Download Spatial and temporal changes in the expression of fibroglycan

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Signal transduction wikipedia , lookup

Cell cycle wikipedia , lookup

Cell growth wikipedia , lookup

Cytokinesis wikipedia , lookup

Endomembrane system wikipedia , lookup

Mitosis wikipedia , lookup

Cell encapsulation wikipedia , lookup

Cell culture wikipedia , lookup

Tissue engineering wikipedia , lookup

Extracellular matrix wikipedia , lookup

Cellular differentiation wikipedia , lookup

Organ-on-a-chip wikipedia , lookup

List of types of proteins wikipedia , lookup

Amitosis wikipedia , lookup

Transcript
841
Development 119, 841-854 (1993)
Printed in Great Britain © The Company of Biologists Limited 1993
Spatial and temporal changes in the expression of fibroglycan
(syndecan-2) during mouse embryonic development
Guido David*, Xiao Mei Bai, Bernadette Van der Schueren, Peter Marynen, Jean-Jacques Cassiman and
Herman Van den Berghe
Center for Human Genetics, University of Leuven, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium
*Author for correspondence
SUMMARY
Fibroglycan (syndecan-2) is a member of a family of cell
surface heparan sulfate proteoglycans that interact with
adhesion molecules, growth factors and a variety of
other effector systems that support the shaping, maintenance and repair of an organism. To investigate this
apparent redundancy of proteoglycans at the cell
surface, we have studied the expression of fibroglycan in
the mouse embryo and compared this expression with
that of syndecan-1. The characterisation of mouse
embryo cDNA clones that crosshybridized to human
fibroglycan-cDNA predicted that murine and human
fibroglycan were highly similar in structure. Consistently, the analysis of transfectant cells, murine cell lines
and embryo extracts indicated that the murine proteoglycan reacted specifically with monoclonal antibody
10H4 developed against the human protein. Fibroglycan,
as detected by monoclonal antibody 10H4 in sections of
embryonic tissues, occurred exclusively on mesenchymal
cells that represented the putative precursors of the hard
and connective tissue cells. No fibroglycan was detected
in epithelia or in muscle cells. Areas where fibroglycan
was particularly abundant were sites of high morphogenetic activity where intense cell-cell and cell-matrix
interactions are known to occur (e.g. the epithelial-mesenchymal interfaces, the prechondrogenic and preosteogenic mesenchymal condensations). The expression
of fibroglycan was weak in the early embryo, culminated
during the morphogenetic phase and at the moment of
cell lineage differentiation, and persisted in the perichondrium, periosteum and connective tissue cells.
Syndecan-1, in contrast, was primarily detected in
epithelia, and transiently in some mesenchymal cells,
with mesenchymal localisations that did not or only
partially overlap with those of fibroglycan. In situ
hybridization analyses confirmed these expression
patterns at the transcriptional level, identifying mesenchymal cells as the major source of fibroglycan production. These data indicate that the expression of fibroglycan occurs along unique and developmentally
regulated patterns, and suggest that fibroglycan and
syndecan-1 may have distinctive functions during tissue
morphogenesis and differentiation.
INTRODUCTION
syndecan-1 (Saunders et al., 1989; Mali et al., 1990) and
fibroglycan (Marynen et al., 1989; Pierce et al., 1992), and
two additional proteoglycans (Gould et al., 1992; Carey et
al., 1992; David et al., 1992b; Kojima et al., 1992) that are
type I integral membrane proteins. The other category
includes forms like glypican (David et al., 1990) that are
linked to the cell surface through the intermediate of a
glycosyl phosphatidylinositol anchor. The four transmembrane proteoglycans show little structural similarity to
glypican, and their extracellular moieties differ also substantially from each other. Yet, the transmembrane and cytoplasmic domains of these proteins are nearly 65% identical,
suggesting that together they may constitute a family of
integral membrane proteoglycans, the syndecans, with
similar domain organization and shared structures,
metabolic fates and functions (Bernfield et al., 1992).
Controlled cell proliferation, the establishment of differential and selective cellular adhesions, and the stabilization of
the generated forms and associations through the deposition
and remodelling of the extracellular matrix are fundamental
to the development of an organism. Cell surface heparan
sulfate proteoglycans are potentially important modulators
of these processes by acting as receptors and stabilizers for
matrix components and growth factors, and by accelerating
the rate at which certain proteinases are inactivated by their
specific inhibitors (Gallagher, 1989; Ruoslahti, 1989;
Kjellén and Lindahl, 1991).
These cell surface-associated heparan sulfate proteoglycans comprise several distinct molecular species and belong
to at least two categories. One category is represented by
Key words: fibroglycan, cell surface heparan sulfate proteoglycan,
mouse embryonic development, mesenchyme
842
G. David and others
Syndecan-1, in its proteoglycan form, can bind a variety
of extracellular matrix components (Saunders and Bernfield,
1988; Sun et al., 1989; Elenius et al., 1990), acts as a ligand
for bFGF (Kiefer et al., 1990) and, to a certain extent, binds
antithrombin III with high affinity (Kojima et al., 1992;
Mertens et al., 1992). None of these properties are unique
for syndecan-1 and the functional significance of the
apparent redundancy of proteoglycans at the cell surface is
therefore not understood. Syndecan-1, however, shows
specific expression patterns and remarkable changes in
expression during embryonic development (Thesleff et al.,
1988; Solursh et al., 1990; Trautman et al., 1991; Sutherland et al., 1991) and woundhealing (Elenius et al., 1991).
Syndecan-1 is therefore thought to function as a matrix and
growth factor receptor, whose expression is developmentally regulated (Bernfield and Sanderson, 1990). Fibroglycan (syndecan-2 in the nomenclature proposed by Bernfield
et al., 1992), on the other hand, shares at least some ligands
with syndecan-1 (Mertens et al., 1992) but may have alternative expression patterns, since it was originally identified
as a major cell surface heparan sulfate proteoglycan of
human lung fibroblasts, which express rather low levels of
syndecan-1 (Lories et al., 1989; 1992). Possibly, the
different cell surface proteoglycans have similar functional
properties but operate in different contexts.
As part of initial approaches to test whether syndecan-1,
fibroglycan and other related proteoglycans might be
engaged in different morphogenetic processes or assume
different tasks in these processes, we have isolated cDNA
clones coding for murine fibroglycan and investigated the
expression of this proteoglycan in murine cells and
embryonic tissues. The expression of fibroglycan was found
to be limited to mesenchymal tissues, to undergo important
changes during development, and to ‘complement’ the
expression of syndecan-1, as the latter occurred mainly in
epithelial tissues, over other parts of the mesenchyme and
during other developmental stages.
MATERIALS AND METHODS
Cell culture
The NMuMG (ATCC #CRL 1636), L-M(TK−) (ATCC #CLL 1.3),
and Swiss 3T3 (ATCC #CCL 92) cell lines were cultured in
Dulbecco’s modified Eagle’s medium (Gibco Europe) containing
10% foetal calf serum as described before (David and Van den
Berghe, 1985). CHO-K1 (ATCC #CCL 61) transfectants expressing recombinant fibroglycan and controls were cultured in media
containing charcoal-treated serum and supplemented with or
without 10 ng/ml of dexamethasone.
Characterisation of the 10H4 epitope in mouse embryos
15-day-old mouse embryos were suspended in 4 M GdnHCl buffer
supplemented with proteinase inhibitors (Lories et al., 1989), triturated by gentle pipetting, and extracted for 24 hours in this buffer.
The extracts were cleared by ultracentrifugation (Beckman
SW60Ti rotor, 40000 rpm, 90 minutes), and fractionated by gelfiltration over Sepharose Cl-4B in 4 M GdnHCl buffer with or
