parathyroid hormone (PTH), 1,25dihydroxyvitamin D3 and prostaglandin Ez have been shown to stimulate collagenolytic activity by explants of osteopetrotic microphthalmic (mi/mi) bone [IO], and cultured calvarial cells from (mi/mi) mice respond to PTH by producing latent collagenase [II]. Previously, we reported that vitamin A (retinol) stimulates the synthesis of collagenase by calvarial explants from normal mice in vitro, and reduces that of its specific inhibitor TIMP [12]. In this paper, we show that calvarial explants from osteopetrotic grey-lethal (gl/gl) mice and heterozygous (gl/+) littermates respond to vitamin A in a similar fashion, providing additional evidence that the mechanisms of osteoid removal in (gl/gl) osteopetrotic bone are not affected by the osteoclast deficiency.
MATERIALS
AND METHODS
Animals. The grey-lethal colony bred at the Strangeways Research Laboratory (SRI,) was originally from stock donated by Professor H. Griineberg and Dr. H. Murphy, Department of Animal Genetics, University College, London. The (gl/gl) mice are most reliably distinguished from their littermates by failure of tooth eruption by 7-8 d of age. The (gl/+) and (+/+) mice are normally phenotypically identical and their precise genotypes can only be assigned by further breeding. This was not possible since for these experiments, the animals were killed at 7 d. To solve this problem, the autosomal recessive which is closely linked to the gl locus, was mutation called “downlessness”, introduced into the SRL breeding stock. This was done in such a way that whenever the alleles at the gl locus were homozygously wildtype the downless mutation would also be present homozygously: any phenotypically normal mice were therefore all (gl/+). An unrelated group of mice from the SRL animal house provided the normal (+/+) group. Bone resorvtion assay. Bone resorption was measured by the release of radiolabelled Ca*+ from explants of neonatal mouse calvariae in culture [IS]. Mice (3-4 d old) were injected with 4 nCi [45Ca2+]-CaC13 (Amersham International) per mouse. After a further 4 d to allow isotope to be incorporated into the skeleton, the mice were killed by chloroform anaesthesia. At this point, the different genotypes in each litter could be reliably distinguished immediately prior to the dissection procedure; explants of half-calvariae were obtained by microdissection. Following a 24 h preincubation, 0.5 ml medium was removed for scintillation counting and the cultures continued in fresh P4 medium (141 with or without vitamin A (7.26ng/ml of retinol in ethyl alcohol) for a further 48 h. Again, 0.5 ml medium was removed for scintillation counting, the bones were dried, digested in formic acid and the remaining 45Ca2+ in the bones measured. To determine 45Ca2+ release due to passive exchange of isotope with the cold Ca*+ in the culture medium, dead bones (obtained by 3 x 20 min cycles of freezing and thawing) were included in each experiment, Bone resorption was expressed as the percentage of the total bone isotope released into the medium. Assays for collagenase and TIMP. Collagenase activity in culture supernatants was measured by the release of [14C]-labelled peptides from [I%]-acetylated rat skin collagen [IS]. One unit of collagenase hydrolyses 1 ng of reconstituted type I collagen fibrils per min at 35-C. The latent form of collagenase was activated by including 4-aminophenylmercuric acetate (APMA) the fibril assay (0.67 r&l final concentration). TIMP was measured by incubating a standard amount (0.05 unit) of activated rabbit skin collagenase in the fibril assay with samples of culture medium. One unit of TIMP is 1172
in
Vol. 168, No. 3, 1990 defined as the by 50%.
BIOCHEMICAL
amount
of inhibitor
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
required
to inhibit
2 units
of collagenase
Characterization of collaxenolvtic activity and inhibitory activitv bv gelfiltration chromatogranhv. Conditioned medium from (gl/gl) explants treated with retinol for 48h was concentrated ten-fold by ultrafiltration and applied as a 4ml sample to a column (86 x 1.5 cm) of Ultrogel AcA 54 (LKB, Croydon, Surrey) equilibrated with 50 mM Tris pH 7.4, 1 H NaCl, 10 mM CaClz and 0.05% Brij-35. The column was eluted with the same buffer and 4 ml fractions were collected and assayed for collagenase activity in the presence of 0.67 mM APMA, and for TIMP as described above. RESULTS AND DISCUSSION Both (+/+) and (gl/+) explants responded to vitamin A by releasing significantly more 4sCa2* than unstimulated control explants. Explants from (gl/gl) mice, however, were unresponsive (data not shown) in keeping with data from earlier experiments (14). Moreover, endogenous 45Caz+ release from (gl/gl) explants was little different from dead (+/+) or (gl/+) explants, confirming that little or no cell-mediated demineralization was occurring in (gl/gl) bone.
