/ . Embryol. exp. Morph. Vol. 53, pp. 179-202, 1979
Printed in Great Britain © Company of Biologists Limited 1979
179
Changes in collagen ultrastructure,
macroscopic properties and chemical composition
of chick embryo cartilage induced by
administration of a A-D-xyloside
By J. T. HJELLE 1 AND K. D. GIBSON 2
From the Department of Physiological Chemistry and Pharmacology,
Roche Institute of Molecular Biology, New Jersey
SUMMARY
Nine-day chick embryos were injected with a /?-xyloside and their sternal cartilage was
examined 3 days and a week later. Sterna from 16-day embryos showed a reduction in size as
compared to controls, with little or no change in the fraction of extracellular space, and a
significant decrease in tensile strength. At the ultrastructural level, collagen fibrils in control
sterna were dispersed evenly in the interstitial space, with few contacts between adjacent
fibrils. In sterna from treated embryos, almost all collagen fibrils were aggregated into clumps
and arrays throughout the interstitial space, withfibril-freeareas in between. No abnormalities
could be detected in the morphology of individual fibrils or in the ultrastructure of the
chondrocytes. The changes in spatial distribution of collagen were fully evident 3 days after
drug administration.
The hydroxyproline/DNA ratio was the same in control and treated sterna, and no change
was observed in the type of collagen. The uronic acid/DNA ratio was reduced by 14% 3 days
after drug administration and by 40% after a week. The degree of sulfation of chondroitin
sulfate was reduced from 80% in control sterna to 40% in treated sterna; almost all of this
chondroitin sulfate was attached to peptide and the sedimentation pattern of the proteoglycan resembled that of normal cartilage proteoglycan.
The function of chondroitin sulfate in embryonic cartilage is discussed in terms of our
results and others. It is suggested that a major physiological role of the proteoglycan is to
control the spatial distribution of collagen fibrils as they assemble to form a cross-linked gel.
INTRODUCTION
The interstitial matrix of cartilage contains a three-dimensional network of
fibrils of Type-II collagen held together by apparently random covalent cross
links (Lane & Weiss, 1975; Miller, 1976,1977). Chemically, the cross links are of
several types, all of which probably result from reactions of the aldehyde
function of allysine or hydroxy-allysine (Tanzer, 1976). Interspersed in the
1
Author's address: International Institute of Cellular and Molecular Pathology, Brussels,
Belgium.
2
Author's address (for all correspondence): Department of Physiological Chemistry and
Pharmacology, Roche Institute of Molecular Biology, Nutley, New Jersey 07110, U.S.A.
180
J. T. HJELLE AND K. D. GIBSON
network of collagen fibrils are molecules of proteoglycan aggregate, each
consisting of a large number of molecules of proteochondroitin sulfate held by
non-covalent bonds to a long strand of hyaluronic acid, and stabilized by one
or more 'link' proteins (Comper & Laurent, 1978). The proteoglycan aggregate
may be held in place by entrapment within the cartilage network (Fessler, 1957;
Hamerman & Schubert, 1962) or by non-covalent interaction with collagen
(Lee-Owen & Anderson, 1976; Toole, 1976).
The physiological function of the collagen network is reasonably well
understood. The network probably provides most of the tensile strength of adult
cartilage (Kempson, Muir, Pollard & Tuke, 1973). The overall morphology of
adult cartilage is retained when it is subjected to rather stringent extraction
procedures, which remove more than 80% of the proteoglycans but leave
cross-linked collagen (Sajdera & Hascall, 1969). Agents that interfere with the
cross linking of collagen in vivo, such as lathyrogens, drastically reduce the
tensile strength of cartilage and cause abnormalities of its morphology if applied
during development (Barrow, Simpson & Miller, 1974). From studies of this
type it is likely that the collagen network forms the structural framework of the
cartilage (Sokoloff, 1969), in embryos and in adults.
The function of the proteoglycan aggregate is less well defined. In load-bearing
cartilages, such as occur at adult joints, the proteoglycan is probably largely
responsible for the ability of the cartilage to withstand deformation and to
recover its shape after local pressure has been applied (Kempson, Muir,
Swanson & Freeman, 1970). These properties of adult cartilage would not be
expected to be important in embryos. In spite of much speculation, no fully
satisfactory physiological role has been assigned to the proteoglycan in embryonic
cartilage. However, the existence of the genetic abnormalities nanomelia in
chicks, in which extremely stunted growth is associated with a large reduction
in the synthesis of the major cartilage proteoglycan (Landauer, 1965; Pennypacker & Goetinck, 1976), and brachymorphy in mice, in which reduced growth
is associated with undersulfation of cartilage chondroitin sulfate (Lane & Dickie,
1968; Orkin, Pratt & Martin, 1976), does suggest that correct embryonic growth
requires the presence in cartilage of a full complement of fully sulfated chondroitin sulfate.
In this study, we have used the teratological syndrome induced by administration of /?-D-xylosides to investigate the effects of altering the structure of
chondroitin sulfate on the ultrastructure of cartilage. Both in vitro and in vivo,
synthetic /?-xylosides prime the synthesis of chains of chondroitin sulfate
(Fukunaga et a/. 1975; Galligani, Hopwood, Schwartz & Dorfman, 1975;
Gibson & Segen, 1977; Gibson, Segen & Doller, 1979). When /?-xylosides are
administered to 9-day chick embryos, there is an increase in the total embryonic
synthesis of chondroitin sulfate and a decrease in the average degree of sulfation
of chondroitin sulfate, from about 80% to 40% (Gibson, Doller & Hoar, 1978;
Gibson et al. 1979). The major morphological changes in the embryos are
f}-D-Xyloside and cartilage in ovo
181
edema of all soft tissues and marked dwarfism; however, the skeleton shows no
morphological abnormalities and appears to ossify normally (Gibson et al. 1978).
Since there is probably little interference with other biochemical pathways, the
xyloside-induced syndrome appears to be an excellent model for investigating
the role of chondroitin sulfate in the growth and development of cartilage. In.
this paper, we present the results of an ultrastructural and chemical analysis of
sterna from 12- and 16-day chick embryos that had been treated with a /?xyloside at 9 days of age. The sternum was chosen because it does not undergo
ossification in the chick until the embryo reaches term (Romanoff, 1960).
MATERIALS AND METHODS
Nine-day chick embryos (stage 34-35 of Hamburger & Hamilton, 1951),
from an inbred block of White Leghorns, were obtained from Spring Lake
Farms, Wyckoff, NJ. Eggs were injected once with a suspension of 10 mg
4-methylumbelliferyl /?-D-xyloside in 0-1 ml sterile corn oil (Gibson et al. 1978,
1979); control eggs received corn oil only. Eggs were maintained in a humidified
incubator at 37 °C until sacrifice.
