Three mutations of the B4GALT7 gene [encoding β1,4-GalT7 (β1,4-galactosyltransferase 7)], corresponding to A186D, L206P and R270C, have been identified in patients with the progeroid form of the Ehlers–Danlos syndrome and are described as being associated with the reduction or loss of β1,4-GalT7 activity. However, the molecular basis of the reduction or loss of activity remained to be determined. In the present study, wild-type, A186D, L206P and R270C β1,4-GalT7 were expressed in CHO618 cells as membrane proteins and in Escherichia coli as soluble proteins fused to MBP (maltose-binding protein). The ability of the expressed proteins to transfer galactose from donor to acceptor substrates was systematically characterized by kinetic analysis. The physicochemical properties of soluble proteins were explored by isothermal titration calorimetry, which is a method of choice when determining the thermodynamic parameters of the binding of substrates. Together, the results showed that: (i) the L206P mutation abolished the activity when L206P β1,4GalT7 was either inserted in the membrane or expressed as a soluble MBP–full-length fusion protein; (ii) the A186D mutation weakly impaired the binding of the donor substrate; and (iii) the R270C mutation strongly impaired the binding of the acceptor substrate. Moreover, the ex vivo consequences of the mutations were investigated by evaluating the priming efficiency of xylosides on GAG (glycosaminoglycan) chain initiation. The results demonstrate a quantitative effect on GAG biosynthesis, depending on the mutation; GAG biosynthesis was fully inhibited by the L206P mutation and decreased by the R270C mutation, whereas the A186D mutation did not affect GAG biosynthesis severely.

INTRODUCTION

EDS (Ehlers–Danlos syndrome) is a heterogeneous group of more than ten inherited connective-tissue disorders characterized by joint hypermobility, cutaneous fragility and hyperextensibility [1]. All involve a genetic disorder defect in collagen and connective-tissue synthesis and structure. The different types of EDS are classified according to the signs and symptoms that are manifested and each type of EDS is a distinct disorder that ‘runs true’ in a family. To date, among these, only five patients have been described with the progeroid variant of EDS, which is a very rare disorder characterized by a premature aging phenotype with a loose elastic skin, failure to thrive, joint laxity, psychomotor retardation, hypotonia and macrocephaly [24]. Biochemically, the replacement of PGs (proteoglycans) with aberrant GAGs (glycosaminoglycans) has been observed in these patients [5,6].

PGs are found predominantly either in the ECM (extracellular matrix) or associated with the cell surface of most eukaryotic cells. Most often, PGs act as molecular organizers of the ECM and as promoters of cell adhesion [7]. They also fulfil a variety of other biological functions, such as growth modulation, ionic filtration and biomechanical lubrication [811]. Their biological roles are closely related to the presence and the composition of their GAG chains, which contain binding sites for various growth factors, cytokines, morphogens, enzymes and other signalling molecules [1215]. Decorin is classified as a small PG, with a core protein of approx. 36 kDa and monoglycosylated with a DS (dermatan sulfate) chain; defective glycosylation of this PG is involved in the progeroid variant of EDS [5,6]. Recently, structural alteration of HS (heparan sulfate) chains has also been described and is associated with the altered wound-repair phenotype of EDS patients [16].