into There
Unstimulated the culture were
released increased
no
after the
bones medium
all
released
during
significant
each
APMA-activatable of
differences
3 d by the three groups production of collagenase
TABLE 1. Daily
production
the
three
between
(latent)
days the
(Fig. 1). by (gl/gl)
in
total
collagenase
culture levels
1).
(Table of
collagenase
Vitamin A significantly and (+/+) explants,
of collagenase by mouse to Vitamin A
calvarial
explants
but
in
response
Collagenase
Phenotype
(units/ml)
N O-24h
24-48h
4a-72h
(+/+I Untreated + vit A
5
0.06to.01 0.08tO.01
0.16M.01 0.18kO.01
0.26+0.01 0.33ti.02~ (0.02+0.02)
Untreated t vit A
6
0.05M.02 0.09tO.01~
0.16+0.04 0.27?0.02~ (0.10?0.03)
0.25f0.04 0.41M.03~ (0.27k0.06)b
Untreated
4
0.03M.01 0.05ti.01
0.14kO.02 0.18tO.01 (0.02?0.02)
0.25kO.03 0.28tO.01 (0.06kO.04)
(sl/sl)
(gl/+)
+ Vit
A
Mouse half-calvariae were cultured for 3 d (with a daily medium change) with or without retinal (7.26 &ml). Culture media were assayed in either the presence (total enzyme) or absence (active enzyme) of APMA. Results are expressed as mean ?: SEM for the number of cultures (N). Figures in parentheses indicate the proportion of collagenase that was active. a Significantly higher than untreated controls, P < 0.05. b Significantly higher than active collagenase production by (gl/t) and (+/+) bones, P < 0.05 and < 0.01 respectively.
1173
not
Vol. 166, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
t
* 5) 7 It A
r+onelVlt Addmons to
NonelVit A medaa
A
Fig. 1. Cumulative (3 d) release of collagenase by explants of mouse calvariae obtained from phenotypically distinct mice. The medium harvested each day was assayed for collagenase activity either in the presence or absence of APMA, (0.67mbt). 0 total collagenase released (detected with APMA),B active collagenase released (detectable in the absence of APMA). Results are expressed as the mean of cumulative values f SRM with the number of cultures in each group shown in parentheses. * Significantly higher than paired control, PxO.05 ** Significantly higher than (gl/+) explants, P
BIOCHEMICAL
Osteopetrotic
AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 1171-1176
(grey-lethal)
and TIUP in organ culture: Joan K. HeathI*, 1 Cell
Physiology
2 Connective Hospital
Received
March
John
J.
ReynoldsI,
bone produces collagenase regulation by vitamin A and
Murray
Department, Strangeways Research Cambridge CBI 4RN, England,
C. Meiklell*,+
Laboratory, UK
Worts
Causeway,
Tissue Research Unit, Department of Orthodontics, Eastman Dental and Institute of Dental Surgery, University of London, 256 Gray’s Inn Road, London WCIX 8LD, England, UK 12,
1990
Evidence has recently accumulated suggesting that osteoblasts play a direct role in bone resorption by producing collagenase. In this paper we describe studies carried out with explants of bone from osteopetrotic grey lethal (gl/gl) mice and show that despite the lack of osteoclastic activity the production of both active and latent collagenase and its specific inhibitor TIMP (tissue inhibitor of metalloproteinases) is similar to that of normal bones. Synthesis of collagenase was stimulated by the bone resorptive agent vitamin A (retinol); concomitantly, TIMP levels fell to zero and active enzyme was detected in the culture medium. This work supports the view that bone collagenase is produced by cells other than osteoclasts, since the response of the osteoblastic population to resorptive signals appears normal. 01990 Academic Press, Inc. Collagenase production by rodent osteoblasts in response to bone resorptive agents [l-43 led to the hypothesis that these bone cells play a major role in resorption, possibly by degrading the surface osteoid layer and thereby exposing the underlying mineralized matrix to osteoclastic action [for review see ref.51. More recent work suggests that osteoid removal is regulated in a complex manner which involves the stimulation of procollagenase production by osteoblasts, followed by activation of the enzyme extracellularly by proteolytic cleavage requiring plasmin [6,7]. Vitamin A was first shown to stimulate bone and cartilage resorption in organ cultures of embryonic limb bones by Fell and Mellanby C81, and recent evidence suggests that retinol and retinoic acid can stimulate bone resorption by a direct action on osteoclasts [9]. Vitamin A metabolites thus appear to be unique among bone-resorbing agents in being able to act directly on osteoclasts. Osteopetrotic animals (which lack active osteoclasts) provide * Present address: St.Vincent‘s Parade, Fitzroy 3065, Melbourne,
Institute Australia.
