| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
3(VI) chain interferes with protein folding
* Max-Planck-Institut für Biochemie, D-82152 Martinsried, Germany;
Biophysics Section, Blackett Laboratory and Division of Medicine, Imperial College, London SW7 2AZ, U.K.;
Department of Dermatology and Cutaneous Biology, and Biochemistry and Molecular Pharmacology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107, USA;
§ Department of Medicine, Section of Medical Genetics, Duke University Medical Center, Durham, North Carolina 27710, USA;
|| Department of Neurology, University of Vermont College of Medicine, Burlington, Vermont 05405, USA
1Correspondence: Max-Planck-Institut für Biochemie, Am Klopferspitz 18A, D-82152 Martinsried, Germany. E-mail: TIMPL{at}biochem.mpg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3 chain of type VI collagen was found in a large family affected
with Bethlem myopathy. Recombinant production of N2 (~200 residues)
in transfected mammalian cells has now been used to examine the
possibility that the mutation interfered with protein folding. The
wild-type form and a G1679A mutant were produced at high levels and
shown to fold into a stable globular structure. Only a small amount of
secretion was observed for mutants G1679E and G1679Q, which apparently
were efficiently degraded within the cells. Homology modeling onto the
related von Willebrand factor A1 structure indicated that substitution
of G1679 by the bulky E or Q cannot be accommodated without
considerable changes in the folding pattern. This suggests protein
misfolding as a molecular basis for this particular mutation in Bethlem
myopathy, in agreement with radioimmunoassay data showing reduced
levels of domain N2 in cultured fibroblasts from two patients.—Sasaki,
T., Hohenester, E., Zhang, R.-Z., Gotta, S., Speer, M. C., Tandan,
R., Timpl, R., Chu, M.-L. A Bethlem myopathy Gly to Glu mutation in the
von Willebrand factor A domain N2 of the collagen 3(VI) chain
interferes with protein folding.
Key Words: haploin sufficiency • inherited disease • protein misfolding • recombinant production
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The collagen triple
helix is the most typical case, with an exclusive requirement for
glycine at every third sequence position in order to allow the
zipper-like folding of the three chains into a rigid structure. Glycine
is placed close to the common axis and provides hydrogen bonds to
adjacent chains (2
3
4)
. It is therefore not surprising
that many glycine substitutions of fibril-forming collagen types are
considered to cause various dominant and recessive genetic diseases,
including osteogenesis imperfecta and chondrodysplasias
(5)
. These substitutions are likely to introduce
flexibility into the triple helix, which causes premature degradation
of procollagens or interferes with fibrillogenesis. The same
substitutions in the triple-helix of collagen type VII, which forms
anchoring fibrils for basement membranes, are found in various forms of
dystrophic epidermolysis bullosa and cause skin blistering (6
, 7)
.
Substitutions of glycine are found in various other genetic disorders,
which do not include the collagen triple helix. They were identified in
the globular domains of some collagen types (5
, 8
, 9)
and
various modules of noncollagenous microfibrils (10)
, and
are more difficult to interpret in terms of their struction–function
relationship. These substitutions are frequently found in calcium
binding EG modules of fibrillin in patients with Marfan syndrome and
related disorders (11)
. A Gly to Ser change in one of
these modules was shown to cause defective folding because the Ser side
chain interferes with the loop c disulfide bridge (12)
. A
Gly to Val substitution was identified in the carboxyl-terminal
globular domain of collagen type X in a family with Schmid metaphyseal
chondrodysplasia. In vitro expression of the mutant collagen
demonstrated lack of trimerization, but it was not clear whether this
was due to defective globular folding (9)
.
Several missense and deletion mutations have also been reported for the
triple helical domains of the three chains that constitute the
microfibrillar collagen type VI (13
14
15
16)
. Patients
involved suffer from Bethlem myopathy, a dominantly inherited,
childhood-onset mild muscular dystrophy with joint constructures
(17)
. A further mutation in such patients was a Gly1679Glu
change in domain N2 from the large amino-terminal globular structure
N9-N1 of the 3(VI) chain (8)
. Each individual N domain
corresponds to a von Willebrand factor domain A-like (VWA) module of
~200 residues (18)
.
