/ . Embryol. exp. Morph. Vol. 37, pp. 49-57, 7977
Printed in Great Britain
49
The effects of ascorbic acid
deficiency on collagen synthesis by mouse molar
tooth germs in organ culture
By JOHN R. SCHILTZ,1 JOEL ROSENBLOOM2
AND GORDON E. LEVENSON23
From the School of Dental Medicine,
University of Pennsylvania, Philadelphia
SUMMARY
Second molar tooth germs from 2-day-old Swiss-Webster mice, grown in organ culture for
7 days in ascorbic-acid-deficient medium, synthesized about 65 % as much protein (measured
by incorporation of [14C]proline during a 24-h pulse) as did ascorbic-acid-supplemented
controls. The newly synthesized proteins from ascorbic-acid-deficient cultures contained
only about 7 % of the hydroxyproline content of the controls. Collagenase digestion of the
newly synthesized proteins showed that collagen comprised the same fraction of the total
protein synthesized under both culture conditions. This result indicates that the ascorbatedeficient cultures made significant quantities of underhydroxylated collagen. Partial characterization of the collagen alpha chains on carboxymethyl cellulose columns showed an al/a2
ratio of about 5, suggesting that at least two different species of collagen were synthesized.
The a] /a2 ratio of the chains recovered from the ascorbate-deficient cultures was also about 5
but the chains were slightly underhydroxylated and the total amount of these chains which
could be identified accounted for only a small fraction of the total collagen which was
synthesized. A large fraction of the synthesized collagenous protein was found in the culture
medium, mostly in the form of lower molecular weight peptides. It is concluded that
most of the collagen which is synthesized by ascorbate-deficient tooth-bud cultures is
not utilized by the component tissues, but is probably degraded and released into the
medium.
INTRODUCTION
Studies on the effects of vitamin C (ascorbic acid) on tooth germ development (Levenson, 1976) showed that in ascorbate-deficient medium the differentiated state of ameloblasts and odontoblasts was not maintained, dentine
deposition ceased and tooth germs generally became severely disorganized and
aplastic. The large pulp mass and odontoblasts were particularly affected by
ascorbate deficiency. Since these resemble connective tissue in origin and
1
Author's address: Department of Medicine, Division of Dermatology, Case Western
Reserve University of Medicine, Cleveland, Ohio 44106, U.S.A.
2
Authors' address: Department of Histology-Embryology, University of Pennsylvania
School of Dental Medicine, Philadelphia, Pennsylvania 19174, U.S.A.
3
Reprint requests to Dr Levenson.
50
J. R. SCHILTZ, J. ROSENBLOOM AND G. E. LEVENSON
function, it was suspected that altered collagen synthesis might be the major
lesion causing the altered development.
Since collagen is known to be produced by growing tooth germs (Weinstock & LeBlond, 1974), and ascorbic acid is known to be a cofactor for the
enzymic hydroxylation of prolyl and lysyl residues in collagen alpha chains, it
was of interest to determine the effects of ascorbate deficiency on the amount of
collagen synthesized, the type(s) of collagen alpha chains synthesized1 and the
extent of hydroxylation of the alpha chains.
MATERIALS AND METHODS
Organ culture techniques
Paired second molar tooth germs from 2-day-old Swiss-Webster mice were
grown on millipore strips in medium BGJ supplemented with 15 % fetal calf
serum (Flow Labs) and 2 % beef embryo extract (Grand Island Biological Co.).
Medium was changed every 48 h. Ascorbate medium received 150/^g/ml
powdered ascorbic acid at the time of each feeding. Cultures were maintained
at 37 °C under 50 % O2, 45 % N 2 and 5 % CO2 in humidified MaclntoshFieldes jars at atmospheric pressure. Details of the culture techniques have been
presented (Levenson, 1976). On the seventh day of culture, 6-8 tooth germs,
still on small pieces of millipore, were pooled in single dishes and handled as
follows.