of Medical
Research,
41 Victoria
90 whom correspondence should be addressed at Strangeways Research Cambridge CBl 4RN, England, UK. Laboratory, Worts Causeway,
1171
0006-291x/90 $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
Vol.
168, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
valuable experimental models for investigating the role of cells other than osteoclasts in bone resorption. Parathyroid hormone (PTH), 1,25dihydroxyvitamin D3 and prostaglandin Ez have been shown to stimulate collagenolytic activity by explants of osteopetrotic microphthalmic (mi/mi) bone [IO], and cultured calvarial cells from (mi/mi) mice respond to PTH by producing latent collagenase [II]. Previously, we reported that vitamin A (retinol) stimulates the synthesis of collagenase by calvarial explants from normal mice in vitro, and reduces that of its specific inhibitor TIMP [12]. In this paper, we show that calvarial explants from osteopetrotic grey-lethal (gl/gl) mice and heterozygous (gl/+) littermates respond to vitamin A in a similar fashion, providing additional evidence that the mechanisms of osteoid removal in (gl/gl) osteopetrotic bone are not affected by the osteoclast deficiency.
MATERIALS
AND METHODS
Animals. The grey-lethal colony bred at the Strangeways Research Laboratory (SRI,) was originally from stock donated by Professor H. Griineberg and Dr. H. Murphy, Department of Animal Genetics, University College, London. The (gl/gl) mice are most reliably distinguished from their littermates by failure of tooth eruption by 7-8 d of age. The (gl/+) and (+/+) mice are normally phenotypically identical and their precise genotypes can only be assigned by further breeding. This was not possible since for these experiments, the animals were killed at 7 d. To solve this problem, the autosomal recessive which is closely linked to the gl locus, was mutation called “downlessness”, introduced into the SRL breeding stock. This was done in such a way that whenever the alleles at the gl locus were homozygously wildtype the downless mutation would also be present homozygously: any phenotypically normal mice were therefore all (gl/+). An unrelated group of mice from the SRL animal house provided the normal (+/+) group. Bone resorvtion assay. Bone resorption was measured by the release of radiolabelled Ca*+ from explants of neonatal mouse calvariae in culture [IS]. Mice (3-4 d old) were injected with 4 nCi [45Ca2+]-CaC13 (Amersham International) per mouse. After a further 4 d to allow isotope to be incorporated into the skeleton, the mice were killed by chloroform anaesthesia. At this point, the different genotypes in each litter could be reliably distinguished immediately prior to the dissection procedure; explants of half-calvariae were obtained by microdissection. Following a 24 h preincubation, 0.5 ml medium was removed for scintillation counting and the cultures continued in fresh P4 medium (141 with or without vitamin A (7.26ng/ml of retinol in ethyl alcohol) for a further 48 h. Again, 0.5 ml medium was removed for scintillation counting, the bones were dried, digested in formic acid and the remaining 45Ca2+ in the bones measured. To determine 45Ca2+ release due to passive exchange of isotope with the cold Ca*+ in the culture medium, dead bones (obtained by 3 x 20 min cycles of freezing and thawing) were included in each experiment, Bone resorption was expressed as the percentage of the total bone isotope released into the medium. Assays for collagenase and TIMP. Collagenase activity in culture supernatants was measured by the release of [14C]-labelled peptides from [I%]-acetylated rat skin collagen [IS]. One unit of collagenase hydrolyses 1 ng of reconstituted type I collagen fibrils per min at 35-C. The latent form of collagenase was activated by including 4-aminophenylmercuric acetate (APMA) the fibril assay (0.67 r&l final concentration). TIMP was measured by incubating a standard amount (0.05 unit) of activated rabbit skin collagenase in the fibril assay with samples of culture medium. One unit of TIMP is 1172
in
Vol. 168, No. 3, 1990 defined as the by 50%.