Sterna were removed from the embryos, dissected free of adhering muscle and
fixed for 1 h in cold (4 °C) glutaraldehyde/formaldehyde fixative, buffered with
0-2 M sodium cacodylate at pH 7-3. Post-fixation in 2 % OsO4, buffered with
0-144 M sodium cacodylate buffer, at room temperature was followed by
dehydration in an ascending series of graded acetone (50-100 %) and propylene
oxide before embedding in Epon 812 (Ladd Research Industries, Inc., Burlington
VT). Sectioning was carried out with an LKB-8800 A Ultramicrotome III. Thin,
sections were mounted on uncoated 200 mesh copper grids and stained with lead
citrate and uranyl acetate (5 % in absolute ethanol). Tissue examination was
carried out on a Jeol 100B electron microscope at original magnifications of
2000-20000.
For determination of their gross chemical composition, seven or eight pooled
sterna (wet weight 200-250 mg) were homogenized in 2 ml ice-cold water, using
a Brinkman Polytron homogenizer operated at full speed. Portions of the
homogenates were used for the determination of DNA (Burton, 1956) with calf
thymus DNA (Calbiochem., LaJolla, Calif.) as standard, total protein (Lowry,
Rosebrough, Farr & Randall, 1951) with bovine serum albumin (Sigma
Chemical Co., St Louis, Mo.) as standard, and total hydroxyproline (Kivirikko,
Laitinen & Prockop, 1967). Uronic acid was determined by the carbazole
reaction (Bitter & Muir, 1962) with glucuronic acid as standard, after digestion
of the homogenates with papain (Gibson et al. 1979). The degree of sulfation of
chondroitin sulfate was determined by subjecting crude glycosaminoglycans,
prepared from these papain digests, to digestion with chondroitin lyase AC
(EC 4.2.2.5) followed by paper chromatography as described (Gibson et al.
1979). Crude glycosaminoglycans were also analyzed by electrophoresis in
0-15 M-ZnSO4 on cellulose acetate strips (Gibson et al. 1979).
182
J. T. HJELLE AND K. D. GIBSON
The distribution of chain lengths of chondroitin sulfate, and the extent to
which it was attached to peptide or to the unnatural aglycone, were determined
by cleavage with alkaline [3H]NaBH4 according to Hopwood & Robinson
(1973) with modifications. Six sterna were suspended in 1 ml 0-05 M-[3H]NaBH4
(19 mCi/mmol)-0-2 M-NaOH and kept at 4 °C for 7 days. Acetone (50 (i\) was
then added, followed after 1 h by 50 /A glacial acetic acid; the precipitates were
removed by centrifugation and the supernatants were lyophilized. The residues
were dissolved in 0-15 ml water and applied to a column (0-8x93 cm) of
Sepharose 6B (Pharmacia Fine Chemicals, Inc., Piscataway, NJ) equilibrated
with 0-2 M-NaCl at 4 °C. The column was developed with 0-2 M-NaCl at a flow
rate of 2 ml/h. Fractions of 0-49 ml were collected with a 2112 Redirac fraction
collector (LKB Instruments, Inc., Hicksville, NY). Portions (0-1-0-2 ml) of each
fraction were used for estimating total glycosaminoglycan by precipitation with
Alcian blue (Whiteman, 1973), radioactivity expressed as nmol by comparison
with a standard of [3H]sorbitol prepared with the same batch of [3H]NaBH4
(Audhya, Segen & Gibson, 1976), and methylumbelliferone by fluorimetry after
acid hydrolysis (Gibson & Segen, 1977).
Proteochondroitin sulfate was extracted by stirring groups of six to eight
sterna with 15-20 vol of 4 M guanidinium chloride-0-05 M sodium acetate
(pH 5-7) at 4 °C for 24 h. The extraction was repeated once. Cesium chloride
(0-59 g/g of extract) was added to the combined extracts, and the solutions were
centrifuged at 40000 rev/min in the 65 rotor of a Spinco L2-65B ultracentrifuge
for 48 h at 10 °C. The contents of the tubes were split into five approximately
equal fractions, which were analyzed for uronic acid. Appropriate fractions (see
Fig. 13) were combined, dialyzed exhaustively against water and lyophilized.
The residues were dissolved in 0-15 ml 4 M guanidinium chloride buffered with
sodium acetate and layered on 5-20% linear gradients of sucrose in buffered
4 M guanidinium chloride (Kimata, Okayama, Oohira & Suzuki, 1974), The
gradients were centrifuged at 38000 rev/min in a SW40 rotor of a Spinco
L2-65B ultracentrifuge at 5 °C, until the u)2t value was 9 x 1011 rad 2 /sec (approx.
16 h). Fractions of 12 drops were collected by upward displacement and analyzed
for proteoglycan and glycosaminoglycan by precipitation with Alcian blue
(Whiteman, 1973).
For determination of prolyl hydroxylase (EC 1.14.11.2), extracts of pooled
sterna were prepared and assayed according to Hutton, Tappel & Udenfriend
(1966). We are grateful to Dr A. Fallon for performing these assays. For comparison of the type of collagen, lyophilized sterna (1-3 mg) were suspended in
70 % formic acid containing a 15-fold excess of cyanogen bromide and kept at
30 °C for 4 h with occasional vigorous shaking (Orkin et ah 1976). Excess
cyanogen bromide was removed by addition of ice-cold water followed by
lyophilization. The residues were analyzed by electrophoresis on 11 % polyacrylamide gels (0-5 x 6-5 cm) in the presence of sodium dodecyl sulfate, using
the buffer system of Neville (1971) with 0-4244 M Tris-O-0308 N-HCl (pH 9-18) in
fi-D-Xyloside and cartilage in ovo
183
Table 1. Properties of sterna from 16-day embryos
DNA
Treatment of Length of ventral Wet weight
Sorbitol space
embryos
edge of keel (mm)* C"g/mg)t (% of wet weight)*
Control
/?-Xyloside
1 3 0 ± 0 1 (13)
ll-5±0-2(8)J
3-33
3-25
78-1 ±0-5 (13)
83-4+1-7 (8)§
Work-to-fracture
(ergs)*
293-2±500 (9)
68-1 ±13-3 (8)||
* Results are expressed as mean ±S.E. for the numbers in parentheses,
t Mean of two determinations.
t P < 0001 by /-test.