without detergent, and by ion-exchange chromatography over
Mono Q in TUT buffer (0.5% Triton X-100, 6 M urea, 50 mM
Tris-HCl, pH 8.0), essentially as described before (Lories et al.,
1987). The elution of the 10H4 epitope was monitored by
immunodot-blotting using 125I-iodinated mAb 10H4 and autoradi -
ography. The Mono Q fractions that were positive for 10H4 were
reabsorbed on a small volume (50 µl) of DEAE-Trisacryl M beads
and 125I-iodinated as described before (Lories et al., 1987). The
125I-iodinated materials were dialysed against TTBS (0.1% Triton
X-100, 150 mM NaCl, 50 mM Tris-HCl, pH 7.4), and mixed with
mAb 10H4 that had been coupled to CNBr-activated Sepharose 4B
(2 mg/ml). After an overnight incubation at 4˚C, the 10H4Sepharose beads and bound materials were extensively washed
with TTBS and then with heparitinase buffer (Lories et al., 1987).
Half of the washed beads was treated with 1 mIU of heparitinase
for 3 hours at 37˚C, while the other half was left untreated. Finally,
the suspensions of heparitinase-digested and of non-digested
immunobeads were boiled for 10 minutes in sample buffer, and
loaded on a 4-16% SDS-polyacrylamide gradient gel (Lories et al.,
1987). After electrophoresis, the gel was processed for autoradiography as described before (Lories et al., 1987).
Isolation and analysis of murine cDNA clones
Poly(A)+ RNA was isolated from whole 14-day-old mouse
embryos and used as template for the synthesis cDNA as described
before (Marynen et al., 1989). The resulting library was ligated into
λZAP II arms and screened without amplification, using the 32Poligolabeled insert from the cDNA clone 48K5EP, an EcoRI and
PstI-generated subclone (bases 1-1389) of the human fibroglycan
clone 48K5 (Marynen et al., 1989). Five positive phage clones
were plaque purified, converted to Bluescript II SK− plasmids and
further analyzed by restriction mapping, subcloning and sequencing. Both strands of the inserts were sequenced using the dideoxy
chain termination method (Sanger et al., 1977), and using both
dGTP and 7-deaza dGTP (Pharmacia-LKB Biotechnology,
Uppsala, Sweden). All other procedures for molecular analyses
were essentially as described before (Marynen et al., 1989; David
et al., 1990, 1992b).
Expression of recombinant fibroglycan
The 48KM2 plasmid was digested with XbaI and SalI, and the
restriction fragment consisting of the 48KM2 insert and short
flanking sequences derived from the multiple cloning site of the
vector was ligated into a NheI and SalI-restricted pMAMneo vector
(Clontech Laboratories Inc, Palo Alto, CA, USA), yielding the
plasmid pMAM-48KM-neo for the expression of sense mRNA for
murine fibroglycan under the control of the dexamethasoneinducible MMTV-LTR promoter. pMAMneo and pMAM-48KMneo plasmids were propagated in HB101 E. coli cells, linearized
with PvuI, and introduced into CHO cells by electroporation. After
two weeks of selection, G418-resistant cells were tested for dexamethasone-dependent expression of the 10H4 epitope by western
blotting and chemiluminescence (Lories et al., 1989).
In situ hybridisation
Labeled single-stranded cDNA probes for the in situ detection of
the messages for fibroglycan were prepared by two different
methods. In the first approach, 15 ng aliquots of the EcoRI and
HincII-generated fragment of the 48KM4 insert (bases 85 to 862
of the 48KM sequence) were radiolabeled by random oligonucleotide priming using [35S]dCTP, denatured and prehybridized
with a molar excess (500 ng) of unlabeled single-stranded competitor DNA from M13 phages containing either the + or the −
strand of the 48KM4 insert (Boehm et al., 1991). Alternatively,
cDNA corresponding to the part of the 48KM sequence (bases 524958) that codes for the ectodomain of fibroglycan was prepared by
PCR, using the 48KM4 plasmid as a source of DNA and the
oligonucleotides 5′-ATGCAGCGCGCGTGGATCCTGCTCACCTTGGGC and 5′-TTCTGTCGACTAAAACAGACTGTCTGAGTG as sense and antisense primers respectively. The PCR
product was restricted with BamHI and SalI and ligated into a
pGEM3Z plasmid, which had been restricted with the same
Developmental expression of fibroglycan
enzymes to produce the plasmid 48KM-ECTO. A second set of
PCR reactions, using the 48KM-ECTO plasmid as a source of
DNA and the combinations of a biotinylated-T7 primer with the
antisense primer and of a biotinylated-SP6 primer with the sense
primer, yielded double-stranded 48KM-ECTO cDNAs biotinylated
on either the + or the − strand. Approximately 2 µg of these biotinylated PCR products were incubated with 600 µl of Streptavidin
MagneSphere paramagnetic particles (Promega, Madison, WI) and
used to produce [35S]dCTP-labeled single-stranded cDNA probes
(essentially as described by Espelund et al., 1990). Both
approaches were also used to produce labeled sense and antisense
cDNA probes corresponding to the part of the syndecan-1 message
that codes for the ectodomain of the proteoglycan (bases 240-869
of the sequence published by Saunders et al. (1989)). The in situ
hybridization procedure on pronase-treated and refixed sections of
paraffin-embedded tissues was essentially as described by Boehm
(Boehm et al., 1991).
Immunohistochemistry
For light microscopy, 10- to 17-day-old mouse embryos were fixed
in 4% paraformaldehyde in PBS for 3 hours at 4˚C, and embedded
in paraffin. Deparaffinated tissue sections were preincubated in 1%
casein in phosphate-buffered saline, and then incubated for 2 hours
in PBS-casein with or without 50 µg/ml of mAb 10H4. After
extensive washing in PBS-casein, test and control sections were
incubated for 1 hour with horseradish peroxidase-conjugated rabbit
anti-mouse Ig (Dako, Glostrup, Denmark) and stained as described
before (David et al., 1992a). For the developmental stages
examined, all control sections were negative. Staining reactions for
syndecan-1 using the monoclonal antibody 281-2 were essentially
as described by Hayashi et al. (1987).
For immunogold labeling and electron microscopy, the lungs
were dissected from 17-day embryos and fixed in 4%
paraformaldehyde in PBS for 30 minutes at 4˚C and for 30 minutes
at room temperature. Ultracryosections were made from these
tissues and incubated for 1 hour at room temperature in PBS containing 0.5% bovine serum albumin, 1% normal goat serum and 50
µg/ml of either mAb 10H4 or the control mAb 3G10 (David et al.,
1992a). The sections were washed, incubated with goat antimouse-Au (Janssen Pharmaceuticals, Beerse, Belgium) and further
processed for electron microscopy as described before (Heremans
et al., 1989).
843
Fig. 1. Fibroglycan mRNA analysis. Poly(A)+RNA isolated from
14-day mouse embryos (1) and from cultured human foetal lung
fibroblasts (2) was fractionated in denaturing formaldehyde
agarose gels, blotted on a nylon membrane and hybridized to the
human probe 48K5EP (F-H). Total RNA samples from adult
mouse brain (3), liver (4) and lung (5) and from the murine
NMuMG (6), L-M(TK−) (7), and 3T3 (8) cell lines were similarly
fractionated and blotted, and hybridized to the 5′ moiety (up to the
internal EcoRI site) of the murine probe 48KM2 (F-M) or to a
probe for murine syndecan (S-M).
as an initiation codon, fulfilling the need for a purine at
position −3 (Kozak, 1989). The open reading frame was terminated by a TAA stop codon at position 1181. The
presumed 3′ untranslated region contained three AATAAA
polyadenylation signals, but did not terminate in a poly(A)
sequence. Overall, this sequence showed 72% of similarity
to the 48KH5 sequence that codes for human fibroglycan
(Marynen et al., 1989).
Northern analysis of the murine cell lines 3T3, L-M(TK−)
and NMuMG revealed ∼3.5, ∼2.3 and ∼1.2 kb fibroglycan
RESULTS
Murine fibroglycan cDNA clones
Poly(A)+ RNA isolated from whole 14-day-old mouse
embryos was used for the synthesis of cDNA, after northern
analyses had indicated that the human fibroglycan probe
48K5EP was detecting a weak but discrete signal in this
RNA preparation (Fig. 1). This murine cDNA was ligated
into the λZAPII vector, and 1.7×105 independent plaques
were screened with the 48K5EP probe under high stringency
conditions of hybridisation. Five positive clones were