A recombinant fragment corresponding to N9-N2 was shown to fold into
eight small globular domains, which can take up different arrangements
relative to one another (19)
. Fragment N9-N2 or its
subdomains were also shown to bind heparin and hyaluronan. Thus, the
Bethlem myopathy mutation could have affected either binding properties
or protein folding. The latter appeared more likely based on the
recently elucidated structure of von Willebrand factor domain A1
(20
, 21)
.
In the present study, we approached these questions by recombinant
production in mammalian cells of domain N2 and several mutants of the
critical Gly. These cell systems would immediately degrade a product
not properly designed to represent an autonomously folding unit, as
shown in previous studies (22
23
24
25)
. The data demonstrated
autonomous folding properties for domain N2, which could accommodate an
Ala substitution but not amino acids at position 1679 with a bulkier
side chain. These observations were rationalized by homology modeling
of the domain N2 structure.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3(VI) collagen
chain, F019 cDNA (18)
was amplified by polymerase chain
reaction (PCR) with primers HCOL6A3–41 (GTC AGC TAG CCG AGA AGA AGA
AAG CAG AC) and HCOL6A3–42C GTC ACT CGA GTC AAA GGG TTT CAT GCA TCG
C), which introduced an Nhel and Xhol site at the 5'- and 3'- end,
respectively. N2 with the Gly to Glu mutation was prepared by RT-PCR
amplification of total RNA from fibroblasts of the patient carrying
this mutation, using the same primer pairs (8)
. N2 domains
with Gly to Ala and Gly to Gln mutations were prepared by two
sequential PCR reactions with the above primers and two internal
primers containing the mutations, as described (26)
. The
internal primers for the Ala mutant were HCOL6A3-N2AF (CCA AGT GGC GCT
TGT CCA G) and HCOL6A3-N2AR (CTG GAC AAG CGC CAC TTG G), and for the
Gln mutant were HCOL6A3-N2QF (CCA AGT GCA GCT TGT CCA G) and
HCOL6A3-N2QR (CTG GAC AAG CTG CAC TTG G).
Total RNA was reverse transcribed with Superscript II reverse
transcriptase (Life Technologies, Gaithersburg, Md.) and PCR
amplification was performed with the Ampli Taq PCR kit
(Perkin Elmer, Foster City, Calif.) following the manufacturer’s
instructions. The PCR products were digested with Nhel and Xhol, then
cloned into the corresponding sites of the expression vector pCEP-Pu,
which contains the signal sequence for BM-40 (27)
. The
expression vectors were sequenced to ensure no other mutations were
introduced by PCR. These vectors were then used to transfect 293-EBNA
cells (Invitrogen, San Diego, Calif.) for episomal production
(27)
. Efficiency of transfections were analyzed by
Northern blots of total mRNA prepared from puromycin-resistant cells
using cDNA encoding N2, following standard protocols.
Protein purification
Serum-free conditioned medium (0.5–1l) was collected from
transfected cells, dialyzed against 0.05M Tris-HCl, pH 8.6, and passed
over a DEAE cellulose column (2.5x15 cm) equilibrated in the same
buffer. Elution with a linear 0–0.4M NaCl gradient (300/300 ml)
displaced the fragments at ~0.2M NaCl. They were concentrated by
ultrafiltration and passed over a Superose 12 column (HR16/50,
Pharmacia, Piscataway, N.J.) equilibrated in 0.2M ammonium acetate, pH
6.8. Purified fragments were lyophilized and dissolved in 0.2M ammonium
bicarbonate.
Protein analysis
Protein concentrations were determined after hydrolysis with 6M
HCl (16 h, 110°C) on a Biotronik LC 3000 analyzer. Sodium dodecyl
sulfate (SDS) gel electrophoresis followed standard protocols. Edman
degradation on a 473 sequencer was performed according to the
manufacturer’s instructions. Electron microscopy was done after rotary
shadowing of protein samples (28)
and circular dichroism
spectroscopy followed previous methods (19)
. Chymotrypsin
digestions were carried out at an enzyme-substrate ratio of 1:50 for
24 h at 37°C.