Biochemical techniques
Culture dishes containing 6-8 tooth-germs received fresh medium and were
allowed to equilibrate for 1 h. At this time, 10/*Ci[14C]proline (New England
Nuclear, NEC 285) was added and the cultures were incubated for a prescribed
period of time. In all these experiments, cultures which had been grown without
ascorbate were pulsed without ascorbate. The tooth-germs were then removed
from the medium, washed with Earle's balanced salt solution (the wash solution
was added to the medium) and homogenized. The homogenates and media
were then dialyzed extensively against cold running tap water, hydrolyzed with
6 N-HC1 at 110 °C for 24 h and the content of [14C]imino acid determined by
the method of Juva & Prockop (1966). The values are expressed as ([14C]hydroxyproline/[14C]hydroxyproline + [14C]proline) x 100.
1
The collagen molecule which is found in vertebrate bone or tendon consists of three
alpha chains, each with a molecular weight of about 96000. Two of these chains, called al(I)
have identical primary structures which are different from the third, which is termed the a2
chain. The composition of this trimer is indicated as [al(I)]2a2 (Piez, Eigner & Lewis, 1963).
Other collagens have been described which have the molecular composition of [<*l(x)]3,
where the al(x) can be the al(II) chain of cartilage (Miller & Matukas, 1969; Trelstad, Kang,
Igarishi & Gross, 1970; Strawich & Nimni, 1971), the al(III) chain which partially comprises
the collagen fraction from skin, smooth muscle and vascular tissue (Miller, Epstein & Piez,
1971; Chung & Miller, 1974); or the al(IV) chain of basement membrane (Kephalides,
1971).
Collagen synthesis in ascorbate deficiency
51
In order to separate and identify collagen a-chains, cultures were pulsed with
[14C]proline in the presence of 100/ig/ml /?-amino-proprionitrile-fumarate and
the organs extracted for 2 days at 4 °C in 1 M-NaCl in 50 mM Tris-HCl, pH 7-4.
The extracts were mixed with 5 mg lathyritic chick skin collagen to serve as a
carrier, partially purified by precipitation with NaCl and subjected to carboxymethyl cellulose (CMC column chromatography as previously described
(Schiltz, Mayne & Holtzer, 1973)). The [14C]imino acid content of the labelled
chains isolated from the carboxymethyl cellulose columns was determined by
hydrolyzing the chains and separating the [14C]imino acids on the long column
of a Beckman amino acid analyzer. Separation of the [14C]-labeled proteins by
polyacrylamide disc gel electrophoresis in sodium dodecyl sulfate and mercaptoethanol was performed as previously described, using 5 % acrylamide (Harsch,
Murphy & Rosenbloom, 1972). Collagenase digestion of labeled proteins was a
modification of the method described by Peterkofsky & Diegelmann (1971).
Tooth-buds and medium were combined and homogenized with a Dounce
homogenizer (Wheaton). The homogenates were then heated at 100 °C for
15 min in a boiling water bath in order to destroy endogenous proteolytic
enzyme activity and the homogenates were dialyzed extensively against cold
distilled water. A portion of the dialyzed suspension was used for determination
of [14C]hydroxyproline and [14C]proline content and the remaining portion
used for the collagenase sensitivity assay. Replicate 1 ml aliquots of the homogenate suspension were added to 0-5 ml of a 0-15 M Tris-HCl (pH 7-2) buffer
containing 0-015 M-CaCl2 and 0-15 mg/ml clostridial collagenase (Worthington,
CLSPA). The mixtures were then incubated in capped tubes in a shaking water
bath at 37 °C for 6 h. The digests were dialyzed against distilled water and the
retentates and dialysates counted in a Beckman liquid scintillation counter using
an aqueous scintillation counting fluid containing 0-33 g POPOP x 3-33 g PPO +
333 ml Triton X-100 (Rohm & Haas, Philadelphia) and toluene to 1 1. of
solution. Control tubes without collagenase were incubated and dialyzed to
determine loss due to non-specific adherence of the labeled proteins to the
dialysis bags. The results are expressed in terms of the percentage of the counts
which were made dialyzable by collagenase.