§ P < 001 by /-test.
|| P < 005 by /-test, after correcting for the average difference in cross-sectional area of
the sterna (see the text).
the resolvent gel. The gels were fixed and stained according to Furthmayr &
Timpl (1971) and scanned in a Gilford Model 210 spectrophotometer equipped
with a Linear Transport system and a Honeywell recorder, using a wavelength
of 545 nm and a slit width of 0-1 mm. Samples of Type-I and Type-TI collagen
from lathyritic chick cartilage, prepared by the method of Trelstad, Kang, Toole
& Gross (1972), were subjected to the same treatment.
For determination of sorbitol space, up to 10 sterna were incubated for 4 h
at 37 °C in 2 ml of a synthetic medium (Audhya et al. 1976) containing 0-1 nmol
[3H]sorbitol (19 mCi/mmol, prepared by reducing glucose with [3H]NaBH4
according to Audhya et al. (1976)). The sterna were rapidly and thoroughly
blotted, weighed and dissolved by incubation in 1 ml 90% formic acid for 48 h
at 37 °C. Portions (50 fi\) of the incubation medium were treated in the same
way. Radioactivity was determined after addition of 10 ml Aquasol and the
percentage sorbitol space of each cartilage was calculated from its wet weight and
radioactivity and the radioactivity of the medium.
Load-extension curves of sternal fragments from 16-day embryos were
measured in an Instron Universal Testing machine. The posterior two thirds of
a sternum was excised and the keel was cut off under a dissecting microscope,
leaving a 7-9 mm long fragment of cartilage of very nearly constant width
(1-3-1-7 mm, depending on the individual cartilage) and thickness (0-2-0-3 mm).
Specimens were stretched at a rate of 10 cm/min until fracture and load was
recorded on a continuously moving chart. After fracture the samples were
examined for evidence of slip; if this had occurred the recording was discarded.
Work-to-fracture was calculated by graphical integration of the load-extension
curve.
RESULTS
Administration of /?-xylosides to 9-day chick embryos results in a significant
decrease in skeletal size with little change in the relative proportions of the
calcified skeletal elements (Gibson et al. 1978). The same changes were seen in
184
J. T. HJELLE AND K. D. GIBSON
C2
Cl
C3
I
1
5 ,
1
4 ,
1
2 _
1 1
0
1
J
/\
El
1
11u
1
3 -
\
•J V.
-
4
2
E2
0 4 2 0 4
Extension (mm)
E3
2 0
Fig. 1. Load versus extension curves for 16-day sterna. Cartilage fragments were
prepared as described and stretched in a tensitometer to fracture. C1-C3, curves
for three representative control cartilage fragments; E1-E3; curves for three
xyloside-treated fragments.
sterna from these embryos. The overall appearance of sterna from 16-day
xyloside-treated embryos was indistinguishable from that of control sterna, but
their size was clearly decreased. This is expressed quantitatively in Table 1, in
which the lengths of the ventral edge of the keels of several sterna from 16-day
control and treated embryos are compared. Treatment with the xyloside
induced an 11-5% decrease in this linear dimension, corresponding to a 31 %
decrease in total volume of the sternum (Table 1). The wet weight/dry weight
ratio had the same value (11-1-11-2) in sterna of both types, indicating that the
large relative increase in the water content of whole embryos treated with
/?-xylosides (Gibson et ah 1979) is not reflected in the cartilage. The wet
weight/DNA ratio was also the same in sterna of both types, suggesting that
there was no great change in the ratio of extracellular to intracellular space. This
contrasts with the situation in nanomelia, in which there is an obvious decrease
in the space between chondrocytes (Pennypacker & Goetinck, 1976). Further
evidence that the volume of extracellular space is not decreased in sterna from
xyloside-treated embryos comes from a determination of sorbitol space, which
was slightly greater in the treated sterna than in the control group (Table 1).
Thus in contrast to the edema of the soft tissues of xyloside-treated embryos,
which is associated with a preferential increase in the extracellular space (Gibson
fl-D-Xyloside and cartilage in ovo
'•' '.'-"'• " ft "•'•'['
•'"'^•v-
Fig. 2. Low-power electron micrograph of sternum from a control 16-day chick
embryo. The interstitial space contains a network of evenly spaced fibrils and
'electron-dense granules'. Nu, nucleus, er; endoplasmic reticulum. The bar
represents 1 /<m.
et al. 1979), the dwarflsm is accompanied by little change in the relative volume
of the extracellular space of cartilage.
Sterna from xyloside-treated embryos were noticeably more fragile than
control sterna during dissection and handling, although their fragility was not as
marked as that of lathyritic sterna. To quantitate the change in tensile strength,
we made use of the fact that the posterior two thirds of the sternum of a 16-day
embryo has a nearly constant width. Fragments of sterna were prepared by
removing the keel from the posterior two thirds of the sternum and stretched in
a tensitometer until fracture. Plots of load versus extension for three representative control sterna and three sterna from treated embryos are shown in Fig. 1.
The shape of the curves is similar in both cases, with fracture occurring at
2-3 mm extension (from a nominal initial length of 1-5 mm). However, the load
at fracture of the treated sterna was much smaller than that of the control
186
J. T. HJELLE AND K. D. GIBSON
Fig. 3. Low-power electron micrograph of sternum from a xyloside-treated 16-day
chick embryo. Fibrils in the interstitial space are aggregated into bundles (arrows).
Nu, nucleus. The bar represents 1 /*m.
sterna. Values of work-to-fracture were obtained by integrating plots like those
in Fig. 1 for nine control sterna and eight sterna from xyloside-treated embryos.
The value for one treated sternum fell within the range for the control sterna,
while the remainder were below the lower limit. The mean work for the xylosidetreated was 27 % of the mean for the control sterna (Table 1). Since the fragments of cartilage used were approximately rectangular and of constant width,
the work-to-fracture values are reasonably representative of the actual tensile
strengths of the cartilages. Allowing for the fact that the cross-sectional area of
the fragments from the xyloside-treated embryos was on the average 78 % of
that of the controls, the difference in work-to-fracture was significantly lower
after xyloside treatment (Table 1).
An electron micrograph of a section of sternum from a normal 16-day chick
embryo is shown in Fig. 2. Chondrocytes displaying ultrastructural features that
have been described in detail by others (Anderson, Chacko, Abbott & Holtzer,
P-D-Xyloside and cartilage in ovo
Fig. 4. Electron micrograph of sternum from a control 16-day embryo. A tract of
fibrils with nearly parallel orientations crosses the field from left to right, and is
spanned by the double arrow. The bar represents 1 /tm.
187
188
J. T. HJELLE AND K. D. GIBSON
4
,
Fig. 5. Electron micrograph of sternum from a xyloside-treated 16-day chick
embryo. A long aggregate of fibrils (f) separates two chondrocytes. The bar
represents 1 /*m.