Structure homology modeling
The program CLUSTALW (29)
was used to align the
sequences of domains N2 to N9 of human 3(VI) collagen with those of
A domains of known structure, namely, domains A1 (PDB accession codes
1oak, 1auq) and A3 (1atz, 1au3) of human von Willebrand factor (VWF)
and the I domains of human integrins CD11a/CD18 (1ido) and CD11b/CD18
(1lfa). From this multiple alignment, the VWF A1 domain was judged to
be the best template for homology modeling of the region around Gly1679
in 3(VI) collagen. Residue changes were done with the program O
(30)
. Side chain conformations were taken from the rotamer
database of O and then optimized. The main chain conformation was not
changed significantly, with the exception of a short stretch of amino
acids following helix 1 (residues 1732 to 1736), which is at some
distance from the site of mutation.
Immunological assays
A rabbit antiserum against fragment N9-N2 (19)
was
used throughout the study. Part of it was affinity-purified on a column
of fragment N2. Antibodies eluted from the column reacted in
enzyme-linked immunoassay equally well with N2 and N9-N2, whereas the
nonbinding antibodies showed a high titer for N9-N2
(1:2.104) but not for N2 (titer less than 1:400).
Both sets of antibodies were used in immunoblots following a previously
described procedure (31)
. Radioimmuno-inhibition assays
were established with the antiserum and 1 ng each of
125I-labeled N2-N9 (antiserum dilution
1:3.104) or N2 (dilution 1:2500) following
established protocols (32)
. They were used for
quantitation of samples with either fragment N2 or N9-N2 as reference
inhibitor. All the concentrations determined are recorded as
equivalents of N2 or N9-N2, respectively.
Cell cultures
Skin fibroblasts from patients and nonaffected controls
(8)
were grown to confluency in Dulbecco’s minimal
essential medium containing 10% fetal calf serum. They were then
incubated in serum-free medium (6 ml) before collecting the cells and
medium. Intracellular and matrix deposited collagen VI was solubilized
by detergent extraction (1 ml) as described previously
(31)
. The medium and the cell lysate were used in
radioimmuno-inhibition assays and contained approximately equal amounts
of collagen VI antigens.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). This level of variation is a common observation and reflects how well
the transfected cells were selected by the puromycin treatment.
Quantitation of domain N2 in serum-free culture medium and cell lysates
by radioimmunoassays (see below) revealed striking differences in
protein production, however (Table 1
). Comparably high levels were found for wild-type N2 and mutant G1679A,
but levels were 30- to 100-fold lower for mutants G1679E and G1679Q.
This strongly indicated extensive intracellular degradation of the
latter two mutants.
|
|
Serum-free culture medium was used to purify N2 and mutants G1679A and
G1679E by ion-exchange and molecular sieve chromatography. Fragments N2
and mutant G1679A were obtained in good purity and yields (6–12 mg/l
medium) and showed a main electrophoretic band of 26 kDa (Fig. 2
, lanes 1, 2). Both products started with a single amino-terminal
sequence, APLAEKKK, where APLA is derived from the signal peptide
cleavage region of the expression vector (27)
. Because of
low production, mutant G1679E could only be partially purified, but
contained a major band of ~28 kDa (Fig. 2
, lane 3). This band was
identified as the mutant by immunoblotting and amino-terminal
sequencing. Its slower electrophoretic mobility is presumably explained
by a greater amount of unfolding caused by SDS used during
electrophoresis.