RESULTS
Mouse molar tooth germs cultured for 7 days in an ascorbic-acid-deficient
medium show a progressive morphological deterioration and disorganization
(Levenson, 1976). In order to determine if this deteriorated condition is reflected in the rates of synthesis of collagen and non-collagen proteins, organ cultures grown for 7 days in the presence or absence of ascorbic acid were pulsed
for 24 h with [14C]proline. At the end of the incubation period, the toothbuds were separated from the medium and analyzed separately for [14C]proline
and [14C]hydroxyproline content as described in the Materials and Methods. The
results in Table 1 represent the average values from three separate experiments.
52
J. R. SCHILTZ, J. ROSENBLOOM AND G. E. LEVENSON
-
£
0
10
20
30
20
30
5
40
50
Fraction no.
Fig. 1. Carboxymethyl cellulose chromatography of newly synthesized collagen from
2-day mouse tooth buds grown for 7 days in the presence or absence of ascorbic
acid. Both cultures were then pulsed with io/*Ci of [14C]proline for a 24-h period
and the collagen extracted, partially purified and co-chromatographed with carrier
chick skin collagen. Upper panel, with ascorbate; lower panel, without ascorbate.
O—O, cpm/0-5 ml; solid line, O.D. 230 nm.
The total incorporation of [14C]proline in the ascorbate-deficient cultures was
only about 64 % of the ascorbate-supplemented controls, whereas the synthesis
of collagen as measured by the presence of [14C]hydroxyproline was about 7 %
of the controls.
In an attempt to determine the nature of the collagen synthesized by control
and ascorbate-deficient tooth-germs, 7-day cultures were labeled with [14C]proline
for 24 h in the presence of ytf-APN (/?-aminoproprionitrile). /?-APN prevents
crosslinking of the newly synthesized collagen a-chains and renders the molecule
more soluble. The cultures were then extracted, the extracted collagen partially
purified by salt precipitation with 5 mg of carrier lathyritic chick skin collagen and
the collagen then subjected to chromatography on columns of carboxymethyl
cellulose. The chromatograms are shown in Fig. 1. In both control and ascorbate-
Collagen synthesis in ascorbate deficiency
53
Table 1. Effects of ascorbic acid on the synthesis of collagen and
non-collagen protein by cultured tooth-germs
Tooth-germs from 2-day mice were grown in organ culture in the presence or absence
of 150/Ag/ml ascorbic acid. The cultures were then incubated 24 h with 10/tCi/
culture [14C]proline in the same media. The organs were separated from the medium
and analyzed for [l4C]proline and [14C]hydroxyproline content as described in
Materials and Methods. The values represent the mean from three separate experiments which involved six tooth-germs per culture. The values for ([14C]hypro/[14C]hypro+ [14C]pro) x 100 are followed by the respective standard deviations.
Percentage
of total
cpm
Culture
With ascorbate
Without
ascorbate
Percentage
of total
hypro
Total
cpm per
culture Cells Med Cells Med
553518
351538
730
750
270
250
78-7 21-3
53-3 46-7
([14C]Hypro/
[14C]hypro + [14C]pro) x 100
Cells
10•7 ±1-75
0 •7 ±0-28
Med
Per culture
6-4 ±0-92 9-60+ 1-85
1-8 ± 0 1 6 0-98 ± 0-24
deficient culture conditions, radioactivity (open circles) co-eluted with the
carrier a-chains (solid line), although the total counts recovered were much less
in the ascorbate-deficient cultures. The radioactivity in these peaks migrated on
sodium dodecyl sulfate polyacrylamide gels with carrier a-chains (96000
Daltons, data not shown). When these labeled peaks were hydrolyzed and
analyzed for [14C]imino acid content, the ([14C]hypro/[14C]hypro + [14C]pro)
x 100 values were as follows: with ascorbate al = 43-1, a2 = 40-7; without
ascorbate al = 41-3, a2 = 30-7. Thus it appears that, in the absence of ascorbic
acid, a much smaller proportion of the radioactivity which is incorporated into
proteins can be recovered as collagen a-chains and the extent of hydroxylation of
the recovered chains is slightly reduced (the usual values range from 44 to 46 %
for collagen a-chains). Whether or not ascorbate was present, the al/a2 ratios
were between 5 and 6. Since an exclusive synthesis of [al(I)] 2 a2-type collagen
(as found in tendon) would yield an al/a2 ratio of 2, it can be estimated by
calculation that approximately 54 % of the total collagen synthesized by these
tooth-germs is of the form [al(X)]3, where al (X) could be the a 1(11) chain found
in cartilage, the al(III) chain found in a number of tissues or the al(IV) which
appears to be unique to basement membrane. Although we have not identified
the chain types made by these organ cultures, since the ratios of recovered enchains made in the presence or absence of ascorbate were similar, the data
suggest that the types of collagen molecules synthesized in the presence or
absence of ascorbate are similar.