1970; Pennypacker & Goetinck, 1976) are surrounded by a uniform extracellular matrix composed of randomly oriented, evenly spaced fibrils and
'electron-dense granules'. In general, the chondrocytes are not completely
surrounded by well-differentiated lacunae, and the matrix shows approximately the same density of fibrils and granules throughout. In accordance with
many published studies (Martin, 1954; Anderson, et al. 1970; Lane & Weiss,
1975; Pennypacker & Goetinck, 1976), we intrepret the fibrils as being composed
of Type-II collagen, The nature of the electron-dense granules is less certain.
Some of these may be cross sections of collagen fibrils; however, most workers
consider the larger granules at least to be composed of proteochondroitin
sulfate (Matukas, Panner & Orbison, 1967; Anderson et al. 1970; Kochhar,
Aydelotte & Vest, 1976; Pennypacker & Goetinck, 1976).
Comparison of the extracellular matrix in Fig. 2 to that of a sternum from a
16-day embryo injected on day 9 with 10 mg of 4-methylumbelliferyl /?-Dxyloside (Fig. 3) reveals striking differences. Almost all collagen fibrils in the
fi-n-Xyloside and cartilage in ovo
189
Fig. 6. Electron micrograph of sternum from a xyloside-treated 16-day chick embryo
A large bundle offibrils(f) is at least 1 /tm from the nearest cell process (p). The bar
represents 1 /tm.
treated cartilage are aggregated into bundles or arrays throughout the extracellular space. The bundles are in general well separated from each other and
make infrequent contact. Large areas devoid of fibrils are prominent. The
number of electron-dense granules may be somewhat decreased relative to the
control cartilage. Examination of many similar sections from control and
xyloside-treated sterna indicated that the average distance between chondrocytes may have been slightly lower in the treated than in control cartilage, but
the difference was not great. The cells are very much less crowded than in
nanomelic chick embryo sternum, which exhibits almost total loss of electrondense granules (Pennypacker & Goetinck, 1976).
Figures 4-6 are electron micrographs of control and xyloside-treated cartilage
at higher magnifications. Despite the differences in the extracellular matrix,
chondrocytes from control and xyloside-treated embryos are ultrastructu rally
identical (Figs. 4 and 5). In the control matrix, collagen fibrils often exhibit a
13
EMB 53
190
J. T. HJELLE AND K. D. GIBSON
Fig. 7. Electron micrograph of sternum from a control 12-day chick embryo. Most
fibrils in the extracellular space are well separated from each other. Nu, nucleus;
er, endoplasmic reticulum; m, mitochondrion. The bar represents 1 /im.
fi-D-Xyloside and cartilage in ovo
191
Fig. 8. Electron micrograph of sternum from a xyloside-treated 12-day chick
embryo. Almost all extracellular fibrils are gathered into bundles (arrows), m, Mitochondrion. The bar represents 1 /tm.
nearly parallel alignment, but there are almost always electron-lucent spaces
between neighboring fibrils (Fig. 4). There is a marked difference from the
longitudinally associated bundles of fibrils which frequently adjoin, chondrocytes in the xyloside-treated cartilage (Fig. 5) but are also seen in areas of the
extracellular space that appear to be at some distance from cells (Fig. 6).
Similarly oriented tracts of fibrils, which however are well spaced from one
another, can be discerned in corresponding locations of control cartilage. Except
for the difference in distribution, the collagen fibrils of xyloside-treated cartilage
are not distinguishable from those of the control.
The structure of the fibrillar network is altered after only 3 days of xyloside
treatment. Cartilage from a 12-day-old control embryo shows a fairly homogeneous distribution of fibrils (Fig. 7), while that of a xyloside-treated embryo
contains only clumps and arrays of fibrils (Fig. 8). In general, chondrocytes in
the control sternum have a uniformly distributed matrix within 0-5 urn of the
13-2
192
J. T. HJELLE AND K. D. GIBSON
Table 2. Chemical composition of sterna
Age of
embryos
(days) Treatment
12
16
Control
/?-Xyloside
Control
/?-Xyloside
Protein
(/*g//*g
DNA)
Hydroxyproline
0"g//*g
DNA)
23-4
19-7
22-2
18-4
0-50
0-46
1-49
1-43
Prolyl
Per cent of chondroitin
hydro- Uronic
sulfate as
acid i
xylase
N
6Un4(units//ig 0*g//*g
DNA) DNA) Sulfate Sulfate sulfated
*
1-36
51
30
19
21
60
117
19
*
690
34
36
3-88
40
17
19
64
2-32
700
* Not determined.
cell surface, and isolated collagen fibrils can be observed at all distances from the
cell (Fig. 7). In the xyloside-treated cartilage, almost all fibrils are associated into
bundles, whether they are close to the cell surface or at some distance from it
(Fig. 8). This suggests that collagen fibrils aggregate into bundles as soon as they
are formed.
Chemical analysis of chick embryo sternal cartilage after administration of
xyloside is shown in Table 2. There was a slightly lower protein/DNA ratio in
cartilage from treated embryos relative to control embryos. In spite of the
marked ultrastructural changes in the organization of the collagen component
of the extracellular matrix, no consistent difference in the hydroxyproline/DNA
ratio could be detected. There was also no difference in prolyl hydroxylase
(EC 1.14.11.2) activity of sterna from control and xyloside-treated 16-day
embryos. Sterna from 16-day control and xyloside-treated embryos were treated
with cyanogen bromide and analyzed by gel electrophoresis in the presence of
sodium dodecyl sulfate (Fig. 9). The same major peptides were present in both
cases. The electrophoretic pattern showed major bands in positions corresponding to the CNBr peptides of Type-II collagen, supporting the ultrastructural
evidence which indicates that the major collagen in xyloside-treated sterna, as in
control sterna, was Type II.
The uronic acid content of xyloside-treated sterna was 14% less than that of
control sterna 3 days after drug administration and 40 % less in 16-day embryos
(Table 2). This contrasts with the analyses of whole 16-day embryos, which show
increased accumulation of uronic acid after xyloside treatment (Gibson et al.