BIOCHEMICAL
amount
of inhibitor
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
required
to inhibit
2 units
of collagenase
Characterization of collaxenolvtic activity and inhibitory activitv bv gelfiltration chromatogranhv. Conditioned medium from (gl/gl) explants treated with retinol for 48h was concentrated ten-fold by ultrafiltration and applied as a 4ml sample to a column (86 x 1.5 cm) of Ultrogel AcA 54 (LKB, Croydon, Surrey) equilibrated with 50 mM Tris pH 7.4, 1 H NaCl, 10 mM CaClz and 0.05% Brij-35. The column was eluted with the same buffer and 4 ml fractions were collected and assayed for collagenase activity in the presence of 0.67 mM APMA, and for TIMP as described above. RESULTS AND DISCUSSION Both (+/+) and (gl/+) explants responded to vitamin A by releasing significantly more 4sCa2* than unstimulated control explants. Explants from (gl/gl) mice, however, were unresponsive (data not shown) in keeping with data from earlier experiments (14). Moreover, endogenous 45Caz+ release from (gl/gl) explants was little different from dead (+/+) or (gl/+) explants, confirming that little or no cell-mediated demineralization was occurring in (gl/gl) bone.
into There
Unstimulated the culture were
released increased
no
after the
bones medium
all
released
during
significant
each
APMA-activatable of
differences
3 d by the three groups production of collagenase
TABLE 1. Daily
production
the
three
between
(latent)
days the
(Fig. 1). by (gl/gl)
in
total
collagenase
culture levels
1).
(Table of
collagenase
Vitamin A significantly and (+/+) explants,
of collagenase by mouse to Vitamin A
calvarial
explants
but
in
response
Collagenase
Phenotype
(units/ml)
N O-24h
24-48h
4a-72h
(+/+I Untreated + vit A
5
0.06to.01 0.08tO.01
0.16M.01 0.18kO.01
0.26+0.01 0.33ti.02~ (0.02+0.02)
Untreated t vit A
6
0.05M.02 0.09tO.01~
0.16+0.04 0.27?0.02~ (0.10?0.03)
0.25f0.04 0.41M.03~ (0.27k0.06)b
Untreated
4
0.03M.01 0.05ti.01
0.14kO.02 0.18tO.01 (0.02?0.02)
0.25kO.03 0.28tO.01 (0.06kO.04)
(sl/sl)
(gl/+)
+ Vit
A
Mouse half-calvariae were cultured for 3 d (with a daily medium change) with or without retinal (7.26 &ml). Culture media were assayed in either the presence (total enzyme) or absence (active enzyme) of APMA. Results are expressed as mean ?: SEM for the number of cultures (N). Figures in parentheses indicate the proportion of collagenase that was active. a Significantly higher than untreated controls, P < 0.05. b Significantly higher than active collagenase production by (gl/t) and (+/+) bones, P < 0.05 and < 0.01 respectively.
1173
not
Vol. 166, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
t
* 5) 7 It A
r+onelVlt Addmons to
NonelVit A medaa
A
Fig. 1. Cumulative (3 d) release of collagenase by explants of mouse calvariae obtained from phenotypically distinct mice. The medium harvested each day was assayed for collagenase activity either in the presence or absence of APMA, (0.67mbt). 0 total collagenase released (detected with APMA),B active collagenase released (detectable in the absence of APMA). Results are expressed as the mean of cumulative values f SRM with the number of cultures in each group shown in parentheses. * Significantly higher than paired control, PxO.05 ** Significantly higher than (gl/+) explants, P