After a 24-h pulse with [14C]proline (Table 1), approximately one-fourth of
the total labeled proteins were released into the media (27 % for the controls
and 25 %for the ascorbate-deficient cultures). The partitioning of hyeroxyproline,
however, is quite different. In the controls, 21 % of the total [14C]hydroxyproline
54
J. R. S C H I L T Z , J. ROSENBLOOM AND G. E. LEVENSON
40
Fig. 2. Gel electrophoresis of labeled proteins released by ascorbate-supplemented
and ascorbate-deficient cultures. Labeled proteins from the media of the experiment
described in Fig. 1 were treated with 1 % mercaptoethanol and 1 % sodium dodecyl
sulfate and then subjected to electrophoresis on 5 % polyacrylamide disc gels. The gels
were fractionated into 1-5 mm slices, solubilized by heating in 15 % H2O2 at 55 °C for
6 h and counted. Duplicate gels were run containing marker rat tail a-chains.
These gels were stained with Coomassie blue. The a-chains were found in slices
11-13.
was released into the medium as contrasted to 47 % in the ascorbate-deficient
cultures. This suggests a differential release of collagenous components
in these deficient cultures. In order to characterize the newly-synthesized
proteins with respect to molecular weight, the radioactive proteins found in
the media of the experiment described in Fig. 1 were subjected to electrophoresis on polyacrylamide gels in sodium dodecyl sulfate (Fig. 2). In both
culture conditions there was a wide spectrum of molecular weights, with values
ranging from greater than 200000 Daltons to less than 10000 Daltons, including
proteins with the mobility of pro-a-chains and a-chains of collagen. The major
difference appeared to be that the ascorbate-deficient cultures contained a
Collagen synthesis in ascorbate deficiency
55
Table 2. [uC]fmino acid content and sensitivity to digestion by collagenase
of the radioactive proteins synthesized by cultured tooth-germs
Tooth-germs from 2-day mice were grown 7 days in organ culture in the presence or
absence of 150/tg/ml ascorbic acid. The cultures were then incubated 3 h with
10/jCi/culture [14C]proline in the same medium. The organs and the respective
media were mixed and the labeled proteins analyzed for sensitivity to digestion by
collagenase and for their [14C]imino acid content as described in Materials and
Methods. The values represent the mean and standard deviation from three cultures,
each containing seven tooth-germs.
Culture
With ascorbate
Without ascorbate
Total cpm per culture
152020±6364
124410±5840
([»C]Hypro/
Percentage digested
[14C]hypro +
[14C]pro) x 100
by collagenase
3-51 ±009
0-96±005
14-7±3-04
14-1 ±1-25
larger fraction of the total labeled proteins in the smaller molecular weight
regions of the gel.
The foregoing results suggest that either less collagen was synthesized in the
absence of ascorbate or the collagen which was synthesized was greatly underhydroxylated. In order to decide between these two alternatives, the newly
synthesized radioactive proteins were tested for collagenase sensitivity. Cultures
which had been grown for 7 days with or without ascorbate were given a 3-h
pulse with [14C]proline in the same media and the cells and media were combined and immediately homogenized. The homogenates were then heated for
15 min at 100 °C and dialyzed extensively against cold H 2 O. The heat denaturation step was found to be necessary in order to destroy endogenous proteolytic
enzyme activity released by homogenization. The radioactive proteins were then
tested for digestibility with purified collagenase (Worthington, CLSPA) as
described in Materials and Methods. This experiment (Table 2) showed that
during the 3-h pulse the percentage of proline which had been converted to
hydroxyproline was much less in the ascorbate-deficient cultures than in the
supplemented cultures. During these 3-h pulses the percentage was always lower
than during the 24-h pulses (i.e. 3-51 % in Table 2 as compared to 9-6% in
Table 1), probably because the turnover of collagen is much less than that for
most other cellular proteins and consequently the [14C]hydroxyproline content
becomes enriched as the time of incubation increases. More important, the
percentage of labeled proteins digested by collagenase was the same in both
cultures. Since the specificity of this collagenase is such that only collagen or
collagen fragments are rendered dialyzable, we conclude that the rate of collagen
synthesis relative to total protein synthesis was the same in the presence or
absence of ascorbic acid, but that the collagen made in ascorbate-deficient
cultures was underhydroxylated.