1979). Electrophoresis of sternal glycosaminoglycans revealed a polydispersity
of charge in the xyloside-treated sterna, suggesting varying degrees of sulfation
of the predominant glycosaminoglycan, chondroitin sulfate; whereas the charge
density of the chondroitin sulfate from control sterna was more nearly constant
(Fig. 10). Quantitation of the disaccharides released by chondroitin lyase
digestion showed a marked undersulfation in the xyloside-treated sterna
(Table 2). Sixty per cent of all chondroitin disaccharides were unsulfated in
fi-D-Xyloside and cartilage in ovo
193
Bottom
Fig. 9. Cyanogen bromide cleavage of sterna. Sterna from 16-day embryos were
cleaved with CNBr and the resulting peptides separated by electrophoresis on SDS
polyacrylamide gels. The gels were stained and scanned at 545 nm. The positions of
the major CNBr peptides from Type-I and Type-II chick collagen are shown at the
top of the figure.
xyloside-treated cartilage, while only 20-25 % were unsulfated in control
cartilage. Undersulfation was fully evident at day 12 and continued through
day 16.
Gibson et al. (1979) reported that chondroitin sulfate extracted from whole
16-day embryos treated with methylumbelliferyl ^-xyloside was markedly
undersulfated and shorter in chain length, with 75 % of the chains being linked
to the fluorescent aglycone. To determine whether this was true for cartilage
also, control and treated sterna were reduced with sodium borotritide to simultaneously cleave chondroitin chains from the core protein and stoichiometrically radiolabel the freed chondroitin chains (Hopwood & Robinson, 1973).
When analyzed on a gel nitration column, chondroitin sulfate prepared in this
way from control or xyloside-treated 16-day embryos migrated in an identical
manner (Fig. 11). The amount of fluorescence associated with the chondrotin
sulfate peak was insignificant (Fig. 11, lower panel), indicating that very little
methylumbelliferyl chondroitin sulfate was present. These findings indicate that
the chondroitin sulfate chains of cartilage from xyloside-treated embryos are
linked predominantly to a core peptide and have the same distribution of chain
length as the chondroitin sulfate chains from control cartilage.
Proteoglycans were extracted from control 16-day sterna and subjected to
isopycnic centrifugation in a 'dissociative' cesium chloride gradient. Nearly
194
J. T. HJELLE AND K. D. GIBSON
Fig. 10. Electrophoresis of sternal glycosaminoglycans from control and xylosidetreated embryos. Standards of chondroitin 6-sulfate (1), dermatan sulfate (2),
heparan sulfate (3) and hyaluronic acid (4) were applied to the left-hand strip;
glycosaminoglycans from 16-day control (C) or xyloside-treated (X) sterna were
applied to the center strip; and glycosaminoglycans from 12-day control (Q or
xyloside-treated (X) sterna to the right-hand strip. The arrow irrdicates the point of
application; the direction of migration is towards the top.
fi-D-Xyloside and cartilage in ovo
195
— 4-0 =
-
30 5
20
Elution vol. (ml)
Fig. 11. Gel-filtration chromatography of end-labeled sternal chondroitin sulfate.
End-labeled chondroitin sulfate was produced by cleavage in alkaline [3H]NaBH4
and analyzed on a column of Sepharose 6B. The void volume and total volume of the
column were at 13-6 and 31-6 ml respectively. Upper panel, chondroitin sulfate from
control 16-day sterna; lower panel, chondroitin sulfate from xyloside-treated 16-day
sterna.
90 % of the uronic acid from control sterna was found at the bottom of the
gradient (Fig. 12, upper left panel). When this material was analyzed by zone
sedimentation, most of the proteoglycan migrated well into the gradient
(Fig. 12, right panel, CTT), indicating that it had a large molecular weight
(Kimata et al. 1974). There was also a minor peak near the top of the sucrose
gradient, which may correspond to the 'ubiquitous' proteoglycan found in
chick embryo cartilage as well as other tissues (Palmoski & Goetinck, 1972;
Okayama, Pacifici & Holtzer, 1976). The very small amount of uronic acid at
the top of the original CsCl gradient did not correspond to any high molecular
weight proteoglycan when analyzed by zone sedimentation (Fig. 12, right panel,
CI). A different picture was obtained when proteoglycan from xyloside-treated
sterna was subjected to isopycnic centrifugation in a CsCl gradient. Only 75 %
of the uronic acid was in the bottom three fifths of the gradient, while the remainder banded at a density close to 1-44 g/ml (Fig. 12, lower left panel).
Each of these fractions was subjected to zone sedimentation and gave a pattern
very similar to that shown by the proteoglycan from control sterna (Fig. 12,
196
J. T. HJELLE AND K. D. GIBSON
1-40 3
1 2
3 4
5
Fraction number
10 15
5
Fraction number
Fig. 12. Isopycnic centrifugation and zone sedimentation of proteoglycans from
sterna. Proteoglycans were extracted from 16-day sterna and subjected to isopycnic
centrifugation in dissociative CsCl gradients (left panel). The gradients were cut into
five fractions, which are numbered from top to bottom. Fractions were combined as
shown and subjected to zone sedimentation in a linear 5-20 % sucrose gradient under
dissociative conditions (right panel).CI and CII, proteoglycans from control 16-day
sterna; El and EH, proteoglycans from xyloside-treated 16-day sterna. Cl and El
represent the combined material from the upper two fifths of the CsCl gradients in
the left panel; CII and EII represent the combined material from the lower three
fifths of those gradients. The direction of sedimentation was towards the right.
right panel, El and EII). Thus the proteoglycan from xyloside-treated sterna had
approximately the same sedimentation properties as that of control sterna, but
at least 25 % of it had a much lower buoyant density. These data are compatible
with the presence of a normal core peptide together with a change in the degree
of sulfation. Essentially the same results were obtained with proteoglycans
extracted from 12-day control and xyloside-treated sterna.
DISCUSSION
From a morphological point of view, administration of /?-xylosides produces
two major changes in the cartilage of chick embryos. Microscopically, there is
a decrease in size of the cartilage with no significant change in its shape, and at
the ultrastructural level there is a marked distortion in the spatial distribution of
collagen fibrils. Dwarfism seems to be a common feature of all syndromes in
which the structure or synthesis of proteochondroitin sulfate is specifically
fi-D-Xyloside and cartilage in ovo
197
impaired. The most marked effect occurs in nanomelia (Landauer, 1965), a
recessive genetic abnormality in which the synthesis of the major cartilage
proteoglycan is greatly reduced (Pennypacker & Goetinck, 1976). A less marked
change is seen in brachymorphy (Lane & Dickie, 1968), which is characterized
by undersulfation of cartilage chondroitin sulfate (Orkin et al. 1976). The
dwarfism in these conditions is not accompanied by extreme skeletal defects.