detected and plaque-purified by several rounds of selection.
All clones appeared to contain different parts of the same or
closely related messages. The 5′ moiety of clone 48KM2 (up
to the EcoRI site) and clone 48KM4 were completely
sequenced, following the strategy shown in Fig. 2, yielding
the composite 3311 bp cDNA sequence shown in Fig. 3. The
first ATG codon in this sequence occurred at position 524.
This ATG was preceded by several inframe stop codons and
formed the start of a 606 bp open reading frame. The adenine
at position 521 was consistent with a function of this ATG
Fig. 2. Restriction map and alignment of the murine fibroglycan
clones. The total sizes, coding sequences (boxes) and position of
the polyadenylation signals (triangles) of the murine clones
48KM2 and 48KM4 (solid symbols) are shown next to a
representation of the human fibroglycan cDNA clone 48KH5
(open symbols) and next to a restriction map of the murine
cDNAs. The scale of this map is in kb. P, PstI; S, SacI; B, BamHI;
H, HindIII; C, HincII; E, EcoRI; X, XbaI; V, PvuII; A, AccI. The
arrows illustrate the strategy followed to sequence the 48KM2 and
48KM4 clones.
844
G. David and others
G TGA GAT CGA GGG GGA GGA CCC AGG AGG AGG CGG CGC GTA GAA GGA GGG GCG CTG CAG CCA AGC CAC CAG TCT AGC TGT AGA GTG CTG TA
9 0
C TCC AGA TAC TCC TGG AGC TCC AGC CCC TCA GAC CGC GCG TCC CCG CCG CCG TTC GCC TTT CGG CAT CCT AAA CTT GAG CGG CAG CTC GG
180
T TTC AAC CCA TCG ACT GCT TGC TTC AAA TCA GAC AGC ATT GCG ACC CAG ACA CCG GAG TCC GCC GAG TGA AAG CAC AAC GCT GCC CTG TA
270
G GAC CAG ACC AAG AAA ACA GCC TCG CGA AGC AGC AAC TCT GGG TTG GGA GTA CAG AAG CCA ACA AGT GAG AAG GCG CCG GGT TTC CGG GG
360
C GCA GGG AGA ACG CTA GAG CAG GCG CCA GAG AAG ACA GCT AGA GCT CGG AAT CGG AGC CCA AAC CTC TCT CCT GGA GGC GGC TCA GCT TC
450
T TTT TTT TCT AGG CCC TCT TAG GCT TGC TCT GGG CTT CCT TTT ATC CGG GTA GGA GCC ACA TCC CTG GGG AAT ATG CAG CGC GCG TGG AT
540
M
Q
R
A
W
I
C CTG CTC ACC TTG GGC TTG ATG GCC TGT GTG TCC GCA GAG ACG AGA ACA GAG CTG ACA TCC GAT AAG GAT ATG TAC CTT GAC AAT AGC
630
L
L
T
L
G
L
M
A
C
V
S
A
E
T
R
T
E
L
T
S
D
K
D
M
Y
L
D
N
S
O
C ATT GAG GAA GCT TCA GGA GTA TAT CCT ATT GAT GAT GAT GAC TAT TCT TCT GCC TCA GGC TCA GGA GCT GAT GAA GAC ATA GAG AGT
720
I
E
E
A
S
G
V
Y
P
I
D
D
D
D
Y
S
S
A
S
G
S
G
A
D
E
D
I
E
S
•
•
•
TC
S
CC
P
•
A GTT CTG ACA ACA TCC CAA CTG ATT CCA AGA ATC CCA CTC ACT AGT GCT GCT TCC CCC AAA GTG GAA ACC ATG ACG TTG AAG ACA CAA AG
810
V
L
T
T
S
Q
L
I
P
R
I
P
L
T
S
A
A
S
P
K
V
E
T
M
T
L
K
T
Q
S
C ATT ACA CCT GCT CAG ACT GAG TCA CCT GAA GAA ACT GAC AAG GAG GAA GTT GAC ATT TCT GAG GCA GAA GAG AAG CTG GGC CCT GCT AT
900
I
T
P
A
Q
T
E
S
P
E
E
T
D
K
E
E
V
D
I
S
E
A
E
E
K
L
G
P
A
I
A AAA AGC ACA GAT GTG TAC ACG GAG AAA CAT TCA GAC AAT CTG TTT AAA CGG ACA GAA GTT CTA GCA GCC GTC ATT GCT GGT GGT GTG AT
990
K
S
T
D
V
Y
T
E
K
H
S
D
N
L
F
K
R
T
E
V
L
A
A
V
I
A
G
G
V
I
C GGC TTT CTC TTT GCC ATT TTC CTC ATC CTG CTA TTG GTG TAC CGC ATG CGG AAG AAA GAT GAA GGA AGC TAC GAC CTT GGA GAA CGC AA
1080
G
F
L
F
A
I
F
L
I
L
L
L
V
Y
R
M
R
K
K
D
E
G
S
Y
D
L
G
E
R
K
A CCA TCC AGC GCA GCT TAC CAG AAG GCA CCC ACT AAG GAG TTT TAT GCA TAA AGC TCC CAT TTA GTG TCT CTA TTT AGG AGA GCA CTG AA
1170
P
S
S
A
A
Y
Q
K
A
P
T
K
E
F
Y
A
C TTT TCA AAA TAA AGC TTT GGC ATA GAA TAA TGA AGA TCT TTG TTA TCT GTT TTG TTC ATT ACA GAG CCA TAC TGG CCC TTT AAT GAT GA
1260
G AAT TCC ATT GTA TTT AAA TTT TTT CGT GTA TTT CTT TAG AAT GAC ATA AAA GTA AAA ATC TAA CAT CTG CAG TGT TCT GTG AAT AGC AG
1350
T GGC AAA ATA TTA TGT TAT GAA AAC CCC TGA CAT TCA TGG AAT TGA TTT GAA CAT CTA TGT GCA AAT ACA AAA TGA TTG TGT TTG TCC TC
1440
T GGT TCA AAG ATG ACT GCT ACT CCC CCT TGT CAG CAA GTC TCC AGT TAA CCT TAC TGA GTT GGT CTT CAT TTA TTT ATC TCT TGT CCC TC
1530
T TCT CTA CCC TCC CTT CTT GTC ACC TTT CTT AGA AAC AAA ACC TTA TGC CTT TTG TAG CTG TCA TGG TGC AAT TTG TCT TTG AAT GAT TA
1620
C AAT AAT GGT AAT TTA GTG TAT ATG TGA TTT TTC TCA AAT ATG TTA ACC TTA ACC TCC ACT TTA TAT AAT GTT TTT AAA TAT CAG ACT AT
1710
C CAT TTT ATG CTT GCT TTA AAT TTC ATT ACT TGT AGC TTC AGG CAG ATT TGC AAA GGC AAA TTA ATG TGT AAA ATT GGA TTA TTA CTA CG
1800
A AAC TGT TTA GTC CTA GCT ATC TTC TTT TGG GAG GAT TTG ACA TAA CGG ACA AGC CTC AGC AAA CCC AAA GAT GCT AAC AGT ATT TTG AG
1890
A AGT TGC TGC AGA TTC CTC TGG CCA CTG TAC TTG TTA ATT TCT TGT CAT TTG AAG GTA CGA GAA AGA GTT TAA AGA AAA AAA AAA AAG AT
1980
C AGT TTT TGT TCT TAA CAT GCA TTT AAA TTG CAA ACA TCT TTT TAA GCC TTT GAA GTG CCT ATG ATT CTA TGT AAC TTG TTG CGG ACT GG
2070
T GTT AAT GAG TAT ATA TAA CAG TTT TTA AAA AGT TGG TAT TTT ATA AGC ACA GAC AAT TCT AAT GGT AAC TTT TGT AGT CTT ATG AAT AG
2160
A CAT AAA TTG TAA TTT GGA AAC AAC AAA AAA ACC TGC TGA ATA AAT CAC ATG GCC TAA TAT TGA AAA TGT CAC TGT TAT AAA TTT TGT AC
2250
Fig. 3. Merged sequences of clones 48KM2 and 48KM4 and corresponding predicted sequence of murine fibroglycan. The three
AATAAA polyadenylation sequences are boxed. The potential N-glycosylation ( ) and glycosaminoglycan chain attachment sites ( )
have been indicated. The effective glycosylation of these sites has not been proven. The hydrophobic putative signal peptide (single line)
and transmembrane sequences (double line) have also been indicated. These sequence data are available from EMBL/GenBank/DDBJ
under accession number U00674.
Developmental expression of fibroglycan
messages in the 3T3 and L-M(TK−) cell lines, but yielded
weak signals in the mammary gland epithelial cells (Fig. 1).
Yet, the latter RNA sample gave strong signals for
syndecan-1. RNA extracts from embryonic tissue and adult
mouse liver also contained fibroglycan messages of ∼3.5,
∼2.3 and ∼1.2 kb, but adult brain contained only the 3.5 and
2.3 kb messages, while very little message was found in
adult lung (Fig. 1). All three fibroglycan messages
hybridized to the full-length 48KM4 insert or to the insert
of an 48KM4 subclone spanning from base 85 to base 862
(HincII site) of the 48KM sequence. Inserts ranging from
base 1333 to base 1901 (PstI sites), and from base 2746
(XbaI site) to base 3311, in contrast, hybridized only to the
3.5 and 2.3 or solely to the 3.5 kb messages respectively,
indicating that the three messages differed at least in their
3′ ends (not shown). Three different messages for fibroglycan have also been detected in human foetal lung fibroblasts
(Marynen et al., 1989) and in rat tissues (Pierce et al., 1992)
and shown to result from differential use of alternative
polyadenylation signals.
Sequence of murine fibroglycan
The open reading frames of all five murine cDNAs coded for
a peptide of 202 amino acids with a sequence typical for type
I integral membrane proteins. The murine protein had a
predicted Mr of 22,130, featuring an NH2-terminal hydrophobic signal peptide of 18 amino acids, a presumptive extracellular domain of 127 amino acids, a potential membrane
spanning segment of 25 hydrophobic residues and a COOHterminal cytoplasmic domain of 32 amino acids. The extracellular domain of the protein potentially accommodated up
to three or four glycosaminoglycan chains, featuring three or
four serines (four if A may be substituted for G in the SGXG
consensus sequence) in appropriate sequence contexts (Mann
et al., 1990), and a single N-linked oligosaccharide, and
contained serine-threonine rich sequences that may also be
glycosylated. The predicted sequences of murine and human
fibroglycan were very similar from the first methionine down
to the sites for glycosaminoglycan chain attachment (91%
similarity), less conserved for most of the remainder of the
ectodomain (65% similarity), but completely identical in the
terminal segment of the ectodomain, in the membrane
spanning segment and in the cytoplasmic domain (Fig. 4). The
ectodomains of both proteins also featured a conserved
dibasic sequence in proximity to the junction with the transmembrane segment. A second dibasic sequence present in the
ectodomain of the human protein, however, was not
conserved in murine fibroglycan. Mouse fibroglycan in turn
was 91% identical to rat fibroglycan (Pierce et al., 1992), with
also complete identity of the transmembrane and cytoplasmic
parts of the two rodent versions of the protein (Fig. 4).