56
J. R. SCHILTZ, J. ROSENBLOOM AND G. E. LEVENSON
DISCUSSION
Ascorbic acid has been shown to be a cofactor for the enzymic hydroxylation of prolyl and lysyl residues of collagen (Hutton, Tappel & Udenfriend,
1967). The present study was undertaken to determine if the gross morphological alterations which occur in organ cultures of mouse molar tooth germs grown
in ascorbic-acid-deficient media are due in part to an altered synthesis, modification or processing of collagen. The results indicated that the relative rate of
collagen synthesis in the ascorbate-deficient cultures was comparable to that of
ascorbate-supplemented cultures, but the hydroxyproline content of the collagen in the ascorbate-deficient cultures was markedly reduced. These findings
are in agreement with the work of Bates, Prynne & Levine (1972) and Peterkofsky (1972), who observed the production of hydroxyproline-deficient collagen
by fibroblasts grown in ascorbate-free media. What, then, is the disposition of
the underhydroxylated collagen synthesized by these ascorbate-deficient organ
cultures? Recent studies have demonstrated that hydroxyproline stabilizes the
triple helix of collagen (Berg & Prockop, 1973; Jimenez, Harsch & Rosenbloom,
1973; Rosenbloom, Harsch & Jimenez, 1973) and that underhydroxylated chains
synthesized by 3T6 fibroblasts in ascorbate-free medium were found in a random coil conformation at 37° (Ramaley, Jimenez & Rosenbloom, 1973). In this
state the molecules were sensitive to proteolytic enzymes, in contrast to the
resistance of the normal triple helical structure of collagen. Since the collagen
synthesized by the tooth germs in the ascorbate-free media was significantly
underhydroxylated, it is likely that it too was in a random coil conformation at
37 °C. As such this collagen would be sensitive to any proteolytic enzymes
found in the tissue either intra- or extra-cellularly. Supporting this notion that
the underhydroxylated molecules were preferentially degraded was the observation that a major fraction of the hydroxyproline in the ascorbate-free cultures
was found in the medium, largely in smaller molecular weight peptides. Furthermore, only a small proportion of the collagenous protein in the tissue incubated
in the ascorbate-free media could be extracted and recovered as alpha chains, in
contrast to the ascorbate-supplemented controls. These results are similar to
those of Chen & Raisz (1975), who found that embryonic rat bones cultured in
ascorbate-free medium synthesized underhydroxylated collagen and released it
into the medium, mostly in the form of degraded peptides.
In conclusion, the net result of ascorbate deficiency in these tooth-bud cultures is to deprive the organ of the normal quantity and quality of collagen.
This deficiency, at least in part, must explain the altered gross morphology of the
tissues.
This work was supported by National Institutes of Health Grants AM 14439 a n d
14526; also by U.S.P.H.S. Grant DE-02623 to The Center for Oral Health Research, University of Pennsylvania, Philadelphia, Pa. 19174.
Collagen synthesis in ascorbate deficiency
57
REFERENCES
C, PRYNNE, C. & LEVINE, C. (1972). Ascorbate-dependent differences in the hydroxylation of proline and lysine in collagen synthesized by 3T6 fibroblasts in culture. Biochim.
biophys. Acta 278, 610-616.
BERG, R. & PROCKOP, D. (1973). The thermal transition of a non-hydroxylated form of
collagen. Evidence for a role for hydroxyproline in stabilizing the triple-helix of collagen.