Dwarfism is also a feature of the action of the glutamine analog 6-diazo-5oxonorleucine (DON), both in vivo (Greene & Kochhar, 1975) and in mouse
limb-bud cultures in vitro (Aydelotte & Kochhar, 1975). Among other metabolic
effects, DON inhibits the synthesis of chondroitin sulfate by interfering with
the formation of glucosamine (Ghosh, Blumenthal, Davidson & Roseman,
1960; Telser, Robinson & Dorfman, 1965). The skeletal malformations seen in
embryos treated with DON probably result from interference with other
metabolic pathways, whereas the dwarfism can be reversed by administration of
glucosamine (Aydelotte & Kochhar, 1975; Greene & Kochhar, 1975).
The proteoglycan aggregate in the extracellular space of cartilage exerts a
significant Donnan osmotic pressure (Ogston, 1970; Comper & Laurent, 1978),
which should influence the total volume of the interstitial fluid. A decrease in the
rate of secretion of the proteoglycan or its glycosaminoglycan sidechains will lead
to a lower Donnan osmotic pressure and less interstitial fluid per chondrocyte.
This can account for part of the dwarfism observed in nanomelia, where there is
a clear reduction in the relative volume of interstitial space (Pennypacker &
Goetinck, 1976). Since administration of /?-xylosides led to some decrease in. the
glycosaminoglycan content of the cartilage (Table 2), a similar but smaller
reduction in interstitial space might be expected to occur under these conditions.
However, our observations seem to indicate little or no decrease in the relative
volume of interstitial space (Table 1), and suggest that the dwarfism induced by
/?-xylosides is due to a reduction in the rate of cell division of chondrocytes. The
effect may be rather specific for cartilage, since there is little reduction in the
total DNA content of 16-day xyloside-treated embryos (Gibson et al. 1979).
A decrease in the rate of cell division of chondrocytes is a likely contributory
factor in the other cases of dwarfism associated with abnormal proteoglycan
structure or content discussed above. Little is known about the influence of the
components of the extracellular matrix on the growth of differentiated embryonic
cartilage, although both collagen and proteochondroitin sulfate promote the
differentiation of chondrogenic tissue to cartilage in vitro (Kosher, Lash &
Minor, 1973; Kosher & Church, 1975).
The changes in spatial distribution of collagen fibrils induced by administration of /?-xylosides resemble changes seen in nanomelia and in limb-bud
cultures incubated in the presence of DON. In nanomelia, the synthesis and type
of cartilage collagen are normal, but the fibrils in the interstitial matrix are
closely packed in long arrays resembling those in Fig. 5 (Pennypacker &
Goetinck, 1976). In DON-treated cultures, collagen fibrils in the interstitial
198
J. T. HJELLE AND K. D. GIBSON
matrix of cartilage are gathered into bundles and clumps (Kochhar et al. 1976).
The fibrils tend to be thicker than usual and show some periodicity, a feature
that is not normally obvious in Type-II collagen in chick embryo cartilage
(Martin, 1954). We have not been able to document changes in periodicity of
individual fibrils in cartilage from xyloside-treated chick embryos, but the
changes in spatial distribution of collagen fibrils (Figs. 5, 6 and 8) are very
similar to those induced by DON.
Obvious mechanisms by which changes in proteoglycan structure could
influence the spatial distribution of collagen fibrils include alterations in the
primary structure of collagen, interference with cross linking and changes in the
rate of synthesis or secretion. Morphologically, the fibrils in xyloside-treated
cartilage closely resemble those in control cartilage, and are very different from,
for instance, the fibers of Type-I collagen which abound in the perichondria on
the same sterna. In addition, cyanogen bromide cleavage experiments indicate
that the major type of collagen is the same in control and xyloside-treated
cartilage. Thus changes in the structure of the matrix are not accompanied by a
significant shift in the type of collagen produced by the chondrocytes. Interference with cross linking has not been examined directly. However, cross-linkdeficient cartilage, produced by treating chick embryos with /9-aminopropionitrile, shows a distribution of collagen fibrils that is different from that found
in control or xyloside-treated cartilage. While the ultrastructure and spatial
distribution of most of the collagen fibrils resemble normal cartilage, the
interstitial matrix contains many individual thin and twisted collagen fibrils
that are not cross linked to other fibrils (J. T. Hjelle and K. D. Gibson, unpublished). Since no such fibrils are seen in xyloside-treated cartilage, we infer
that the xyloside probably does not directly affect the mechanism for forming
cross links. Administration of xyloside does not change the rate of synthesis
or deposition of collagen in sterna in short-term in vitro incubations (J. T. Hjelle
and K. D. Gibson, unpublished). Further, the amount of hydroxyproline per
mg DNA was the same in sterna from xyloside-treated and control embryos
3 days and a week after drug administration (Table 2), indicating that in vivo
the rate of synthesis of collagen per cell was unchanged. Thus the changes in
spatial distribution of collagen fibrils described here are probably not secondary
consequences of changes in chemical structure or rate of synthesis of collagen,
but result from a direct influence of proteochondroitin sulfate on the formation
of the fibrillar network.
Some insight into the physiological significance of the changes in collagen
spacing induced by /?-xylosides or DON may be obtained by noting that there
is a strong similarity between the formation of a cross-linked collagen network
in the interstitial space of cartilage and the formation of a synthetic network
polymer in the presence of a diluent. When a synthetic network polymer such
as moderately cross-linked polystyrene is formed in the presence of a good
solvent, polymer chains and monomer subunits remain freely dispersed in the
fi-D-Xyloside and cartilage in ovo
199
solution at all stages of the polymerization and the final network is homogeneous, with all polymer chains evenly distributed in space (Millar, Smith,
Marr & Kressman, 1963). If the solvent is replaced by a non-solvating diluent,
for which the polymer chains and monomers have a low affinity, the reaction
mixture tends to separate locally into two phases, one enriched in polymer and
monomer and the other consisting mainly of diluent. The polymer chains become
linked into a heterogeneous, macroporous network, in which areas of tightly
packed polymer are enmeshed with pores that are free of polymer (Millar,
Smith & Kressman, 1965). Macroporous gels formed in the presence of a nonsolvent have reduced tensile strength and elasticity as compared to homogeneous
gels with the same chemical composition (Millar et al. 1965). Both differences
are due to the existence within the network of many surfaces that are crossed by
only a few polymer chains; such surfaces offer weak spots at which fracture will
occur readily. It should be emphasized that the weakness of these networks is
not due to deficiencies in the number of cross links, but results from the erratic
spatial distribution of the polymer chains.
The analogy with the geometry of the collagen network in control and
xyloside-treated cartilage is evident, although the underlying physical causes
may be quite different. The distribution of collagen fibrils in control cartilage
resembles the distribution of polymer chains in a homogeneous network polymer
whereas the collagen network in treated cartilage resembles a macroporous
network. The reduced tensile strength of the treated cartilages (Table 1) is
consistent with the propertes of macroporous networks, although as discussed
above, the possibility that cross linking is abnormal in these cartilages has not
been ruled out. The increase in sorbitol space of the treated cartilages is also
consistent with macroporosity, since the excluded volume associated with the
macromolecular components of the extracellular space (Ogston, 1958; Comper
& Laurent, 1978) will be partly compensated by the presence of freely permeable
pores.