|
Properties of the recombinant proteins
Electron microscopy of rotary-shadowed recombinant N2 (Fig. 3A
) showed small globular particles that corresponded to the
size of individual N modules previously predicted from the larger
fragment N9-N2 (19)
. The same were observed for mutant
G1679A (Fig. 3B
). The circular dichroism spectra of both
recombinant fragments were identical to that previously shown for
fragment N9-N2 (19)
, with two distinct troughs of
ellipticity (-13000 deg cm2
dmol-1) at 210 and 222 nm. This indicates
~40% helix and some ß structure. Together, the data support
the correct folding of recombinant N2 and mutant G1679A.
|
Chymotrypsin was previously shown to cleave fragment N9-N2 at the
borders between individual domains (19)
. In agreement with
this, recombinant N2 and mutant G1679A were now shown to be stable
against chymotrypsin proteolysis (Fig. 2
, lanes 4, 5). The 28 kDa band
of the mutant G1679E disappeared completely on this treatment, however
(Fig. 2
, lane 6), and no smaller fragments could be detected by
immunoblotting. A chymotrypsin fragment N3-N2 was previously shown to
bind heparin (19)
. This property is not shared by
recombinant N2, which therefore implies involvement of domain N3 in the
binding epitope.
A specific radioimmuno-inhibition assay for N2 epitopes was developed
using an antiserum against N9-N2 (Fig. 4
). This assay could be equally well inhibited by fragments N2 (26 kDa)
and N9-N2 (180 kDa) with half-maximal inhibition at 0.15 nM. Culture
medium and cell lysates from cells transfected with N2 or the mutant
G1679A showed similar inhibition gradients, which demonstrated
identical antigenic epitopes and allowed us to determine precisely the
concentrations of the recombinant products (Table 1)
. These
concentrations were lower in cell lysate and medium of mutants G1679E
and G1679Q (Table 1)
; because the inhibition curves were distinctly
less steep (Fig. 4)
, they could not be determined with great accuracy.
The lower steepness of these curves also indicated that these mutants
share some but not all epitopes (32)
, presumably due to
imperfect folding. This also suggested rapid intracellular degradation
that prevents any significant secretion of the mutants (Table 1)
.
|
A second inhibition assay with 125I-labeled fragment N9-N2 and the same antiserum could similarly be inhibited by N9-N2, with half-maximal inhibition at 0.06 nM (data not shown). Fragment N2, however, showed only 8–10% inhibition over a broad concentration range (0.02–14 nM), indicating a relatively minor contribution of antigenic epitopes. Furthermore, since the assays for N2 and N9-N2 show a two- to threefold difference in sensitivity, they could only be used for a relative quantitation of antigenic epitopes in biological samples (see below).
Structural modeling of the mutations
To understand the effect of the G1679E mutation on protein
folding, we initially tried to obtain crystals of the N2 domain of the
type VI collagen 3(VI) chain, but these attempts were not
successful. We therefore used homology modeling to understand this
detrimental effect. A pairwise sequence alignment of the collagen
3(VI) N2 domain and the A1 domain of van Willebrand factor is shown
in Fig. 5A
(20.1% sequence identity, 42.3% similarity). We used the
structure of the VWF A1 domain, determined independently at 2.2 Å
(21)
and 2.3Å resolution (22)
, as a template
to model the local structure around the site of the G1679E mutation in
domain N2. Only small adjustments in main chain conformation were
required to accommodate the residue changes. Figure 5B
shows
the resulting model. Gly1679 in N2 corresponds to Ala554 in VWF A1; the
requirement for a glycine, alanine, or serine residue at this position
throughout all A domains has been noted previously (8)
.
Val1641, Val1681, Leu1693, and Phe1731, which surround Gly1679, are
strictly conserved between N2 and VWF A1. These residues are also
highly conserved in all other members of the A domain superfamily. The
salt bridge between Asp1689 and His1725 appears to be unique to the
domain N2 and replaces a pair of conserved hydrophobic or aromatic
residues in other A domains. Likewise, His1730 replaces an invariant
apolar residue (valine, leucine, or isoleucine in most cases). Finally,
Phe1692, the side chain of which stacks with that of His1730 in our
model, replaces Gly606 of the VWF A1 domain. It is noted, however, that
an apolar residue at this position is found in five of the nine N
domains of collagen 3(VI) chains (18)
.
