different usage of the glycosaminoglycan-attachment sites of
TRANSCRIPT
Different Usage of the Glycosaminoglycan-Attachment Sites of Biglycan∗
Hans Kresse1, Daniela G. Seidler1, Margit Müller1, Egon Breuer1, Heinz Hausser1,2,
Peter J. Roughley3, and Elke Schönherr1
1Institute of Physiological Chemistry and Pathobiochemistry, University of Münster,
Germany, and 3Shriners Hospital for Children, McGill University, Montreal, Canada
Running title: Glycosaminoglycan Substitution of Biglycan
Correspondence: Dr. Hans Kresse, Institute of Physiological Chemistry and Pathobiochemistry, Waldeyerstr. 15, D-48149 Münster, Germany Phone: +49-251-8355581 Fax: +49-251-8355596 E-mail: [email protected]
∗ This work was supported in part by the Deutsche Forschungsgemeinschaft (SFB 496, Project A6) 2 Present address: Department of Orthopedics, University of Ulm, Germany
Copyright 2001 by The American Society for Biochemistry and Molecular Biology, Inc.
JBC Papers in Press. Published on January 5, 2001 as Manuscript M009321200 by guest on A
pril 16, 2018http://w
ww
.jbc.org/D
ownloaded from
2
Synopsis
Biglycan is a member of the small leucine-rich proteoglycan family. Its core protein
comprises two chondroitin/dermatan sulfate attachment sites on serine-42 and serine-47,
respectively, which are the fifth and tenth amino acid residue after removal of the prepro
peptide. Since the regulation of glycosaminoglycan chain assembly is not fully understood
and because of the in vivo existence of monoglycanated biglycan, mutant core proteins were
stably expressed in human 293 and Chinese hamster ovary cells, in which i) either one or both
serine residues were converted into alanine or threonine residues, ii) the number of acidic
amino acids N-terminal of the respective serine residues was altered, and iii) an hexapeptide
was inserted between the mutated site 1 and the unaltered site 2. Labeling experiments with
[35S]sulfate and [35S]methionine indicated that serine-42 was almost fully used as
glycosaminoglycan attachment site regardless of whether site 2 was available or not for chain
assembly. In contrast, substitution of site 2 was greatly influenced by the presence or absence
of serine-42, although additional mutations demonstrated a direct influence of the amino acid
sequence between the two sites. When site 2 was not substituted with a glycosaminoglycan
chain, there was also no assembly of the linkage region. These results indicate that
xylosyltransferase is the rate-limiting enzyme in glycosaminoglycan chain assembly and
implicate a cooperative effect on the xylosyl transfer to site 2 by xylosylation of site 1 which
is probably becoming manifest before the removal of the propeptide. It is shown additionally
that biglycan expressed in 293 cells may still contain the propeptide sequence and may carry
heparan sulfate chains as well as sulfated N-linked oligosaccharides.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
3
Introduction
The biosynthesis of proteoglycans is a complex process where in the case of
chondroitin and heparan sulfate proteoglycans the assembly of the glycosaminoglycan side
chains begins with the transfer of xylose from UDP-xylose to the hydroxyl group of selected
serine residues of the core protein (1). A recognition consensus sequence for the attachment
of xylose residues may consist of two acidic amino acids closely followed by the tetrapeptide
Ser-Gly-Xaa-Gly, where Xaa is any amino acid (2-4). However, a sequence comparison of
several proteoglycan core proteins indicated that the second glycine residue was present in
only 19 % of the cases, and there was also only an about 50 % abundance of acidic residues
among the three amino acids preceding the glycosaminoglycan attachment site (5).
Furthermore, a Gly-Ser-Ala sequence instead of the Xaa-Ser-Gly sequence is found at the
chondroitin sulfate attachment site of chicken type IX collagen (6), and threonine
substitutions of serine residues can serve as acceptors for chain assembly albeit at low
efficiency (7). The situation is additionally complicated by the fact that established
attachment sites are not necessarily used regularly, giving rise to so-called part-time
proteoglycans. This has been considered as a result of the possibility that the attachment site
is not always recognized by xylosyltransferase and that there may be competition for substrate
between xylosyltransferase and glycoprotein N-acetylgalactosaminyltransferase (8).
Furthermore, in the glycosaminoglycan-free form of the part-time proteoglycan
thrombomodulin the respective serine residue is substituted with the linkage region
tetrasaccharide GlcAβ1-3Galβ1-3Galβ1-4Xyl (9), raising the possibility that it is the transfer
of the fifth monosaccharide (either GalNAc or GlcNAc) which represents the most critical
step in glycosaminoglycan chain biosynthesis.