Cross-reactivity of murine fibroglycan with mAb
10H4
Monoclonal antibody 10H4, which was raised against
human fibroglycan (Lories et al., 1989), reacted with the
∼48×103 Mr core protein of a heparan sulfate proteoglycan
present in detergent extracts from dexamethasone-treated
CHO cells that had been transfected with a pMAMneo
vector containing the 48KM2 cDNA coding for murine
845
HUMAN FIBROGLYCAN
MOUSE FIBROGLYCAN
RAT
FIBROGLYCAN
MRRAWILLTLGLVACVSAESRAELTSDKDMYLDNSSIEEASGVYPIDDDDY
51
: :::::::::: ::::::.: :::::::::::::::::::::::::::::
MQRAWILLTLGLMACVSAETRTELTSDKDMYLDNSSIEEASGVYPIDDDDY
51
::::::::::::::::::::: ::::::::::: ::::::::.::::::::
MQRAWILLTLGLMACVSAETRAELTSDKDMYLDSSSIEEASGLYPIDDDDY
51
ASASGSGADEDVESPELTTTRPLPKILLTSAA-PKVETTTLNIQNKIPAQT
101
::::::::::.::: :::. .:.: ::::: ::::: :: :
::::
SSASGSGADEDIESPVLTTSQLIPRIPLTSAASPKVETMTLKTQSITPAQT
102
:::::::::: :: :::::::::: ::::: : ::::::::::::: :::
SSASGSGAYEDKGSPDLTTSQLIPRISLTSAA-PEVETMTLKTQSITPTQT
101
KSPEETDKEKVHLSDSERKMDPAEEDTNVYTEKHSDSLFKRTEVLAAVIAG
152
:::::::: : .:. : : ::
: :::::::: ::::::::::::::
ESPEETDKEEVDISEAEEKLGPAIKSTDVYTEKHSDNLFKRTEVLAAVIAG
153
::::::: : .::::::: ::.::::::::::::::::::::::::::::
Fig. 4. Comparison of the human, mouse and rat fibroglycan
sequences. The human and rodent sequences have been aligned
using the Genepro programme. Similar but non-identical amino
acids are indicated by one dot; identical residues by two dots. The
solid triangle points to the dibasic sequence in the ectodomain
which has been conserved in all three proteins. The other symbols
are as in Fig. 3.
fibroglycan. This proteoglycan was nearly undetectable in
the transfectant cells in the absence of dexamethasone and
was absent from control transfectants with or without dexamethasone treatment (Fig. 5). The same antibody reacted
also with the ∼48×103 Mr coreprotein of a liposome-intercalatable heparan sulfate proteoglycan in murine cell lines
where the expression of fibroglycan had been observed at
the RNA level (not shown).
The 10H4 epitope could also be extracted from 15-day
mouse embryos in 4 M GdnHCl buffer. When this extract
was run over Sepharose Cl-4B in 4 M GdnHCl in the
absence of detergent, the 10H4 epitope was excluded from
the column (Fig. 6A). In contrast, after the addition of 0.5%
Triton X-100 to the extract and to the column buffer, most
of the 10H4 epitope was eluted in the included fractions as
a broad peak with a maximum near Kav=0.4, suggesting that,
in the original extract, the 10H4 epitope formed part of
micellar aggregates (Fig. 6A). Upon further purification of
the extract by ion-exchange chromatography over Mono Q
in the presence of detergent, the 10H4 epitope bound quantitatively to the column and was eluted at 0.8 M NaCl (Fig.
6B). Finally, when immunoprecipitated from 125I-iodinated
extracts and analyzed by SDS-PAGE and autoradiography,
the 10H4 epitope ran as a sharp band with an apparent Mr
of ∼48×103 or as a broad smear, depending on whether it
was applied with or without prior heparitinase digestion
(Fig. 6C). These findings indicated that also in mouse
embryonic tissues the epitope for mAb 10H4 behaved as
authentic fibroglycan.
846
G. David and others
Fig. 5. Crossreactivity of mAb 10H4 with recombinant
fibroglycan expressed in transfectant CHO cells. Detergent
extracts from uninduced and from dexamethasone-treated CHO
cells transfected with pMAM-neo (lanes 1-4) or with pMAM48KM-neo (lanes 5-8) plasmids were concentrated on DEAE,
digested with heparitinase or left untreated, fractionated by SDSPAGE, blotted and incubated with mAb 10H4 and an alkaline
phosphatase-conjugated goat anti-mouse second antibody.
Differential expression of fibroglycan during
embryogenesis
The 10H4 antibody was then used for the in situ detection
of fibroglycan during mouse embryonic development.
Neither heparitinase nor chondroitinase ABC pretreatments
of the sections had significant effects on the staining patterns
that were obtained, except for cartilage (see below).
The expression of fibroglycan, as detectable by mAb
10H4, was limited to mesenchymal tissues and their derivatives. Intense reactions occurred in the visceral mesenchyme
(e.g. of lung and gut), which originates from lateral plate
mesoderm (Fig. 7); in cells from the paraxial mesoderm that
lead to the formation of connective and hard tissues (e.g. in
the sclerotome, precartilage and osteogenic tissue; Fig. 8);
and in the head mesenchyme, the branchial cartilage and a
group of cells in the medulla of the adrenal gland, which are
of neural crest origin (Fig. 9). The accumulation of 10H4
epitope was particularly pronounced over areas or cells that
were morphogenetically active and where intense interactions occurred between cells (mesenchymal condensations,
epithelial-mesenchymal interactions), such as in the mesenchyme that surrounded the branching lung epithelium, in
the developing cartilage, bone and teeth, and in the mesenchyme around the vibrissae. Fibroglycan was detectable in
the mesenchymal tissues and at interfaces between mesenchyme and epithelium, but not in the embryonic epithelia.
All epithelia derived from ectoderm and endoderm, and even
the tubular epithelium of the kidney, which is derived from
the intermediate mesoderm, were negative for 10H4.
Neuronal cells, including the cells of the peripheral neural
ganglia, which are of neural crest origin, and muscle cells
were also negative. In mature organs, fibroglycan was found
in moderate concentrations all over the connective tissues.
The staining patterns for fibroglycan underwent important
changes during development, with stain occurring both on
Fig. 6. Extraction and purification of the 10H4 epitope from
mouse embryonic tissues. (A) Extracts in 4M GdnHCL were
fractionated over Sepharose-CL4B in 4M GdnHCL buffer without
detergent and monitored by immunodotblotting with mAb 10H4
( ). The 10H4-positive fractions were pooled, supplemented with
detergent and rechromatographed over a similar column in 4M
GdnHCL buffer with detergent ( ). (B) The pooled 10H4positive fractions from the Sepharose CL-4B eluate were dialysed
and further fractionated by ion-exchange chromatography over
Mono Q. (C) Finally, the 10H4-positive Mono Q fractions were
pooled, radio-iodinated, immunoprecipitated, and analysed by
SDS-PAGE and autoradiography with (+) or without (−)
heparitinase treatment.
cell surfaces and on extracellular matrices, sometimes in a
basement membrane-like pattern. This was most striking
during lung development (Fig. 7). At day 12, the antibody
Developmental expression of fibroglycan
847
Fig. 7. Expression of fibroglycan in mouse embryonic lung. Paraffin (A-D) and ultracryosections (E and F) through the lungs of 12- (A),
13- (B), and 17-day (C-F) mouse embryos incubated with the anti-fibroglycan mAb 10H4. Indirect immunoperoxidase (A-D) and
immunogold (E,F) stainings. At day 12, the staining pattern is almost basement membrane-like, occurring at the interface between the
first epithelial branches and the mesenchyme. At day 13, the stain completely surrounds the mesenchymal cells that are in close contact
with the stalks of the epithelial branches but is still basement membrane-like at the tips of the branches. At day 17, a layer of positive
mesenchymal cells surrounds the primary and secondary branches. Mesenchymal cells near the tips and distant mesenchymal cells are
negative. High magnification shows stain on the surface of the juxtaepithelial mesenchymal cells (m), on matrix fibres and on the reticular
part of the basement membrane (bm), but the surfaces of the epithelial cells (e) are negative (E,F). Bars, 100 µm (A,B,C); 50 µm (D); 0.1
µm (E,F).
left the lung epithelial cell surfaces unstained, and stained
the mesenchymal cells only disparately, moderately and
irregularly, but it yielded a fairly strong demarcation of the
interface between the mesenchyme and the first epithelial
branches (Fig. 7A). From 13 days on and in later stages, the
anti-fibroglycan antibody continued to demarcate the epithelial-mesenchymal interfaces of the major epithelial ducts,
but it also encircled the mesenchymal cells in proximity of
the epithelium, the epithelial tissue itself remaining