Based on this analogy, and taking into account the macroscopic and ultrastructural changes seen in cartilages in which the structure or synthesis of
proteochondroitin sulfate is defective, we propose that a major function of the
proteoglycan in the interstitial space of embryonic cartilage is to ensure that
collagen fibrils are well dispersed before they assemble to form a cross-linked
gel. Load-bearing cartilages require adequate tensile strength and elasticity, and
these properties are intimately connected with the geometry and topology of the
collagen network (Sokoloff, 1969; Kempson et al. 1973). If a homogeneous
network with optimal mechanical properties is to be formed, not only must
there be the correct number of cross links, but the collagen fibrils must be
evenly dispersed in the interstitial matrix at the time that they are assembled
into the network. We suggest that the spatial distribution of newly formed
collagen fibrils is determined by the proteoglycan aggregate, probably through
non-covalent binding of Type-II collagen and cartilage proteochondroitin
200
J. T. HJELLE AND K. D. GIBSON
sulfate (Lee-Own & Anderson, 1976; Toole, 1976). Correct spacing of collagen
fibrils in the cartilage of the young or adult animal will thus depend on the
correct distribution of proteochondroitin sulfate within the cartilage during
embryonic growth; if this is impaired because of abnormalities in the chemical
composition or amount of the proteoglycan, collagen fibrils may be permanently
assembled into a mechanically deficient network, such as the macroporous
networks induced by /?-xylosides or DON. This will have serious consequences
for the viability of the animal, and may, in the case of embryos treated with
/?-xylosides, be partly responsible for the observation that they are more fragile
than control embryos (Gibson et al. 1978).
It is a pleasure to acknowledge the advice and assistance of Dr M. Boublik and Mr F.
Jenkins in performing the electron microscopic studies. We are also grateful to Dr A. Fallon
for the determinations of prolyl hydroxylase activity.
REFERENCES
H. C , CHACKO, S., ABBOTT, J. & HOLTZER, H. (1970). The loss of phenotypic
traits by differentiated cells in vitro. Am. J. Path. 60, 289-311.
AUDHYA, T. K., SEGEN, B. J. & GIBSON, K. D. (1976). Stimulation of proteoglycan synthesis
in chick embryo sternum by serum and L-3,5,3'-triiodothyronine. /. biol. Chem. 251,
3763-3767.
AYDELOTTE, M. B. & KOCHHAR, D. M. (1975). Influence of 6-diazo-5-oxo-L-norleucine
(DON), a glutamine analogue, on cartilaginous differentiation in mouse limb buds in vitro.
Differentiation 4, 73-80.
BARROW, M. V., SIMPSON, C. F. & MILLER, E. J. (1974). Lathyrism: A review. Q. Rev. Biol.
49, 101-128.
BITTER, T. & MUIR, H. (1962). A modified uronic acid carbazole reaction. Analyt. Biochem. 4,
330-334.
BURTON, K. (1956). A study of the conditions and mechanism of the diphenylamine reaction
for the colorimetric estimation of deoxyribonucleic acid. Biochem. J. 62, 315-323.
COMPER, W. D. & LAURENT, T. C. (1978). Physiological function of connective tissue polysaccharide. Physiol. Rev. 58, 255-315.
FESSLER, J. H. (1957). Water and mucopolysaccharide as structural components of connective
tissue. Nature, Lond. 179, 426-427.
FUKUNAGA, Y., SOBUE, M., SUZUKI, N., KUSHIDA, H., SUZUKI, S. & SUZUKI, S. (1975).
Synthesis of a fluorogenic mucopolysaccharide by chondrocytes in cell culture with 4-methyl
umbelliferyl /?-D-xyloside. Biochim. biophys. Acta 381, 443-447.
FURTHMAYR, H. & TIMPL, R. (1971). Characterization of collagen peptides by sodium dodecylsulfate-polyacrylamide electrophoresis. Analyt. Biochem. 41, 510-516.
GALLIGANI, L., HOPWOOD, J., SCHWARTZ, N. B. & DORFMAN, A. (1975). Stimulation of
synthesis of free chondroitin sulfate chains by /?-D-xylosides in cultured cells. /. biol. Chem.
250, 5400-5406.
GHOSH, S., BLUMENTHAL, H. J., DAVIDSON, E. & ROSEMAN, S. (1960). Glucosamine metabolism. V. Enzymatic synthesis of glucosamine-6-phosphate. / . biol. Chem. 235, 1265-1273.
GIBSON, K. D., DOLLER, H. J. & HOAR, R. M. (1978). /?-D-Xylosides cause abnormalities of
growth and development in chick embryos. Nature, Lond. 173, 151-157.
GIBSON, K. D. & SEGEN, B. J. (1977). The mode of action of 4-methylumbelliferyl /?-Dxyloside on the synthesis of chondroitin sulfate in embryonic chick sternum. Biochem. J.
168, 65-79.
GIBSON, K. D., SEGEN, B. J. & DOLLER, H. J., Jr. (1979). Changes in chemical composition
of chick embryos treated with a /?-xyloside and a lathyrogen. Teratology 19, 345-356.
ANDERSON,
p-D-Xyloside and cartilage in ovo
201
R. M. & KOCHHAR, D. M.(1975). Limb development in mouse embryos: Protection
against teratogenic effects of 6-diazo-5-oxo-L-norleucine (DON) in vivo and in vitro.
J. Embryol. exp. Morphol. 33, 355-370.
HAMBURGER, V. & HAMILTON, H. L. (1951). A series of normal stages in the development of
the chick embryo. /. Morph. 88, 49-92.
HAMERMAN, D. & SCHUBERT, M. (1962). Diarthroidal joints, an essay. Am. J. Med. 33,
555-590.
HOPWOOD, J. J. & ROBINSON, H. C. (1973). The molecular-weight distribution of glycosaminoglycans. Biochem. J. 135, 631-637.
HUTTON, J. J., Jr, TAPPEL, A. L. & UDENFRIEND, S. (1966). A rapid assay for collagen
proline hydroxylase. Analyt. Biochem. 16, 384-394.
KEMPSON, G. E., MUIR, H., POLLARD, C. & TUKE, M. (1973). The tensile properties of the
cartilage of human femoral condyles related to the content of collagen and glycosaminoglycans. Biochim. biophys. Ada 297, 456-472.