Considering the complexity of glycosaminoglycan chain assembly we have focussed
on the biosynthesis of biglycan. Biglycan belongs to the family of small leucine-rich
proteoglycans (see 10-12 for recent reviews) and carries most often two chondroitin/dermatan
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
4
sulfate chains which are linked to serines 42 and 47 of the full length human core protein
sequence (13). The sequence around these serine residues (marked by an asterisk) is
DEEAS*GADTS*GVLD. However, usage of the two sites is normally not complete (14), and
up to 30 % of the core protein may be substituted with a single chain only (15). Biglycan
contains a prepro sequence of 38 amino acids. The pro-peptide whose function is not fully
known is not necessarily removed prior to or after secretion (16). Additional proteolytic
processing may take place after removal of the pro-peptide accounting for non-glycanated
forms of the molecule (17,18). The presence of only two glycosaminoglycan attachment sites
which are located in close proximity made biglycan an ideal candidate to study
glycosaminoglycan chain assembly in a still simple, yet somewhat more complex system than
in case of the monoglycanated small proteoglycan decorin where glycosaminoglycan
biosynthesis has been studied in great detail (2,7). It will be shown that the attachment of the
two chains is regulated differently.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
5
EXPERIMENTAL PROCEDURES
Expression of Recombinant Biglycan in Human 293, Chinese Hamster Ovary and
COS-7 Cells - Total RNA was isolated from human MG-63 osteosarcoma cells which are
known to produce high quantities of biglycan. It was subjected to reverse transcription with
AMV reverse transcriptase and an oligo-(dT)15 primer. Then, the complete biglycan coding
sequence was amplified by the Expand High Fidelity PCR system (Roche Diagnostics; PCR
conditions: denaturation for 1 min at 94 °C, annealing for 1 min at 58 °C and elongation for 2
min at 72 °C) using the forward primer 5'-GTTGGAATTCGAGTAGCTGCTTTCGGTCC-3'
and the reverse and complement primer 5'-ATAAGAATGCGGCCGCCAAGGCT
GGAAATG-3'. The primers introduce an EcoRI- and a NotI-restriction site, respectively, near
the 5' and 3' end of the amplified sequence. Upon EcoRI/NotI digestion the purified
amplification product was ligated into the EcoRI/NotI digested pcDNA3.1(+) vector
(Invitrogen). The plasmid was isolated after transformation of competent E. coli TOP 10F'.
Biglycan mutations were introduced by using the QuikChange Site Directed
Mutagenesis Kit (Stratagene) and HPSF quality oligonucleotides (MWG Biotech) as primers.
The method uses forward and reverse primers which carry the base exchange in the center of
their sequence and anneal to identical positions of the plasmid. High fidelity Pfu Turbo DNA
polymerase was used for amplification. The PCR conditions were denaturation for 30 sec at
94 °C, annealing for 1 min at 55 °C, and elongation for 14 min at 68 °C: After 30 PCR cycles
the methylated template was degraded by DpnI, and XL1-Blue supercompetent cells were
used for transformation without prior ligation. The forward primers used to introduce
mutations of the wild-type sequence were as follows. BGN S42A:
5'-GATGAGGAAGCTGCGGGCGCTGACAC-3'; BGN S42T: 5'-CGATGAGGAAGC
TACGGGCGCTGACAC-3'; BGN S47A: 5'-CGCTGACACCGCAGGCGTCCTGGAC-3'.
The latter primer was also used to mutate BGN S42A for the generation of the double mutant
BGN S42A/S47A. The primer pair 5-AAGGTGCCCAAGGGAGATTTCAGCGGGCTCC-3'
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
6
and 5-GGAGCCCGCTGAAATCTCCCTTGGGCACCTT-3' served for the construction of
the triple mutant BGN S42A/S47A/V178D. Using the BGN S42A construct as template BGN
S42A/A44D was made by using 5'-GAAGCTGCGGGCGATGACACCTCAGG-3' as forward
primer. The BGN S47A construct served as template to generate BGN E39Q/S47A with the
forward primer 5-'CATGATGAACGATCAGGAAGCTTCGGGCG-3'. Sequencing of all
constructs verified the introduction of the desired point mutations and the absence of
mutations in the other parts of the cDNA sequence. Finally, a hexapeptide - IGPQVP - which
is similar to the hexapeptide following the Ser-Gly attachment site in decorin (IGPEVP), was
placed between Gly-43 and Ala-44 of the S42A biglycan mutant. In the first step 5'- and 3'-
portions of the desired sequence were generated by using the primers corresponding to the
vector sequences mentioned above, the reverse and complement primer
5'-GGGCACCTGGGGGCCGATGCCCGCAGCTTCCTCATC-3' (5' portion) and the
forward primer 5'-ATCGGCCCCCAGGTGCCCGCTGACACCTCAGGCGTCC-3' (3'
portion), respectively. The amplified products were purified by agarose gel electrophoresis
and subjected to a second polymerase chain reaction by using the two primers designed from
the multicloning site. The cDNA thus obtained was ligated into the pcDNA3.1 (+) vector after
EcoRI/XbaI digestion. The mutants used are summarized in Fig. 1.