negative. Around the expanding tips of the branching respiratory epithelium, however, the fibroglycan staining
remained faint, discontinuous and in a near basement
membrane-like fashion (Fig. 7B). On day 17, finally, the
anti-fibroglycan antibody decorated a layer of cells that surrounded the major ducts and tubules, whereas the epithelial
cells and mesenchymal cells near the tips remained negative
(Fig. 7C,D). Immunogold staining of ultracryosections of
these lungs and electron microscopy revealed gold particles
848
G. David and others
Fig. 8. Temporal and regional differences in fibroglycan expression during the formation of hard and connective tissues. Paraffin sections
of mouse embryonic tissues stained with mAb 140H4 showing fibroglycan in developing vertebrae (left) and limbs (right). The
sclerotomes (s) on day 12 stain lightly for fibroglycan, with a slight accentuation of the stain in the more anterior somites. The
dermatomes (d) are weakly stained, the myotomes (m) negative (A). Distinctive staining occurs in the aggregated mesenchymes of the
vertebral primordia (C). Strong stainings of the intervertebral fissurae (if), with progressively decreasing stainings in the hypertrophic
cartilage of the vertebral bodies (v) of a 17-day embryo. The spinal cord (sc) is negative (E). The mesenchyme of the early limb bud (lb)
at day 11 is weakly stained (B), but at day 17 strong stainings occur in the perichondrial, periosteal (po), periarticular (j), tendinous and
dermal tissues of the digits (D) and foot (F). Indirect immuno-peroxidase staining. Bars, 100 µm (A,C,D,E); 200 µm (B,F).
on amorphous materials that were associated with the
surfaces of the mesenchymal cells, with fibres in the pericellular matrix of these cells and with the reticular part of
the basement membrane of the ductular epithelium, but no
gold occurred over the lamina densa or in association with
the epithelial cell surfaces (Fig. 7E,F). Marked changes in
expression were also observed in cell lineages that were
involved in the formation of the skeleton. Fibroglycan was
undetectable in the undifferentiated somite (before day 10),
but became weakly positive in the sclerotome (between day
11 and 13) (Fig. 8A,B), to be found later in prechondrogenic
mesenchymal condensations that formed the rudiments of
the skeletal structures, e.g. vertebrae and ribs (Fig. 8C,E),
long bones, tendons and ligaments (Fig. 8D,F). In the differentiating cartilage, stain occurred over the perichondrium
and, after unmasking with chondroitinase ABC, over the
chondrocytes as well. The strong immunoreactivity in the
chondroblasts of the perichondrium persisted, but the
staining of the chondrocytes decreased gradually with maturation, despite unmasking attempts. In the spinal column
for example, the staining was mainly found in the intervertebral fissure and in the cranial and caudal parts of the
Developmental expression of fibroglycan
849
Fig. 9. Fibroglycan in neural crest-derived tissues. Paraffin sections through the lower lip and developing vibrissae at days 12, 15 and 17
(A,C,E), the maxillary process at day 15 (B), and the adrenal gland (D) and peripheral neural ganglia (F) at days 13 and 17. In the jaw
mesenchyme, fibroglycan accumulates at the interface with the ectoderm and around the invaginating buds and developing vibrissae.
Strong stainings occur in the dental sac and in the flat bone (fb) of the jaw that forms by intramembranous ossification. Fibroglycan is also
expressed by a group of medullar cells (m) in the adrenal gland, but is undetectable in the adrenal cortex (c) and in neural ganglia (ng).
Me, Meckel’s cartilage. Bars, 100 µm.
vertebral segments, while less immunoreactivity persisted in
the vertebral cores (Fig. 8E). Upon the start of the endochondral ossification, strong reactions occurred in the
periosteum (Fig. 8F). During intramembranous ossification,
such as the formation of the mandibula, the 10H4 staining
was hardly detectable on the scattered mesenchymal cells,
became pronounced once the cells aggregated in the
osteogenic core and persisted in the differentiating
osteoblasts (Fig. 9B). The temporal and regional changes in
immunoreactivity were also marked in the mesenchyme that
supports vibrissal development (Fig. 9A,C,D). The mesenchyme of the lip showed little stain before the ingrowth
of the ectoderm but became strongly positive in the areas
around the epidermal buds and intense staining persisted in
the perivibrissal connective tissue.
Finally, the staining patterns obtained for fibroglycan
were clearly distinct from those obtained for syndecan-1
(Fig. 10). In the head, gut, kidney, pancreas and lungs of 12to 17-day mouse embryos, fibroglycan was in the mesenchyme, mostly on cells and structures that were in
proximity with the epithelial tissue. The 10H4 antibody
stained the epithelial- and mesothelial-mesenchymal interfaces in the lung (Fig. 10A) and pancreas (Fig. 10C), the
periepithelial mesenchyme but none of the epithelial structures of the kidney (Fig. 10E), and strongly stained the mesenchyme of the stomach (Fig. 10G). Syndecan-1, in contrast,
850
G. David and others
Fig. 10. Differential expression of fibroglycan and syndecan-1 in mouse embryonic tissues. Paraffin sections of the lungs (A,B), pancreas
(C,D), kidney (E,F) and stomach (G,H) of a 13-day mouse embryo incubated with either the anti-fibroglycan mAb 10H4 (A,C,E,G) or the
anti-syndecan-1 mAb 281-2 (B,D,F,H). Fibroglycan decorates the mesenchymal cells and occurs at the basement membrane interface,
while syndecan-1 shows an epithelial cell surface localisation. Indirect immunoperoxidase staining. Bars, 100 µm.
Developmental expression of fibroglycan
851
prevailed in the epithelial tissues of these organs (Fig.
10B,D,F,H) but, at certain stages, it was also expressed by
mesenchymal cells, e.g. at the level of the developing tooth
(not shown) and transiently in the periductal mesenchyme
of the lung (Fig. 10B). Often these mesenchymal stainings
were not or only partially overlapping with the staining for
fibroglycan. In the lung for example, the mesenchyme near
the secondary branches that stained for syndecan-1 showed
little fibroglycan, whereas the periductular cell layer that had
acquired strong stain for fibroglycan was only poorly stained
for syndecan-1 (compare Fig. 10A,B).
Preferential expression of fibroglycan in mesenchymal
cells and differential expression from syndecan-1 were also
apparent at the transcriptional level (Fig. 11). By in situ
hybridization analyses, strong signals for fibroglycan were
obtained in the peribronchial mesenchyme, in the mesenchyme of the gut, in the mesenchyme of the digits, in prechondrogenic condensations and in perichondria, but no or
weak signals occurred over the epithelia. The signal for
syndecan-1, in contrast, was more intense over the
epidermis, lung and gut epithelium than over the dermal and
visceral mesenchymes. These results for syndecan were in
agreement with previous reports by others (Vainio et al.,
1991; Elenius et al., 1991).
All these data indicated that the expression of fibroglycan
occurred along specific developmental patterns, and in
apparent contrast with that of syndecan-1.
DISCUSSION
Fibroglycan is a structurally conserved integral membrane
heparan sulfate proteoglycan that is related to, but distinct
from, syndecan-1 and at least two other cell surface proteoglycans. In the mouse embryo, fibroglycan is primarily
expressed during middle and late stages of development, and
abounds in mesenchymal cells that form the precursors of
the connective and hard tissues. Its developmental
expression, therefore, contrasts with that of syndecan-1.
Fig. 11. Tissue-specific expression of the messages for fibroglycan.
Detection of the messages for fibroglycan by in situ hybridisation
analysis, using single-stranded antisense cDNA probes for the part of
the message that codes for only the ectodomain of the proteoglycan.
Strong signals are obtained in the 14-day juxtaepithelial lung
mesenchyme (A), the perichondrial tissues in a 17-day embryo (B),
and in the mesenchyme of the digits on day 15 (C). No signal is
obtained over the epithelia. Some of these epithelia, e.g. the