GAG chain initiation begins with the formation of a tetrasaccharide linkage region (GlcAβ1–3Galβ1–3Galβ1–4Xylβ1) bound via the xylose residue to serine residues of the core protein [17]. The biosynthesis of this region is catalysed, step by step, by distinct glycosyltransferases expressed in the Golgi complex. From this linker tetrasaccharide, the sugar chains are extended by addition of two alternating monosaccharides, an amino-sugar and a glucuronic acid [18]. The reaction leads to the formation of HP (heparin)/HS when the aminosugar is N-acetyl-glucosamine, and to CS (chondroitin sulfate) and DS when it is N-acetyl-galactosamine. The progeroid variant of EDS is caused by defects in β1,4-GalT7 (β1,4-galactosyltransferase 7), which catalyses the transfer of the first galactose residue to the xylose residue of the tetrasaccharide linkage region [2,3]. Patients with the autosomal recessive progeroid variant of EDS present with mutations in the coding sequence of the B4GALT7 gene. To date, from these patients, three mutations of B4GALT7, corresponding to A186D, L206P and R270C, have been identified [4,1921], which are associated with either reduction or loss of activity. However, the structural basis of these changes remains to be elucidated, although a three-dimensional structure of the catalytic domain of the β1,4-GalT7 from Drosophila in the presence of UDP has been solved recently [22]. Having access to the kinetic parameters (i.e. the kcat and Km values) brings information, but these should be analysed in conjunction with the thermodynamic parameters of substrate binding in order to interpret the consequences of amino acid substitutions. Recently, we have validated a ITC (isothermal titration calorimetry) method as a tool for detecting and quantifying the interactions between the substrates and the binding sites of the wild-type β1,4-GalT7 [23].

In the present study, mutated β1,4-GalT7s were expressed as soluble MBP (maltose-binding protein) fusions in which the N-terminal cytoplasmic domain, the transmembrane segment and the stem region, which precede the catalytic domain and correspond to amino acids 1–81, were deleted. The kinetic and thermodynamic parameters of all of the mutated β1,4-GalT7s were determined. The consequences of the mutations were also investigated ex vivo by evaluating the priming efficiency of xylosides on GAG chain initiation.

MATERIALS AND METHODS

Cloning and expression of human β1,4-GalT7s in CHO618 cells

The human B4GALT7 sequence fused with a c-Myc tag was subcloned into pcDNA3.1 (Invitrogen) as described previously [23]. The resulting pcDNA-β1,4-GalT7 was then used as a template to construct the expression plasmids pcDNA-A186D, -L206P and -R270C by site-directed mutagenesis using the QuikChange™ site-directed mutagenesis kit (Stratagene).

To express membrane proteins, β1,4-GalT7-deficient CHO618 cells (A.T.C.C. number CRL2241) [24] were cultured in complete DMEM (Dulbecco's modified Eagle's medium)/Ham's F12 (Gibco) supplemented with 10% (v/v) FBS (fetal bovine serum), 1 mM glutamine and 1% (w/v) penicillin/streptomycin. The pcDNA expression plasmids (10 μg) were transfected into CHO618 cells on 100-mm-diameter plates using Exgen 500 (Euromedex) according to the manufacturer's recommendations. At 2 days after transfection, cells were harvested (1000 g for 5 min) and resuspended in Hepes/sucrose buffer (5 mM Hepes, pH 7.4, and 0.25 M sucrose) and briefly sonicated (1× 5 s). The protein concentration of the resulting membrane fraction was determined by the method of Bradford [25]. The expression level of the wild-type, A186D, L206P, L206A and R270C β1,4-GalT7–Myc proteins was evaluated by Western blotting using an anti-Myc antibody (Invitrogen) according to the manufacturer's recommendations; quantification was performed by densitometry using a scanner employing NIH Image software.

Cloning, expression and purification of soluble forms of the human MBP–β1,4-galactosyltransferase 7

Human β1,4-GalT7s were expressed as MBP-fused proteins. To express the entire protein fused with the MBP, the β1,4-GalT7 sequence was subcloned into the expression vector pMALc2x (New England Biolabs) and used as a template to construct the pMALc2x-L206Paa1 expression plasmid by site-directed mutagenesis, performed as described above. To express the soluble catalytic domain of β1,4-GalT7, a truncated form of the β1,4-GalT7 DNA sequence, in which the nucleotides encoding the 81 N-terminal amino acids were deleted, was subcloned into pMALc2x as described previously [23] and then used as a template to construct the expression plasmids pMALc2x-A186Daa82, -L206Paa82 and -R270Caa82 by site-directed mutagenesis. The expression of the soluble MBP-fused proteins in Escherichia coli BL21(DE3) cells (Invitrogen) and the purification on to an amylose column were performed as described previously [23]. The concentration of the purified fusion proteins was determined by fluorescence using a 2100 Bioanalyser (Agilent), according to the manufacturer's recommendations.