Cultured human 293 kidney cells were transfected with the unmodified pcDNA3.1 (+)
vector or with this vector containing one of the above mentioned constructs by using
LipofectAMINE 2000 (GibcoBRL) according to the instructions of the manufacturer. The
cells were selected for neomycine resistance by adding 750 µg/ml G418 (Life Technologies).
The expression of human decorin cDNA in 293 cells had been described before (19). Wild-
type BGN, BGN S42A and BGN S47A were also used to transfect Chinese hamster ovary
cells and COS-7 cells by an analogous procedure. COS cells were used for experiments two
days after transfection.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
7
Metabolic Labeling and Proteoglycan Isolation - Unlabeled reference biglycan was
purified from the culture media of 293 cells stably transfected with wild-type human biglycan
cDNA. Purification included ammonium sulfate precipitation and anion exchange
chromatography as described (20), except that during the latter step NaCl was replaced by
KCl. Biglycan-containing fractions which were eluted from the DEAE column with 0.6 M
KCl were made 2 M with (NH4)2SO4 and directly applied to a phenyl-Sepharose column
(Sigma; 1 ml gel/50 ml of conditioned medium) equilibrated with 10 mM K2SO4, pH 6.8,
containing 2 M (NH4)2SO4. The column was washed with 3 volumes of starting buffer and
then with 3 volumes, each, of this buffer where the (NH4)2SO4 concentration had been
reduced to 1 M, 0.5 M, and 0.2 M, respectively. Analytically pure biglycan (purity >95%)
was desorbed with water.
Metabolic labeling of subconfluent 293, Chinese hamster ovary and COS cells with
[35S]sulfate (incubation period 24 h) and [35S]methionine (incubation period 8 h) was
performed as described earlier (19). When [35S]methionine-labeling was performed in the
presence of p-nitrophenyl-β-D-xyloside (Sigma), the xyloside was present already during the
1 h preincubation periode in methionine-free medium. For labeling with [1-3H]galactose (ICN
Pharmaceuticals; specific radioactivity 18 Ci/mmol) Waymouth MAB 87/3 medium
containing 4 % fetal calf serum and only 1.4 mM glucose was used. Incubation was for 24 h
in the presence of 14 ml of culture medium with 500 µCi of 3H-radioactivity per 75 cm2
culture flask At the end of the incubations, medium was mixed with proteinase inhibitors and
made 70 % saturated with (NH4)2SO4. After centrifugation, the pellet from one 25 cm2 culture
flask was dissolved in 300 µl 0.1 M Tris/HCl buffer, pH 7.4, containing 1 M NaCl, 0.5 %
Triton X-100, 0.5 % sodium deoxycholate, and proteinase inhibitors. The solution was first
allowed to react with protein A-Sepharose (Sigma; 1.5 mg of dry gel per 25 cm2 flask) which
had been preincubated with preimmune rabbit antisera as described (15). Biglycan was
subsequently immune precipitated by applying 6 mg of protein A-Sepharose per 25 cm2 flask
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
8
which had been coated with a monospecific rabbit antiserum against a KLH-coupled
biglycan-derived peptide raised analogously as described (13).
Enzyme Treatment - Protein A-Sepharose-bound immune complexes were subjected to
digestions with chondroitinases ABC and ACII (20) and/or heparinases I and III (21) as
described. Treatment with peptide: N-glycosidase F was for 20 h at 37 °C in 60 µl of sodium
phosphate buffer, pH 7.5, containing 0.3 mM phenylmethylsulfonyl fluoride, 3.3 mM EDTA,
0.4 % (v/v) Nonidet P-40, 0.06 % (w/v) SDS, 0.3 % (v/v) 2-mercaptoethanol, and 500 units of
enzyme (New England Biolabs). After enzyme treatment, the protein A-Sepharose-bound
material was washed twice with 10 mM Tris/HCl, pH 6.8, prior to SDS-PAGE. Use of
[35S]methionine-labeled proteoglycan established the complete removal of the
glycosaminoglycan chains and the N-linked oligosaccharides by this methodology.
Other Methods - SDS-PAGE followed either by fluorography or by Western blotting
was performed as described previously (20,22). Fluorograms were evaluated with the
ImageQuant 5.0 software program (Amersham Pharmacia Biotech). Western blots were
probed with antisera against biglycan (see above), biglycan propeptide (16) and the HNK-1
carbohydrate epitope (a kind gift of Dr. M. Schachner, University of Hamburg, Germany).
Automated protein sequencing was performed on a Procise 492-01 gas phase sequencer (PE
Biosystems). A search for tyrosine O-sulfation was attempted after hydrolysis of core proteins
under nitrogen in 1 M NaOH for 24 h at 110 °C (23).