|
Antigenic epitopes of domain N2 in control and patient fibroblasts
Autosomal dominant mutations of single amino acid residues are
usually difficult to analyze in tissue samples, and no significant
differences in size or ratios of the collagen type VI chains have
been detected before between fibroblasts from patients with the G1679E
mutation and unaffected family members (8)
. We have now
used antibodies against fragment N9-N2 and made them specific for
domain N2 and the remaining N9-N3 structure by immunoadsorption. In
immunoblots of medium from patient and control fibroblasts, both sets
of antibodies reacted with a major broad band of ~250 kDa (Fig. 6
), which corresponds to the natural, processed form of the 3(VI)
chain (33)
. Both sets of antibodies also recognized a
series of weaker bands, particularly in the region 120–200 kDa and
around 70 kDa, indicating further proteolytic processing. However,
there was no significant difference in these additional collagen VI
bands between the patient and control samples.
|
The radioimmuno-inhibition assays for N9-N2 and N2 described above were
used as a second approach to analyze the fibroblast medium and cell
lysate from two patients and two controls (Table 2
). The data showed the presence of both antigenic epitopes, but also a
considerable individual variability between the different sources of
the fibroblasts. The relative ratio of N9-N2 to N2 epitopes was ~3 in
the controls and increased to 4.2 and 4.9 in the patients, which
indicates ~20–30% loss of N2 epitopes. This could be due to
unfolding, as shown for the recombinant products, or may reflect
partial degradation of N2.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1(VI) and 2(VI) chains
(34)
and domain N9-N2 of the 3(VI) chain
(19)
. Furthermore, it was shown that the 3(VI) chain
containing N6-N1 was essential for chain association and secretion of
collagen VI (35)
. This also demonstrated that these large
recombinant proteins possess all the information necessary for proper
folding. Similar studies with other extracellular matrix proteins and
domains, however, showed that this is critically dependent on the
correct choice of module boundaries (22
, 24
, 25)
and also
on individual amino acid substitutions (12
, 36)
. This led
to the conclusion that proteases within the rough endoplasmic reticulum
are important quality controllers for folding (23)
.
Using transfected human cells, we show here that domain N2 of collagen
VI folds autonomously into a globular structure that is resistant to
proteases. When Gly1679, a critical residue in a family with Bethlem
myopathy (8)
, was mutated to Ala, these properties did not
change. However, mutation of this residue to Glu, as found in the
patients, or to Gln was correlated with low levels of secretion very
likely caused by rapid intracellular degradation. Since the N2 mutant
Gly1679Gln could not be detected at all in the culture medium by
radioimmunoassay and immunoblotting, it might cause an even more severe
phenotype than the natural Glu mutation found uniformly in 19 patients
(8)
.
The model of the N2 domain of 3(VI) collagen presented in Fig. 5
readily explains the detrimental effect of the Gly1679Glu mutation on
the folding and stability of this domain. Gly1679 is located in
ß-strand ß2, a highly conserved region of the VWA module, and is
surrounded by conserved apolar residues. The -helix 3, which
covers one face of the central ß-sheet of the A1 domain, contributes
His1730, the side chain of which is in close contact with both the C
atom of Gly1679 and a number of key apolar resides in the area. Due to
the tight packing of residues around Gly1679, the large charged side
chain of a glutamic acid at position 1679 could not be accommodated
without drastic structural changes. In particular, in the present model
of domain N2 every possible conformation of a glutamic acid would
result in a serious clash of the carboxylate group with the imidazole
ring of His1730. Such changes could have an effect on collagen VI
fibril assembly or the interaction with other extracellular matrix
proteins. However, the Gly1679Glu mutation may cause disease simply by
decreasing the amount of 3(VI) chain available for collagen triple
helix formation. It is not clear whether an N2 domain carrying the
Gly1679Glu mutation would fold at all, with misfolded protein being
rapidly cleared in vivo. This scenario seems to be supported
by the somewhat lower content of domain N2 in fibroblast cultures of
patients, as indicated from radioimmunoassays.