epidermis of the digits, yield a strong signal for syndecan-1 (D). In C
and D, the arrows indicate the position of the epithelial cell layer. In
A,C and D, the probes were made using the method of Boehm et al.
(1991); in B, the method of Espelund et al. (1990). No specific
signals were obtained with the sense probes. Bars, 100 µm.
Murine fibroglycan
Fibroglycan has originally been identified in cultured human
foetal lung fibroblasts, as one of the major heparan sulfate
proteoglycans that are expressed at the surface of these cells
(Lories et al., 1989; Marynen et al., 1989). Molecular
cloning of fibroglycan (Marynen et al., 1989) revealed that
this cell surface component was a type I membrane protein,
and that it shared structural features with the cell surface
proteoglycan syndecan (Saunders et al., 1989). However,
unlike syndecan, this proteoglycan was primarily detectable
in fibroblasts and in non-epithelial cell lines (Lories et al.,
1992), hence the name fibroglycan. Later, syndecan and
fibroglycan have also been referred to as respectively
syndecan-1 and syndecan-2 to distinguish these two proteoglycans from two additional related but distinct proteoglycans, the syndecans 3 and 4, whose structural features and
functions have been reviewed recently (Bernfield et al.,
1992). The cDNA structures that are described here are identified as the murine homologues of the human fibroglycan
(syndecan-2) clones based on their extensive structural similarities with the clones of the human 48K series (Marynen
852
G. David and others
et al., 1989). Moreover, the encoded murine and human
proteins, taking the first ATG codon in both sequences as
the translation initiation sites, are 83% identical in sequence,
and the murine protein cross reacts with antifibroglycan
monoclonals when expressed as ectopic proteoglycan in
transfectant CHO cells.
Developmental regulation of fibroglycan
The immunohistochemical and the in situ hybridization
analyses clearly indicate that the expression of fibroglycan is
developmentally regulated. The antibody reveals large
amounts of fibroglycan on and around certain mesenchymal
cells, but fails to detect this proteoglycan in epithelial cells at
all stages of development. The accumulation in the mesenchymal tissue is lineage and stage dependent. It occurs in
mesenchyme derived from the paraxial and lateral plate
mesoderm and from the neural crest, near and on cells that are
solicited to support the formation of the connective and hard
tissues. The initial accumulations are noted at the interfaces
with the invested epithelial tissues, on the juxtaepithelial cells
and on the cells that condense to form the anlage of the skeletal
structures, and the expression persists on the fibroblast-like
cells of the perichondrium, periosteum and connective tissues,
whereby the proteoglycan truly qualifies as a ‘fibroglycan’.
This interpretation of the histochemical results depends
very heavily on the specificity of the 10H4 antibody for fibroglycan and on the expression and accessibility of the epitope
in all forms of fibroglycan. Evidence for specificity was
obtained from the analysis of the murine cell and embryo
extracts whereby the materials traced and purified with mAb
10H4 were identified as a heparitinase-susceptible membrane
proteoglycan with an apparent relative molecular mass of
∼48×103 for the heparitinase-resistant core, similar to that of
fibroglycan in human lung fibroblasts (Lories et al., 1989).
The transfection experiment in CHO cells on the other hand
indicates that the epitope can be expressed in (these) epithelial cells, ruling out differences in postsynthetic modification
or folding of the core as a trivial explanation for the absence
of the 10H4 epitope from epithelial cells. However, fibroglycan may be the product of three different messages that
differ at least in their 3′-moieties and that have not formally
been proven to be identical in their coding parts. The recombinant protein that expresses the 10H4 epitope and that
qualifies as fibroglycan corresponds to the product of the
∼3.5 kb message, but not necessarily to that of the ∼2.3 and
∼1.2 kb messages, where there is evidence that cells may
differ in the relative abundance of these messages and the
translational efficiencies of these messages in the different
tissues are not known. This issue will only rigorously be
resolved by the isolation of the different polyadenylated
cDNAs and the appropriate transfection studies, but the
above expression patterns and the designation ‘fibroglycan’
must apply to the product of at least one of these transcripts.
The in situ hybridization analyses, which in principle trace
all three message forms equally, restrict specific signals for
these messages to mesenchymal tissues that are stained by
the 10H4 antibody, suggesting transcriptional rather than
post-transcriptional mechanisms for the differences in fibroglycan expression, and rendering other possibilities, such as
differences between epithelial and fibroblastic cells in the
routing or turnover of the proteoglycan, or hiding of the
epitope also unlikely. All current evidence therefore suggests
that in situ the expression of fibroglycan is restricted to mesenchymal derivatives. This conclusion is not inconsistent
with the detection of fibroglycan message and protein in adult
liver and brain (which may be considered as largely epithelial in nature), as there too we found the 10H4 epitope to be
concentrated over the connective tissue of the blood vessels,
portal fields and meninges, and the hepatocytes and neuronal
cells to be negative.
Fibroglycan and tissue morphogenesis
It is striking that during development fibroglycan localizes at
sites where intensive epithelial-mesenchymal interactions
shape and transform the epithelia and the mesenchymes into
the morphologically and functionally differentiated tissues
that characterize the different organs. This is obvious during
the organogenesis of the lung, pancreas and kidney, where the
common and main morphologic event consists of the
branching of an epithelium into a mesenchyme, and whereby
the interactions between epithelium and mesenchyme
stimulate branching and determine the branching pattern
(Kratochwil, 1969). In the developing lung for example, fibroglycan is specifically localized on the surface of a layer of
mesenchymal cells that invests the branching epithelium, and
at the interface between epithelium and mesenchyme. The
interface localization occurs mainly at the tips of the branches.
The cell surface expression, on the other hand, occurs on the
condensed and regularly arranged layer of mesenchymal cells
that surrounds the epithelial stalks and the established branch
points. These specific regional differences in fibroglycan
expression parallel differences in mesenchyme function at
those places, where it is known that the tip epithelium induces
the underlying mesenchyme to condense, and that this condensation in turn limits the overexpansion of the epithelium
and results in cleft formation. One possibility is that the lung
epithelium recruits the lung mesenchyme to express fibroglycan. A precedent for this speculation may be found in tooth,
kidney and limb development, where the transient mesenchymal expression of syndecan-1 is induced by the invested
epithelium (Thesleff et al., 1988; Vainio et al., 1989, 1991;
Solursh et al., 1990). Sustained expression of fibroglycan in
the lung mesenchyme, in turn, could consolidate the
branching pattern through the sequestration of growth factors
and proteinases or by guiding the local assembly of extracellular matrix components that stabilize the branch points. Interestingly, the mesenchymal expression of fibroglycan during
murine lung morphogenesis is very similar to that reported for
the integrin α5 subunit, which is also predominantly
expressed by a layer of spindle-shaped mesenchymal cells
that is circumferentially arranged around the developing
airway ducts (Roman et al., 1991). Overlapping distributions
of integrin and fibroglycan in the mesenchyme suggest that
these two may have cooperative matrix receptor functions that
affect lung development, consistent with the effects of both