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
9
RESULTS
Properties of recombinant biglycan synthesized by 293 cells - Stably transfected 293
cells synthesized within 24 h about 0.5 pg of biglycan core protein per cell. Surprisingly,
however, its glycosaminoglycan chains were not completely sensitive towards degradation by
chondroitin ABC lyase. Between 5 and 15 % of total [35S]sulfate incorporated into biglycan
remained in a monoglycanated form. The same observations were made with Chinese hamster
ovary cells. The label could completely be removed by treatment with heparinases I and III
(Fig. 2), indicating that a subset of biglycan, synthesized by these cells, is a hybrid
proteoglycan. However, glycosaminoglycan-free core protein was not observed when
biglycan was treated with heparinases only. Experiments with [35S]methionine-labeled
biglycan corroborated these results (data not shown). Thus, only when two
glycosaminoglycan chains were present, one of the chains could be composed of heparan
sulfate.
It was also investigated whether or not the secreted proteoglycan is still carrying the
propeptide sequence. By using propeptide-specific antibodies the presence of such a sequence
could be verified in Western blots of biglycan purified from 293 cells (Fig. 3). However,
amino acid sequencing revealed that at the most 15 % of the total quantity of secreted core
protein contained the propeptide sequence which comprised amino acids Phe19 - Asn37.
Different Use of Glycosaminoglycan Attachment Sites - For studying the structural
prerequisites for glycosaminoglycan chain attachment to serines 42 and 47, respectively, of
biglycan core protein, a series of mutations was created which is summarized in Fig. 1. In a
first series of experiments the serine residues were replaced by alanine and used for studying
biglycan biosynthesis in human 293 cells. Complementary experiments were performed with
Chinese hamster ovary and COS cells, respectively. The consequences for glycosaminoglycan
chain attachment turned out to be different depending on which one of the two serine residues
was mutated. In the S47A mutant biglycan core protein became substituted with a single
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
10
glycosaminoglycan chain, which was to be expected (shown for 293 cells in Fig. 4). In
contrast, in the S42A mutant about 80 % of the core protein was secreted into the culture
medium without being substituted with a glycosaminoglycan chain. This percentage did not
differ between the three cell types under study. The glycosaminoglycan chain-bearing species
exhibited a somewhat slower and a less heterogeneous migration pattern during SDS-PAGE
than the monoglycanated proteoglycan resulting from the S47A mutation. Such a migration
behaviour is best explained by assuming that the glycosaminoglycan chain of the S42A
mutant is somewhat larger and less heterogeneous in size than in the other mutant. As an
alternative to alanine, serine-42 was also replaced by threonine. In this mutant, too, the
majority of serine-47 residues remained unused (Fig. 5). In the double mutant, S42A/S47A,
an alternative use of other potential glycosaminoglycan attachment sites could not be
observed. As will be discussed below, the glycosaminoglycan-free, but still N-glycanated core
proteins were nevertheless substituted with [35S]sulfate residues.
The proposed consensus sequence for glycosaminoglycan chain attachment is fitted
better for use of serine-42 than of serine-47 (Fig. 1). Further constructs were, therefore,
generated to optimise the use of the latter site by introducing an additional acidic residue and
to make the first site less efficient by removal of an acid residue. Serine-42 was always fully
substituted by a glycosaminoglycan chain regardless of whether or not a preceding acidic
residue had been exchanged against a neutral one. Use of serine-47, however, could
remarkably be stimulated, although complete substitution was never approached (Figs. 5 and
6). The results of three or more separate incubations are summarized in Table I.
From the data described so far it appeared that use of serine-47 was markedly
dependend on a substitution of serine-42. To study this point further, a hexapeptide was
inserted in the S42A mutant in front of the serine-47 residue. This novel mutant protein, too,
was secreted normally. However, [35S]sulfate incorporation into a chondroitin ABC lyase-
sensitive glycosaminoglycan chain was just above the limit of detection (Fig. 7) whereas the
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
11
secretion of the glycosaminoglycan-free core protein remained unaltered. Similarly, attempts
to create a new glycosaminoglycan attachment site at serine-180 in the S42A/S47A double
mutant by inserting aspartate in position 178, were unsuccessful (result not shown). This
negative result should be seen in the context that of all unused Ser-Gly-sites serine-180 should
have the greatest distance from N-glycosylation sites and - in analogy to the proposed horse-
shoe structure of decorin (24) - should be located at the unrestricted convex site of the
molecule.