Bethlem myopathy is dominantly inherited (17)
, and in
affected families the disease has been linked to either chromosomes 2
or 21 (14
, 37)
, where the three collagen type VI chain
genes reside (38)
. Subsequent molecular analyses have
demonstrated that mutations in these collagen genes indeed cause the
disease. Missense mutations that disrupt the invariant glycine residues
in the triple-helical domains of the 1(VI), 2(VI), and 3(VI)
collagen chains are most common, having been reported in three Dutch
families and one Italian family (14
, 16)
.
Haploinsufficiency of collagen type VI due to splice site mutations is
also emerging as a mechanism leading to the disease. This was first
reported in an Australian family, in whom the mutation introduces a
premature stop codon in the 1(VI) gene, leading to mRNA degradation
(15)
. More recently, in two other families from Australia
and Italy, respectively, the disease was shown to be caused by exon
skipping, in which a cysteine residue is deleted in the triple-helical
domain of the 1(VI) chain (39)
. This cysteine is
thought to be involved in the assembly of type VI collagen dimers prior
to secretion (33)
. Moreover, mice with targeted
inactivation of one 1(VI) collagen allele showed histopathology of
myopathy (40)
. This demonstrates that apart from the
Gly1679Glu mutation in a large family of French-Canadian descent
(8)
, several other genetic mechanisms affecting the
collagen VI genes lead to similar myopathy phenotypes. This would
support the interpretation that a reduced collagen VI content and not a
specific functional failure is the common molecular deficiency of
Bethlem myopathy.
Two more glycine mutations have been found to impair folding in protein
modules not related to the collagen triple helix. These involve a Gly
to Ser change in EG modules of coagulation factor IX and fibrillin-1,
mutations that are linked to mild forms of hemophilia B or familial
aortic aneurism, respectively. Both mutations lead to deficient
folding, which is explained by interference of the serine side chain
with the formation of a critical disulfide bridge (12)
. A
Gly to Val change in the carboxyl-terminal C1q-like module of collagen
type X, a mutation linked to Schmid’s metaphyseal chondrodysplasia,
was previously shown to prevent trimerization (9)
and more
recently found to interfere with globular folding (41)
.
Recombinant production of four different mutants in Escherichia
coli showed that apart from failing to trimerize, the mutants
bound extensively to the GroEL chaperone and were highly sensitive to
trypsin. Homology modeling on a related ß jelly roll fold indicated
that two mutations in ß strands caused premature stop codons and a
Tyr598Asp mutation in a ß-strand is close to the trimerization site.
An additional Gly618Val mutation in a loop region may have caused
steric incompatibility, as discussed in another study
(42)
. Here again, it is obvious that different mutations
can cause the same molecular defect and hence the same clinical
phenotype.