KEMPSON, G. E., MUIR, H., SWANSON, S. A. V. & FREEMAN, M. A. R. (1970). Correlation
between stiffness and the chemical constituents of cartilage on the human femoral head.
Biochim. biophys. Ada 215, 70-77.
KIMATA, K., OKAYAMA, M., OOHIRA, A. & SUZUKI, S. (1974). Heterogeneity of proteochondroitin sulfates produced by chondrocytes at different stages of cytodifferentiation.
/. biol. Chem. 249, 1646-1653.
KIVIRIKKO, K. I., LAITINEN, O. & PROCKOP, D. J. (1967). Modifications of a specific assay for
hydroxyproline in urine. Analyt. Biochem. 19, 249-255.
KOCHHAR, D. M., AYDELOTTE, M. B. & VEST, T. K. (1976). Altered collagen fibrillogenesis in embryonic mouse limb cartilage deficient in matrix granules. Expl Cell. Res. 102,
213-222.
KOSHER, R. A., LASH, J. W. & MINOR, R. K. (1973). Environmental enhancement of in vitro
chondrogenesis. IV. Stimulation of somite chondrogenesis by exogenous chondromucoprotein. Devi Biol. 35, 210-220.
KOSHER, R. A. & CHURCH, R. L. (1975). Stimulation of in vitro somite chondrogenesis by
procollagen and collagen. Nature, Lond. 258, 327-330.
LANDAUER, W. (1965). Nanomelia, a lethal mutation of the fowl. /. Hered. 56, .131-138.
LANE, P. W. & DICKIE, M. M. (1968). Three recessive mutations producing disproportionate
dwarfism in mice. / . Hered. 59, 300-308.
LANE, J. M. & WEISS, C. (1975). Review of articular cartilage collagen research. Arthritis
Rheum. 18, 553-562.
LEE-OWN, V. & ANDERSON, J. C. (1976). Interaction between proteoglycan subunit and
Type II collagen from bovine nasal cartilage, and the preferential binding of proteoglycan
subunit to Type I collagen. Biochem. J. 153, 259-264.
LOWRY, O. H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. /. biol. Chem. 193, 265-275.
MARTIN, A. V. W. (1954). An electron microscope study of the cartilaginous matrix in the
developing tibia of the fowl. /. Embryol. exp. Morphol. 2, 38-48.
MATUKAS, V. J., PANNER, B. J. & ORBISON, J. L. (1967). Studies on ultrastructural identification and distribution of protein-polysaccharide in cartilage matrix. /. Cell Biol. 32,
365-377.
MILLAR, J. R., SMITH, D. G. & KRESSMAN, T. R. E. (1965). Solvent-modified polymer
networks. Part IV. Styrene-divinylbenzene copolymers made in the presence of nonsolvating diluents. J. Chem. Soc. 1965, 304-310.
MILLAR, J. R., SMITH, D. G., MARR, W. E. & KRESSMAN, T. R. E. (1963). Solvent-modified
polymer networks. Part I. The preparation and characterisation of expanded-network and
macroporous styrene-divinyl-benzene copolymers and their sulphonates. /. Chem. Soc.
1963, 218-225.
MILLER, E. J. (1976). Biochemical characteristics and biological significance of the geneticallydistinct collagens. Mol. Cellular Biochem. 13, 165-192.
MILLER, E. J. (1977). The collagen of joints. In The Joints and Synovial Fluid (ed. L. Sokoloff),
pp. 205-242. New York: Academic Press.
GREENE,
202
J. T. HJELLE AND K. D. GIBSON
D. M., Jr (1971). Molecular weight determination of protein-dodecyl sulfate
complexes by gel electrophoresis in a discontinuous buffer system. J. biol. Chem. 246,
6328-6334.
OGSTON, A. G. (1958). The spaces in a uniform random suspension of fibres. Trans. Faraday
Soc. 54, 1754-1757.
OGSTON, A. G. (1970). The biological functions of the glycosaminoglycans. In Chemistry and
Biology ofthe Intercellular Matrix (ed. E. A. Balazs), pp. 1231-1240. New York: Academic
Press.
OKAYAMA, M., PACIFIC I, M. & HOLTZER, H. (1976). Differences among sulfated proteoglycans
synthesized in nonchondrogenic cells, presumptive chondroblasts, and chondroblasts.
Proc. natn Acad. Sci. U.S.A. 73, 3224-3228.
ORKIN, R. W., PRATT, R. M. & MARTIN, G. R. (1976). Undersulfated chondroitin sulfate in
the cartilage matrix of brachymorphic mice. Devi Biol. 50, 82-94.
PALMOSKI, M. J. & GOETINCK, P. F. (1972). Synthesis of proteochondroitin sulfate by normal,
nanomelic, and 5-bromodeoxyuridine-treated chondrocytes in cell culture. Proc. natn
Acad. Sci. U.S.A. 69, 3385-3388.
PENNYPACKER, J. & GOETINCK, P. F. (1976). Biochemical and ultrastructural studies of
collagen and proteochondroitin sulfate in normal and nanomelic cartilage. Devi Biol. 50,
35-47.
ROMANOFF, A. L. (1960). The Avian Embryo, pp. 1005-1007. New York: Macmillan Publishing Co.
SAJDERA, S. W. & HASCALL, V. C. (1969). Protein polysaccharide complex from bovine nasal
cartilage. A comparison of low and high shear extraction procedures. /. biol. Chem. 1M,
77-87.
SOKOLOFF, L. (1969). The Biology of Degenerative Joint Disease. Chicago: University of
Chicago Press.
TANZER, M. L. (1976). Cross-linking. In Biochemistry of Collagen (ed. G. N. Ramachandran
& A. H. Reddi), pp. 137-162. New York: Plenum Press.
TELSER, A., ROBINSON, H. C. & DORFMAN, A. (1965). The biosynthesis of chondroitinsulfate protein complex. Proc. natn Acad. Sci. U.S.A. 54, 912-919.
TOOLE, B. P. (1976). Binding and precipitation of soluble collagens by chick embryo cartilage
proteoglycan. J. biol. Chem. 251, 895-897.
TRELSTAD, R. L., KANG, A. H., TOOLE, B. P. & GROSS, J. (1972). Collagen heterogeneity.
High resolution separation of native [al(I)2a2] and [al(II)]3 and their component a chains.
/. biol. Chem. 247, 6469-6473.
WHITEMAN, P. (1973). The quantitative measurement of alcian blue-glycosaminoglycan
complexes. Biochem. J. 131, 343-350.
NEVILLE,
(Received 6 February 1979, revised 22 May 1979)