Absence of Linkage Region Saccharides on Serine-47 in the S42A Mutant - The
majority of serine-47 residues were not substituted with a full-length glycosaminoglycan
chain in the S42A mutant. The substitution of this residue with a linkage region
tetrasaccharide, however, would have escaped detection. To study this possibility 293 cells
harbouring the S42A mutation were, therefore, incubated with a high quantity of [1-
3H]galactose for 24 h. The medium was then passed over a DEAE-Trisacryl M column, and
unbound biglycan core protein was recovered by immune precipitation. The immune complex
was exhaustively digested with endoglycosidase F to remove all asparagine-bound
oligosaccharides. No 3H-radioactivity remained associated with the immune complex. As a
control for a proteoglycan with a single glycosaminoglycan chain, 293 cells transfected stably
with human decorin cDNA were used and labeled under identical conditions. In this case, the
whole medium was subjected to decorin immune precipitation followed by digestion of the
immune complex with chondroitin ABC and ACII lyases and then with endoglycosidase F.
The immune complex still contained 1920 c.p.m. of 3H-radioactivity. It was therefore
concluded that serine-47 was either substituted with a full length glycosaminoglycan chain or
was devoid of a galactose-containing linkage region oligosaccharide. This conclusion was
further supported by the observation of a slightly different migration behaviour during SDS-
PAGE of the unsubstituted S42A core protein and the S42A core protein obtained by
chondroitin ABC lyase treatment.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
12
In further experiments competition for glycosaminoglycan chain assembly by exogeneously
added p-nitrophenyl-β-xyloside was studied. Neither in the S42A nor in the S47A mutant was
the use of the remaining glycosaminoglycan chain-attachment site influenced by increasing
doses of the competitor. However, shorter glycosaminoglycan chains were assembled at high
xyloside concentrations (Fig. 8). These data suggest that the activity of galactosyltransferases
is not the rate-limiting step during the formation of the linkage region. Hence, it seems likely
that serine-47 was also not substituted with xylose in the glycosaminoglycan-free species of
the S42A mutant.
Core Protein-associated [35S]Sulfate - As seen in Fig. 2 and subsequent figures the
core protein of biglycan could be labeled with [35S]sulfate. Combined digestion with
chondroitin ABC lyase and chondroitin ACII lyase for complete removal of the
glycosaminoglycan chains with the exception of the linkage region tetrasaccharide did not
result in complete desulfation. Since small proteoglycans may carry tyrosine O-sulfate
residues (25,26), biglycan core proteins were subjected to alkaline hydrolysis. Thereafter, all
radioactivity could be removed as insoluble barium salt, thereby excluding the presence of
alkali-stable tyrosine O-sulfate residues. However, the label could completely be removed by
treatment with endoglycosidase F, indicating that sulfate ester may be present on one or both
of the two asparagine-linked oligosaccharides of biglycan (Fig. 9). N-linked oligosaccharides
may carry a terminal 3-sulfated GlcA residue as part of the HNK-1 carbohydrate epitope (27).
Westerns blots with antibodies against this epitope were negative for biglycan core protein
although proteins of higher molecular weight in the media of transfected 293 cells gave a
positive immune reaction (result not shown). Whether or not asparagine-linked
oligosaccharides of biglycan core protein contain sulfated N-acetyllactosamine units (28) was
not investigated further.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
13
DISCUSSION
The main result of this study is the observation that of the two glycosaminoglycan-
attachment sites of biglycan the more C-terminally located one is used nearly completely only
when the first attachment site is available. This result should be considered in light of the high
wild-type biglycan production of transfected 293 and CHO cells. Hence, it appears unlikely
that the data obtained with mutant core proteins simply reflect the limited capacity for
glycosaminoglycan chain assembly of these cells. Furthermore, [35S]sulfate-labeling studies
with p-nitrophenyl-β-xyloside as primer molecule also supported this conclusion.
Monoglycanated biglycan has also been extracted from normal tissue. However, it had not yet
been investigated systematically which of the two glycosaminoglycan-attachment sites had
been used in these cases. In bovine articular cartilage both sites were almost completely
substituted (14) whereas, as anticipated from the present study, the second site was less
frequently used in biglycan from human bone (29). Analogous data were obtained with
chicken decorin which resembles mammalian biglycan in so far as it may also carry two
chondroitin/dermatan sulfate chains being linked to serine-4 and serine-16, respectively. In
the monoglycanated proteoglycan form serine-4 was used exclusively (30). Informations on
the usage of the multiple sites of other chondroitin/dermatan sulfate proteoglycans are not
available.
In a study on the glycosaminoglycan attachment sites of perlecan it had been shown
that the serine residues 65, 71, and 76 could be used individually when the two other sites
were mutated to threonine (31). As in our model, the remaining serine residue did not serve as
acceptor site in all molecules but a sequence-dependent change in the efficiency of usage did
not become evident. A direct comparison of the data with respect to site effectiveness,
however, is difficult because a deletion of 110 amino acids was present in the perlecan
construct eight amino acids C-terminal of serine-76, the last glycanated serine residue. The
existence of core protein-specific pathways may also be envisaged (32).