Our study and the data of Dublet et al. (41)
clearly
indicate that recombinant production in both mammalian and bacterial
cells is useful to analyze the effect of mutations on the folding of
globular domains. In the first case, this will require the correct
choice of module boundaries and misfolding can be identified by lack of
production. Bacterial production needs as a prerequisite the proper
folding of the wild-type structure, as described in ref
41
. Misfolded products can then be identified by protease
sensitivity and/or chaperone binding. The two approaches are
complementary but may be also used as alternatives in the
identification of Gly mutations, causing the misfolding of globular
domains.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
3(VI) collagen gene (COL6A3) in a family with Bethlem myopathy. Hum. Mol. Genet. 7,807-8121(X)NC1 domain resulting in Schmid metaphyseal chondrodysplasia. J. Biol. Chem. 270,4558-4562
Ser change causes defective folding in vitro of calcium binding epidermal growth factor-like domains from factor IX and fibrillin-1. J. Biol. Chem. 273,7807-78133 chain: similarity to von Willebrand factor, fibronectin, actin, salivary proteins and aprotinin type protease inhibitors. EMBO J 9,385-393[Medline]
3 chain and its binding to heparin and hyaluronan. EMBO J 11,4281-4290[Medline]
2 chain. FEBS Lett 426,71-76[Medline]
-dystroglycan and sulfatide binding to the G domain of laminin 1 chain by site-directed mutagenesis. J. Mol. Biol. 287,253-264[Medline]
1(VI) and 2(VI) chains of human collagen type VI. Eur. J. Biochem. 221,177-185[Medline]
3(VI) chain in collagen VI assembly. Expression of an 3(VI) chain lacking N-terminal modules N10–N7 restores collagen VI assembly, secretion and matrix deposition in an 3(VI) deficient cell line. J. Biol. Chem. 273,7423-74301 and 2 chains and perlecan to heparin, sulfatides. -dystroglycan and several extracellular matrix proteins. EMBO J 18,863-870[Medline]
1(VI) collagen chain in an Italian family affected by Bethlem myopathy. Biochem. Biophys. Res. Commun. 258,802-807[Medline]
This article has been cited by other articles:
|
|
 |
|
  D. Hicks, A. K. Lampe, R. Barresi, R. Charlton, C. Fiorillo, C. G. Bonnemann, J. Hudson, R. Sutton, H. Lochmuller, V. Straub, et al. A refined diagnostic algorithm for Bethlem myopathy Neurology, April 1, 2008; 70(14): 1192 - 1199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Otten, R Wagener, M Paulsson, and F Zaucke Matrilin-3 mutations that cause chondrodysplasias interfere with protein trafficking while a mutation associated with hand osteoarthritis does not J. Med. Genet., October 1, 2005; 42(10): 774 - 779. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A K Lampe and K M D Bushby Collagen VI related muscle disorders J. Med. Genet., September 1, 2005; 42(9): 673 - 685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A K Lampe, D M Dunn, A C von Niederhausern, C Hamil, A Aoyagi, S H Laval, S K Marie, M-L Chu, K Swoboda, F Muntoni, et al. Automated genomic sequence analysis of the three collagen VI genes: applications to Ullrich congenital muscular dystrophy and Bethlem myopathy J. Med. Genet., February 1, 2005; 42(2): 108 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G C Jackson, F S Barker, E Jakkula, M Czarny-Ratajczak, O Makitie, W G Cole, M J Wright, S F Smithson, M Suri, P Rogala, et al. Missense mutations in the {beta} strands of the single A-domain of matrilin-3 result in multiple epiphyseal dysplasia J. Med. Genet., January 1, 2004; 41(1): 52 - 59. [Full Text] [PDF]
|
|
|
|
 |
|
 B. Leitinger Molecular Analysis of Collagen Binding by the Human Discoidin Domain Receptors, DDR1 and DDR2. IDENTIFICATION OF COLLAGEN BINDING SITES IN DDR2 J. Biol. Chem., May 2, 2003; 278(19): 16761 - 16769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Wiberg, D. Heinegard, C. Wenglen, R. Timpl, and M. Morgelin Biglycan Organizes Collagen VI into Hexagonal-like Networks Resembling Tissue Structures J. Biol. Chem., December 13, 2002; 277(51): 49120 - 49126. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. C. Scacheri, E. M. Gillanders, S. H. Subramony, V. Vedanarayanan, C. A. Crowe, N. Thakore, M. Bingler, and E. P. Hoffman Novel mutations in collagen VI genes: Expansion of the Bethlem myopathy phenotype Neurology, February 26, 2002; 58(4): 593 - 602. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Lamande, M. Morgelin, C. Selan, G. J. Jobsis, F. Baas, and J. F. Bateman Kinked Collagen VI Tetramers and Reduced Microfibril Formation as a Result of Bethlem Myopathy and Introduced Triple Helical Glycine Mutations J. Biol. Chem., January 11, 2002; 277(3): 1949 - 1956. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
). Biological samples were medium (
) and cell
lysate (
) from cells transfected with wild-type N2 and medium (
)
and cell lysate (
) from cells transfected with the G1679E mutant,
used at the dilutions shown at the top.