synthetic RGD peptide (Roman et al., 1991) and heparin
(Bernfield et al., 1984; Platt et al., 1990) on branching morphogenesis. At the minimum, the expression pattern of fibroglycan seems consistent with earlier proposals that proteoglycans and the mesenchymal aspects of the tissues play a
major role in the morphogenesis of branching epithelia and
of lung in particular (Heine et al., 1990; Smith et al., 1990).
Developmental expression of fibroglycan
Fibroglycan and cytodifferentiation
The mesenchymal cells of the embryo ultimately make up
the hard and connective tissues, muscle cells and blood
vessel system of the mature organs. Hard tissues, cartilage
and bone, differentiate from the mesenchyme of the paraxial
mesoderm and from neural crest cells (cranofacial bone) and
fibroglycan, actively or passively, clearly participates in
their formation. The immunostaining for fibroglycan is
negative in the somite, still weak in the sclerotome, reaches
a maximum in the condensing prechondrogenic cores and
decreases progressively in the differentiating chondrocytes,
while persisting in the perichondrium and in the newly
formed periosteum with the onset of osteogenesis. Similarly,
during the process of intramembranous osteogenesis, the
staining first appears in the condensed mesenchyme and
even intensifies once osteogenesis starts. During the process
of condensation, critical cell-cell and cell-matrix interactions take place that are needed to trigger the differentiation
of the mesenchymal cells. Various effector molecules have
been implicated in these interactions and the elevated
expression of fibroglycan on condensing cells suggests a
possible involvement of this cell surface proteoglycan as
well. The decrease of the staining in chondrocytes while
stain persists in the perichondrium and the periosteum would
then indicate a relationship of fibroglycan to the process of
initiation of cell differentiation rather than to the maturation
of the cells itself. It is also interesting to note that the pattern
of fibroglycan expression resembles that of TGF-β throughout chondrogenesis and osteogenesis (Heine et al., 1987),
and that TGF-β and related bone morphogenetic proteins
promote chondrogenesis and osteogenesis (Rosen and Thies,
1992). The codistribution of TGF-β and fibroglycan and the
identification of TGF-β and related osteogenic proteins and
several other growth factors as heparin-binding proteins
(McCaffrey et al., 1992) suggests that fibroglycan might
function as a binding site for growth and differentiation
factors and, directly or indirectly, help modulating the
signals that emanate from these molecules. Alternatively, it
indicates that possible effects of TGF-β-like cytokines on
the expression of fibroglycan in chondrogenic and
osteogenic cells deserve further investigation.
Differential expression of the syndecans
The expression of fibroglycan in the embryo is quite distinct
from that of syndecan-1. In contrast to fibroglycan, syndecan1 is mostly found on epithelial cell surfaces, from where it
may be lost, transiently during morphogenesis or permanently
with terminal differentiation of the cells (Trautman et al.,
1991). Syndecan-1 is also transiently present in mesenchymal
cells when the cells condense (Thesleff et al., 1988; Trautman
et al., 1991; Brauker et al., 1991), but it localizes in different
cell types and is expressed at different developmental stages
compared to fibroglycan. For example, syndecan-1 is
expressed during lung organogenesis, but mainly by the
epithelial cells of the branching tubules and temporarily by
the juxtaepithelial mesenchymal cells, at the initial stage of
lung development when these do not yet express fibroglycan
(Fig. 10; Brauker et al., 1991). During limb development a
similar situation occurs. There is no noticeable staining for
fibroglycan in the mesenchyme at the early morphogenetic
853
phase, but the staining becomes notable with the onset of
chondrogenesis and with the differentiation of the dermis
(Fig. 8). Syndecan-1, on the contrary, is only temporarily
expressed by the mesenchymal cells in the limb bud and
mainly at the time of initial morphogenesis. The staining is
initially widely spread throughout the mesenchyme, but
decreases with the elongation of the bud and disappears when
the cytodifferentiation starts and chondrogenesis and myogenesis are initiated (Solursh et al., 1990). Rather, in limb, fibroglycan mimics syndecan-3, a proteoglycan that belongs to the
same molecular family as fibroglycan and syndecan-1 and
that is expressed in high amounts at the onset of (chick) limb
cartilage differentiation when the expression of syndecan-1
ceases (Gould et al., 1992). A last example of these near
mirror images of syndecan-1/fibroglycan expression can be
seen during vibrissal development. Fibroglycan becomes
notable after budding has started and gradually intensifies in
the mesenchyme that surrounds the vibrissal buds (Fig. 9).
Syndecan-1, on the contrary, is highly expressed by the ectodermal cells and also by the underlying mesenchymal cells
before the epithelial buds invaginate. With the ingrowth of the
epithelial bud, the mesenchyme expression decreases and
soon disappears leaving only an epithelial staining (Trautman
et al., 1991). All these examples indicate that the expression
of fibroglycan occurs in different cells and at a later developmental stage of the mesenchymal cells than the one during
which syndecan-1 prevails. This would suggest that fibroglycan and syndecan-1, and probably still more cell surface proteoglycans, assume different functions during tissue morphogenesis and differentiation, affecting different processes or
different temporal and spatial aspects of a particular developmental process.
We thank Helga Ceulemans, Christien Coomans, Gisèle
Degeest, Staf Doucet, An Rayé and Marleen Willems for their
expert technical assistance. We also thank Dr. M. Bernfield for the
gift of 281-2 antibody. These investigations have been supported
by grants 3.0066.87 and 3.0073.91 from the Nationaal Fonds voor
Wetenschappelijk Onderzoek’ of Belgium, by a grant ‘Geconcerteerde Onderzoeks Acties’ from the Belgian Government, and by
the Interuniversity Network for Fundamental Research sponsored
by the Belgian Government (1991-1995). Guido David is a
Research Director of the ‘Nationaal Fonds voor Wetenschappelijk
Onderzoek’ of Belgium.
REFERENCES
Bernfield, M., Banerjee, S. D., Koda, J. E. and Rapraeger, A. C. (1984).
Remodelling of the basement membrane: Morphogenesis and maturation.
CIBA Found. Symp. 108, 179-197.
Bernfield, M. and Sanderson, R. D. (1990). Syndecan, a
morphogenetically regulated cell surface proteoglycan that binds
extracellular matrix and growth factors. Phil. Trans. R. Soc. Lond.
327,171-186.
Bernfield, M., Kokenyesi, R., Kato, M., Hinkes, M. T., Spring, J., Gallo,
L. and Lose, E. J. (1992). Biology of the syndecans. Annu. Rev. Cell Biol.
8, 365-393.
Boehm, T., Gonzalez-Sarmiento, R., Kennedy, M. and Rabbitts, T. H.
(1991). A simple technique for generating probes for RNA in situ
hybridization: An adjunct to genome mapping exemplified by the RAG1/RAG-2 gene cluster. Proc. Natl. Acad. Sci. USA 88, 3927-3931.
Brauker, J. H., Trautman, M. S. and Bernfield, M. (1991). Syndecan, a
cell surface proteoglycan, exhibits a molecular polymorphism during
lung development. Dev. Biol. 147, 285-292.
854
G. David and others
Carey, D. J., Evans, D. M., Stahl, R. C., Asundi, V. K., Conner, K. J.,
Garbes, P. and Cizmeci-Smith, G. (1992). Molecular cloning and
characterisation of N-syndecan, a novel transmembrane heparan sulfate
proteoglycan. J. Cell Biol. 117, 191-201.
David, G. and Van den Berghe, H. (1985). Heparan sulfate-chondroitin
sulfate hybrid proteoglycan of the cell surface and basement membrane of
mouse mammary epithelial cells. J. Biol. Chem. 260, 11067-11074.
David, G., Lories, V., Decock, B., Marynen, P., Cassiman, J. J. and Van
den Berghe, H. (1990). Molecular cloning of a phosphatidylinositolanchored membrane heparan sulfate proteoglycan from human lung
fibroblasts. J. Cell Biol. 111, 3165-3176.
David, G., Bai, X. M., Van der Schueren, B., Cassiman, J. J., and Van