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
14
Direct biosynthetic labeling of the linkage region together with the results of
competition experiments with p-nitrophenyl-β-xyloside strongly suggested that in the S42A
mutant serine-47 was either linked with a full length glycosaminoglycan chain or not
substituted at all with carbohydrate. This finding indicated that UDP-xylose: core protein
xylosyltransferase (EC 2.4.2.26) is the rate-limiting step in glycosaminoglycan chain
assembly and not - as in case of thrombomodulin - a β-N-acetylgalactosaminyltransferase (9).
Also in other aspects of glycosaminoglycan chain initiation thrombomodulin appears to
exhibit special features, since the attachment consensus sequence is less well preserved and,
most strikingly, different serine residues can be used in different cells (33).
Xylose addition appears to begin in the endoplasmic reticulum and to continue in
intermediate compartments, perhaps including the cis-Golgi, where galactose is added prior to
the assembly of the repetitive disaccharide structures (34-36). For an explanation of the
apparent cooperativity between the usage of the two glycosaminoglycan attachment sites
which are located close to the N-terminus of the mature core protein it is tempting to assume
that the propeptide contributes to optimal recognition of the first serine residue by the
xylosyltransferase. The second serine residue then becomes more easily recognized, although
an independent usage of it may occur and may become facilitated by an appropriate
introduction of an additional acidic amino acid close to its N-terminal site. The compartment
where the propeptide sequence of biglycan and decorin is normally removed has not yet been
directly defined. The recent observation that bone morphogenetic protein-1 (procollagen C-
proteinase) processes probiglycan (37) strongly suggests that this cleavage occurs after
secretion or immediately prior to it. Thus, a regulatory influence of the propeptide on chain
initiation is compatible with the topography of probiglycan processing. The N-terminus of the
propeptide was found to be generated by proteolysis at an unusual cleavage site between Pro18
and Phe19 and not, as often, after a small amino acid (38). The predominant form of biglycan
from the secretions of rat and bovine smooth muscle cells was generated by hydrolysis
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
15
between Ala16 and Leu17. However, an alternative rat propeptide sequence began also with
Phe19, indicating that there might be either cell- and species-specific differences in the
processing of the prepro peptide of biglycan or further degradation of the generated
propeptide by aminopeptidases (39). The functional consequences of these differences are not
yet known.
The present study yielded several additional results. First, the glycosaminoglycan-free
biglycan core protein is transport competent and becomes secreted as the fully N-glycosylated
species. Its extracellular stability, however, may be decreased, considering the finding that the
conformational stability of biglycan depends to a greater extent than in the case of decorin on
the presence of chondroitin/dermatan sulfate chains (40). Second, biglycan as expressed in
293 and CHO cells may be substituted with a single heparan sulfate chain, thereby
representing a hybrid proteoglycan. Recombinant biglycan synthesized by HT-1080 cells also
appears as a hybrid species since its glycosaminoglycan chains cannot completely be removed
by chondroitin ABC lyase (41). The factors which govern the substitution with either two
galactosaminoglycan or a single galactosaminoglycan and an additional glucosaminoglycan
chain and their possible cell-type specificity have not yet been defined. Some experimental
variability was also noted in the present investigation. Third, sulfation of asparagine-bound
oligosaccharides of a chondroitin/dermatan sulfate proteoglycan was shown for the first time.
This sulfate substitution may have escaped detection in previous studies because of its low
relative amount. The functional significance and the detailed structure of the sulfated
oligosaccharides remain to be determined.