Biochem. biophys. Res. Commun. 52, 115-120.
CHEN, T. L. & RAISZ, L. G. (1975). The effects of ascorbic acid deficiency on calcium and
collagen metabolism in cultured fetal rat bones. Calcif. Tiss. Res. 17, 113-127.
CHUNG, E. & MILLER, E. (1974). Collagen polymorphism: characterization of molecules
with the chain composition [al(lll)] 3 in human tissues. Science, N.Y. 183, 1200-1201.
HARSCH, M., MURPHY, L. & ROSENBLOOM, J. (1972). Metabolism by isolated fibroblasts of
abnormal collagens containing analogues of proline or lysine. FEBS Letters 26, 48-52.
HUTTON, J. J., TAPPEL, A. L. & UDENFRIEND, S. (1967). Cofactor and substrate requirements
of collagen proline hydroxylase. Archs Biochem. Biophys. 118, 231-240.
JIMENEZ, S., HARSCH, M. & ROSENBLOOM, J. (1973). Hydroxyproline stabilizes the triple
helix of chick tendon collagen. Biochem. biophys. Res. Commun. 52, 106-114.
3
14
JUVA, K. & PROCKOP, D. (1966). Modified procedure for the assay of H- or C-labelled
hydroxyproline. Analyt. Biochem. 15, 77-83.
KEPHALIDES, N. (1971). Isolation of a collagen from basement membranes containing three
identical a-chains. Biochem. biophys. Res. Commun. 45, 226-234.
LEVENSON, G. (1976). Effect of ascorbic acid deficiency on mouse molar tooth germs cultivated in vitro. J. Embryol. exp. Morph. 36, 73-85.
MILLER, E., EPSTEIN, E., JR. & PIEZ, K. (1971). Identification of three genetically distinct
collagens by cyanogen bromide cleavage of insoluble human skin and cartilage collagens.
Biochem. biophys. Res. Commun. 42, 1024-1029.
MILLER, E. & MATUKAS, V. (1969). Chick cartilage collagen: a new type of al chain not
present in bone or skin of the species. Proc. natn. Acad. Sci., U.S.A. 64, 1264-1269.
PETERKOFSKY, B. (1972). The effect of ascorbic acid on collagen polypeptide synthesis and
proline hydroxylation during the growth of cultured fibroblasts. Archs biochem. Biophys.
152, 318-328.
PETERKOFSKY, B. & DIEGELMANN, R. (1971). Use of a mixture of proteinase free collagenases
for the specific assay of radioactive collagen in the presence of other proteins. Biochemistry,
N.Y. 10,988-994.
PIEZ, K., EIGNER, E. & LEWIS, M. (1963). The chromatographic separation and amino acid
composition of the subunits of several collagens. Biochemistry, N.Y. 2, 58-66.
RAMALEY, P. B., JIMENEZ, S. A. & ROSENBLOOM, J. (1973). Conformation of underhydroxylated collagen synthesized by 3T6 fibroblasts in culture. FEBS Letters 33, 187-191.
ROSENBLOOM, J., HARSCH, M. & JIMENEZ, S. (1973). Hydroxyproline content determines the
denaturation temperature of chick tendon collagen. Archs biochem. Biophys. 158, 474-484.
SCHILTZ, J., MAYNE, R. & HOLTZER, H. (1973). The synthesis of collagen and glycosaminoglycans by dedifferentiated chondroblasts in culture. Differentiation 1, 97-108.
STRAWICH, E. & NIMNI, M. (1971). Properties of a collagen molecule containing three
identical components extracted from bovine articular cartilage. Biochemistry, N.Y. 10,
3905-3911.
TRELSTAD, R., KANG, A., IGARISHI, S. & GROSS, J. (1970). Isolation of two distinct collagens
from chick cartilage. Biochemistry, N.Y. 9, 4993-4998.
WEINSTOCK, M. & LEBLOND, C. (1974). Synthesis migration, and release of precursor collagen by odontoblasts as visualized by radioautography after 3H-proline administration.
/. CellBiol. 60, 92-127.
BATES,
(Received 1 April 1976, revised 27 July 1976)