den Berghe, H. (1992a). Developmental changes in heparan sulfate
expression: in situ detection with monoclonal antibodies. J. Cell Biol.
119, 961-975.
David, G., Van der Schueren, B., Marynen, P., Cassiman, J. J. and Van
den Berghe, H. (1992b). Molecular cloning of amphiglycan, a novel
integral membrane heparan sulfate proteoglycan expressed by epithelial
and fibroblastic cells. J. Cell Biol. 118, 961-969.
Elenius, K., Salmivirta, M., Inki, P., Mali, M. and Jalkanen, M. (1990).
Binding of human syndecan to extracellular matrix proteins. J. Biol.
Chem. 265, 17837-17843.
Elenius, K., Vainio, S., Laato, M., Salmivirta, M., Thesleff, I. and
Jalkanen, M. (1991). Induced expression of syndecan in healing wounds.
J. Cell Biol. 114, 585-595.
Espelund, M., Prentice Stacy, R. A. and Jakobsen K. S. (1990) A simple
method for generating single-stranded DNA probes labeled to high
activities. Nucleic Acids Res. 20, 6157-6158.
Gallagher, J. T. (1989). The extended family of proteoglycans: social
residents of the pericellular zone. Curr. Opin. Cell Biol. 1, 1201-1218.
Gould, S. E., Upholt, W. B. and Kosher, R. A. (1992). Syndecan-3: a
member of the syndecan family of membrane-intercalated proteoglycans
that is expressed in high amounts at the onset of chicken limb cartilage
differentiation. Proc. Natl. Acad. Sci. USA 89, 3271-3275
Hayashi, K., Hayashi, M., Jalkanen, M., Firestone, J. H., Trelstad, R. L.
and Bernfield, M. (1987) Immunocytochemistry of cell surface heparan
sulfate proteoglycan in mouse tissues. A light and electron microscopic
study. J. Histochem. Cytochem. 35, 1079-1088.
Heine, U. I., Munoz, E. F., Flanders, K. C., Ellingsworth, L. R., Lam, HY. P., Thompson, N. L., Roberts, A. B. and Sporn, M. B. (1987). Role
of transforming growth factor-ß in the development of the mouse embryo.
J. Cell Biol. 105, 2861-2876.
Heine, U. I., Munoz, E. F., Flanders, K. C., Roberts, A. B. and Sporn, M.
B. (1990). Colocalization of TGF-beta 1 and collagen I and III, fibronectin
and glycosaminoglycans during lung branching morphogenesis.
Development 109, 29-36.
Heremans, A., Van der Schueren, B., Decock, B., Paulsson, M.,
Cassiman, J. J., Van den Berghe, H. and David, G. (1989). Matrixassociated heparan sulfate proteoglycan: Core protein-specific
monoclonal antibodies decorate the pericellular matrix of connective
tissue cells and the stromal side of basement membranes. J. Cell Biol. 109,
3199-3211.
Kiefer, M. C., Stephans, J. C., Crawford, K., Okino, K. and Barr, Ph. J.
(1990). Ligand-affinity cloning and structure of a cell surface heparan
sulfate proteoglycan that binds basic fibroblast growth factor. Proc. Natl.
Acad. Sci. USA 87, 6985-6989.
Kjellén, L. and Lindahl, U. (1991). Proteoglycans: structures and
interactions. Ann. Rev. Biochem. 60, 443-475.
Kojima, T., Shworak, N. W. and Rosenberg, R. D. (1992). Molecular
cloning and expression of two distinct cDNA-encoding heparan sulfate
proteoglycan core proteins from a rat endothelial cell line. J. Biol. Chem.
267, 4870-4877.
Kozak, M. (1989). The scanning model for translation: an update. J. Cell
Biol. 108, 229-241.
Kratochwil, K. (1969). Organ specificity in mesenchymal induction
demonstrated in the embryonic development of the mammary gland of the
mouse. Dev. Biol. 20, 46-71.
Lories, V., De Boeck, H., David, G., Cassiman, J. J. and Van den
Berghe, H. (1987). Heparan sulfate proteoglycans of human lung
fibroblasts. Structural heterogeneity of the core proteins of the
hydrophobic cell-associated forms. J. Biol. Chem. 262, 854-859.
Lories, V., Cassiman, J. J., Van den Berghe, H. and David, G. (1989).
Multiple distinct membrane heparan sulfate proteoglycans in human lung
fibroblasts. J. Biol. Chem. 264, 7009-7016.
Lories, V., Cassiman, J. J., Van den Berghe, H. and David, G. (1992).
Differential expression of cell surface heparan sulfate proteoglycans in
human mammary epithelial cells and lung fibroblasts. J. Biol. Chem. 267,
1116-1122.
Mali, M., Jaakkola, P., Arvilommi, A. -M. and Jalkanen, M. (1990).
Sequence of human syndecan indicates a novel gene family of integral
membrane proteoglycans. J. Biol. Chem. 265, 6884-6889.
Mann, D. M., Yamaguchi, Y., Bourdon, M. A. and Ruoslathi, E. (1990).
Analysis of glycosaminoglycan substitution in decorin by site-directed
mutagenesis. J. Biol. Chem. 265, 5317-5323.
Marynen, P., Zhang, J., Cassiman, J. J., Van den Berghe, H. and David,
G. (1989). Partial primary structure of the 48- and 90 kilodalton core
proteins of cell surface-associated heparan sulfate proteoglycans of lung
fibroblasts. Prediction of an integral membrane domain and evidence for
multiple distinct core proteins at the cell surface of human lung
fibroblasts. J. Biol. Chem. 264, 7017-7024.
McCaffrey, T. A., Falcone, D. J. and Du, B. (1992). Transforming growth
factor-ß1 is a heparin-binding protein: identification of putative heparinbinding regions and isolation of heparins with varying affinity for TGFß1. J. Cell. Physiol. 152, 430-440.
Mertens, G., Cassiman, J. J., Van den Berghe, H., Vermylen, J. and
David, G. (1992). Cell surface heparan sulfate proteoglycans from human
vascular endothelial cells. Core protein characterization and antithrombin
III binding properties. J. Biol. Chem. 267, 20435-20443.
Pierce, A., Lyon, M., Hampson, I. N., Cowling, G. J. and Gallagher, J. T.
(1992). Molecular cloning of the major cell surface heparan sulfate
proteoglycan from rat liver. J. Biol. Chem. 267, 1-7.
Platt, J. L., Trescony, P., Lindman, B. and Oegema, T. R. (1990).
Heparin and heparan sulfate delimit nephron formation in fetal
metanephric kidneys. Dev. Biol. 139, 338-348.
Roman, J., Little, C. W. and McDonald, J. A. (1991). Potential role of
RGD-binding integrins in mammalian lung branching morphogenesis.
Development 112, 551-558.
Rosen, V., and Thies, R. S. (1992). The BMP proteins in bone formation
and repair. Trends Genetics 8, 97-102.
Ruoslahti, E. (1989). Proteoglycans in cell regulation. J. Biol. Chem. 264,
13369-13372.
Sanger, F., Nicklem, S. and Coulson, A. R. (1977). DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467.
Saunders, S. and Bernfield, M. (1988). Cell surface proteoglycan binds
mouse mammary epithelial cells to fibronectin and behaves as a receptor
for interstitial matrix. J. Cell Biol. 106, 423-430.
Saunders, S., Jalkanen, M., O’Farrell, S. and Bernfield, M. (1989).
Molecular cloning of syndecan, an integral membrane proteoglycan. J.
Cell Biol. 108, 1547-1556.
Smith, C. I., Webster, E. H., Nathanson, M. A., Searls, R. L. and Hilfer,
S. R. (1990). Altered patterns of proteoglycan deposition during
maturation of the fetal mouse lung. Cell Differ. Dev. 32, 83-96.
Solursh, M., Reiter, R. S., Jensen, K. L., Kato, M. and Bernfield, M.
(1990). Transient expression of a cell surface heparan sulfate
proteoglycan (syndecan) during limb development. Dev. Biol. 140, 83-92.
Sun, X., Mosher, D. F. and Rapraeger, A. (1989). Heparan sulfatemediated binding of epithelial cell surface proteoglycan to
thrombospondin. J. Biol. Chem. 264, 2885-2889.
Sutherland, A. E., Sanderson, R. D., Mayes, M., Seibert, M., Calarco, P.
G., Bernfield, M. and Damsky, C. H. (1991) Expression of syndecan, a
putative low affinity fibroblast growth factor receptor, in the early mouse
embryo. Development 113, 339-351.
Thesleff, I., Jalkanen, M., Vainio, S. and Bernfield, M. (1988). Cell
surface proteoglycan expression correlates with epithelial-mesenchymal
interaction during tooth morphogenesis. Dev. Biol. 129, 565-572.
Trautman M. S., Kimelman, J. and Bernfield, M. (1991). Developmental
expression of syndecan, an integral membrane proteoglycan, correlates
with cell differentiation. Development 111, 213-220.
Vainio, S., Lehtonen, E., Jalkanen, M., Bernfield, M. and Saxén, L.
(1989). Epithelial-mesenchymal interactions regulate the stage-specific
expression of a cell surface proteoglycan, syndecan, in the developing
kidney. Dev. Biol. 134, 382-391.
Vainio, S., Jalkanen, M., Vaahtokari, A., Sahlberg, C., Mali, M.
Bernfield, M. and Thesleff, I. (1991). Expression of syndecan gene is
induced early, is transient, and correlates with changes in mesenchymal
cell proliferation during tooth organogenesis. Dev. Biol. 147, 322-333.
(Accepted 26 August 1993)