Acknowledgement: The expert technical help of Ms. Barbara Liel is greatfully
acknowledged.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
16
REFERENCES
1. Esko, J.D., and Zhang, L. (1996) Curr. Opin. Struct. Biol. 6, 663-670
2. Bourdon, M.A., Krusius, T., Campbell, S., Schwartz, N.B., and Ruoslahti, E. (1987)
Proc. Natl. Acad. Sci. U.S.A. 84, 3194-3198
3. Kearns, A.E., Campbell, S.C., Westley, J., and Schwartz, N.B. (1991) Biochemistry 30,
7477-7483
4. Pfeil, U., and Wenzel, K.W. (2000) Glycobiology 10, 803-807
5. Brinkmann, T., Weilke, C., and Kleesiek, K. (1997) J. Biol. Chem. 272, 11171-11175
6. Huber, S., Winterhalter, K.H., and Vaughan, L. (1988) J. Biol. Chem. 263, 752-756
7. Mann, D.M., Yamaguchi, Y., Bourdon, M.A., and Ruoslahti, E. (1990) J. Biol. Chem.
265, 5317-5323
8. Gerlitz, B., Hassell, T., Vlahos, C.J., Parkinson, J.F., Bang, N.U., and Grinnell, B.W.
(1993) Biochem. J. 295, 131-140
9. Nadanaka, S., Kitagawa, H., and Sugahara, K. (1998) J. Biol. Chem. 273, 33728-33734
10. Hocking, A.M., Shinomura, T., and McQuillan, D.J. (1998) Matrix Biol. 17, 1-19
11. Iozzo, R.V. (1998) Annu. Rev. Biochem. 67, 609-652
12. Iozzo, R.V. (1999) J. Biol. Chem. 274, 18843-18846n
13. Fisher, L.W., Termine, J.D., and Young, M.F. (1989) J. Biol. Chem. 264, 4571-4576
14. Neame, P.J., Choi, H.U., and Rosenberg, L.C. (1989) J. Biol. Chem. 264, 8653-8661
15. Schönherr, E., Beavan, L.A., Hausser, H., Kresse, H., and Culp, L.A. (1993) Biochem. J.
290, 893-899
16. Roughley, P.J., White, R.J., and Mort, J.S. (1996) Biochem. J. 318, 979-984
17. Roughley, P.J., White, R.J., Magny, M.C., Liu, J., Pearce, R.H., and Mort, J.S. (1993)
Biochem. J. 295, 421-426
18. Johnstone, B., Markopoulos, M., Neame, P., and Caterson, B. (1993) Biochem. J. 292,
661-666
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
17
19. Kresse, H., Liszio, C., Schönherr, E., and Fisher, L.W. (1997) J. Biol. Chem. 272, 18404-
18410
20. Hausser, H., Hoppe, W., Rauch, U. and Kresse, H. (1989) Biochem. J. 263, 137-142
21. Hausser, H., and Kresse, H. (1999) Biochem. J. 344, 827-835
22. Hausser, H., Ober, B., Quentin-Hoffmann, E., Schmidt, B., and Kresse, H. (1992) J. Biol.
Chem. 267, 11559-11564
23. Mintz, K.P., Fisher, L.W., Grzesik, W.J., Hascall, V.C., and Midura, R.J. (1994) J. Biol.
Chem. 269, 4845-4852
24. Weber, I.T., Harrison, R.W., and Iozzo, R.V. (1996) J. Biol. Chem. 271, 31767-31770
25. Rauch, U., Hollmann, J., Schmidt, A., Buddecke, E., and Kresse, H. (1988) Biol. Chem.
Hoppe-Seyler 369, 595-600
26. Antonsson, P., Heinegård, D., and Oldberg, Å. (1991) J. Biol. Chem. 266, 16859-16861
27. Voshol, H., van Zuylen, C.W., Orberger, G., Vliegenthart, J.F., and Schachner, M. (1996)
J. Biol. Chem. 271, 22957-22960
28. Torii, T., Fukuta, M., and Habuchi, O. (2000) Glycobiology 10, 203-211
29. Fisher, L.W., Hawkins, G.R., Tuross, N., and Termine, J.D. (1987) J. Biol. Chem. 262,
9702-9708
30. Blaschke, U.K., Hedbom, E., and Bruckner, P. (1996) J. Biol. Chem. 271. 30347-30353
31. Dolan, M., Horchar, T., Rigatti, B., and Hassell, J.R. (1997) J. Biol. Chem. 272, 4316-
4322
32. Wong-Palms, S., and Plaas, A. (1995). Arch. Biochem. Biophys. 319, 383-392
33. Lin, J.-H., McLean, K., Morser, J., Young, T.A., Wydro, R.M., Andrews, W.H., and
Light, D.R. (1994) J. Biol. Chem. 269, 25021-25030
34. Lohmander, L.S., Shinomura, T., Hascall, V.C., and Kimura, J.H. (1989) J. Biol. Chem.
264, 18775-18780
35. Kearns, A.E., Vertel, B.M., and Schwartz, N.B. (1993) J. Biol. Chem. 268, 11097-11104
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
18
36. Vertel, B.M., Walters, L.M., Flay, N., Kearns, A.E., and Schwartz, N.B. (1993) J. Biol.
Chem. 268, 11105-11112
37. Scott, I.C., Imamura, Y. Pappanao, W.N., Troedel, J.M., Recklies, A.D., Roughley, P.J.,
and Greenspan, D.S. (2000) J. Biol. Chem. 275, 30504-30511
38. von Heijne, J. (1985) J. Mol. Biol. 184, 99-105
39. Marcum, J.A., and Thompson, M.A. (1991) Biochem. Biophys. Res. Commun. 175, 706-
712
40. Krishnan, P., Hocking, A.M., Scholtz, J.M., Pace, C.N., Holik, K.K., and McQuillan, D.J.
(1999) J. Biol. Chem. 274, 10945-10950
41. Hocking, A.M., Strugnell, R.A., Ramamurthy, P., and McQuillan, D.J. (1996) J. Biol.
Chem. 271, 19571-19577
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
19
Figure Legends
FIG. 1. Alignment of the glycosaminoglycan chain-attachment consensus sequence
with the sequence of mature wild-type and mutant biglycan. Mutated amino acids are
indicated by bold single letters in italics. a designates an acidic and x any amino acid.
FIG. 2. Heparinase-sensitivity of recombinant biglycan. [35S]Sulfate-labeled
biglycan was isolated from 293 cells transfected with wild-type biglycan cDNA and subjected
to treatments with chondroitin ABC lyase (ABC) and/or heparinases I and III (HSase) prior to
SDS-PAGE in a 10 % separation gel followed by fluorography. The migration distance of
reference proteins is indicated on the right margin.
FIG. 3. Identification of biglycan propeptide. Biglycan was purified from the
conditioned media of 293 cells transfected with wild-type biglycan cDNA. After digestion
with chondroitin ABC lyase it was subjected to SDS-PAGE and Western blotting and
immunochemical reaction with a propeptide-specific antibody (lane 1) and an antibody
directed against the mature core protein (lane 2). The prosequence obtained by gas-phase
sequencing is shown in the right panel.
FIG. 4. Glycosaminoglycan substitution of wild-type biglycan and biglycan
mutants S42A, S47A, and S42A/S47A. [35S]Sulfate- and [35SS]methionine-labeled biglycan
was isolated after biosynthetic labeling and subjected to SDS-PAGE and fluorography.
FIG. 5. Glycosaminoglycan substitution of biglycan mutants S42T, S42A/A44D,
and E39Q/S47A as determined by [35S]sulfate incorporation. Secreted biglycan was
obtained by immunoprecipitation from the mutants indicated and from 293 cells transfected
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
20
with wild-type biglycan (WT) or insert-free control vector (CV) prior to SDS-PAGE and
fluorography.
FIG. 6. Glycosaminoglycan substitution of biglycan mutants S42T, S42A/A44D,
and E39Q/S47A as determined by [35S]methionine incorporation. The experiment was
performed analogously as described in Fig. 5. ABC denotes digestion with chondroitin ABC
lyase.
FIG. 7. Further decline of glycosaminoglycan substitution by hexapeptide
insertion into biglycan mutant S42A. Cultures of 293 cells expressing wild-type (WT)
biglycan, biglycan S42A and S42A with a hexapeptide insert were incubated in parallel with
[35S]sulfate prior to SDS-PAGE of secreted biglycan followed by fluorography. Note that
wild-type biglycan was over-exposed.
FIG. 8. Influence of p-nitrophenyl-β -D-xyloside on biglycan biosynthesis. Cells
stably transfected with wild-type biglycan cDNA (WT) or one of the mutant constructs were
preincubated with the xyloside concentrations given in the figure (µM) for 1 h before they
were incubated further for 6 h in the presence of [35S]methionine and the same concentrations
of xyloside. Media were then subjected to immunoprecipitation, SDS-PAGE and
fluorography.
FIG. 9. Sensitivity of core protein sulfation to endoglycosidase F treatment. Cells
transfected with the double mutant S42A/S47A were labeled with either [35S]methionine or
[35S]sulfate before biglycan was obtained by immunoprecipitation and subjected to digestion
with endoglycosidase F prior to SDS-PAGE and fluorography. Note that in this experiment
some biglycan had undergone limited proteolysis of the core protein.
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
21
TABLE I
Composition of glycosaminoglycan - containing biglycan
secreted by 293 cells
wild type
S42A
S42T
S47A
S42A
A44D
E39Q
S47A
% of total
2 GAG chains 1 GAG chain Free core protein
84 - 94
9 - 14
0 - 6
0
15 - 25
75 - 85
0
25 - 35
65 - 75
0
98 - 100
0 - 2
0
50 - 56
44 - 50
0
96 - 100
0 - 4
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
22
FIG. 1.
Consensus sequence
Mature BGN
Consensus sequence
aaxSGxG
||||||
38DEEASGADTSGVL50
|||||
aaxSGxG
GAG KO site 1
GAG KO site 2
GAG KO sites 1 and 2
Site 1, decreased efficiency
Site 2, increased efficiency
Site extension
Novel site V178D in GAG
KO sites 1 and 2
38DEEAAGADTSGVL50
38DEEATGADTSGVL50
38DEEASGADTAGVL50
38DEEAAGADTAGVL50
38DQEASGADTAGVL50
38DEEAAGDDTSGVL50
38DEEAAGIGPEVPADTSGVL56
177GDFSGLRNMN187
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from
Roughley and Elke SchönherrHans Kresse, Daniela G. Seidler, Margit Müller, Egon Breuer, Heinz Hausser, Peter J.
Different Usage of the Glycosaminoglycan-Attachment Sites of Biglycan
published online January 5, 2001J. Biol. Chem.
10.1074/jbc.M009321200Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
by guest on April 16, 2018
http://ww
w.jbc.org/
Dow
nloaded from