β1,4-Galactosyltransferase activity

β1,4-Galactosyltransferase activity was determined as described previously [23] in the presence of 15–30 μg of total proteins containing membrane β1,4-GalT7 (i.e. 0.3–0.6 g of total proteins/l) or 2–10 pmol of purified truncated β1,4-GalT7 (i.e. 0.04–0.2 μM), and various concentrations of 4MU-Xyl (4-methylumbelliferyl-β-D-xylopyranoside) or pNP-Xyl (4-nitrophenyl-β-D-xylopyranoside) (0.02–10 mM 4MU-Xyl or pNP-Xyl in the presence of 10 mM UDP-Gal) and UDP-Gal (0.025–20 mM UDP-Gal in the presence of 10 mM 4MU-Xyl or pNP-Xyl). The estimation of the kinetic parameters was achieved by the incubation of at least ten different concentrations of substrate. Incubations were performed in duplicates. The results were fitted by non-linear regression to the Michaelis–Menten model using the curvefitter program of Sigmaplot™ 9.0.

Reaction products characterization

The characterization of the purified reaction products obtained after glycosylation of 4MU-Xyl in the presence of a saturating concentration of UDP-Gal by MBP-fused truncated wild-type, A186D, L206P and R270C β1,4-GalT7s (i.e. with a deletion of amino acids 1–81) was performed by NMR analysis, as described previously [23].

ITC

ITC experiments were carried out as described previously [23]. UDP (0.5–1 mM) and pNP-Xyl (40 mM) were used as donor and acceptor substrate analogues respectively. Ligand solutions were diluted in the final dialysis solution used for the protein to achieve the expected final concentration (0.5–1 mM UDP; 40 mM pNP-Xyl) and were then injected (35 injections of 4–10 μl each) into the reaction cell containing the degassed purified MBP-fused truncated proteins (1.4 ml, 30–50 μM). Thermodynamic parameters of the binding of the ligands to the proteins were obtained from individual experiments with three different batches of purified MBP–protein, and for each batch determination was performed in duplicate. The thermodynamic parameters n (stoichiometry), ΔH (enthalpy change) and Ka (association constant) were obtained by non-linear least-squares fitting of experimental data using the single-site model of the Origin 7.0 software package provided with the instrument. The Kd (dissociation constant), ΔG (free energy of binding) and ΔS (entropy change) were deduced from the Ka and ΔH values.

Labelling and isolation of radioactive GAGs

Metabolic labelling of GAGs with 35S was performed using a modification of a procedure published previously [24]. Cells were incubated with 0–1 μM 4MU-Xyl for 9 h in sulfate-free medium (Gibco) containing 10 μCi/ml Na2[35S]SO4 (PerkinElmer). GAGs were isolated from the medium using a modification of a procedure published previously [26].

Medium was harvested and digested with papain buffer (20 mM sodium phosphate, pH 6.8, containing 1 mM ethylenediamine tetra-acetic acid, 5 mM L-cysteine hydrochloride and 1 g/l papain) for 2 h at 60 °C. After boiling and centrifugation (10000 g for 20 min), the supernatant was precipitated overnight at 4 °C with 3 volumes of potassium acetate (in 5% ethanol). After centrifugation (10000 g for 30 min), the pellet was resuspended in 0.2 M NaCl and then centrifuged (10000 g for 30 min). The supernatant was supplemented with 100 μg of chondroitin sulfate from shark cartilage (Sigma) and 5% cetylpyridinium chloride (Sigma) and then incubated for 2 h at 37 °C before centrifugation (10000 g for 30 min). After washing twice with NaCl (0.2 M, then 2.5 M), GAGs were resuspended in 0.2 M Tris/HCl, pH 8.0, and the associated radioactivity was counted with 4.5 ml of Ultima Gold™ liquid scintillation counting cocktail (PerkinElmer).

RESULTS

β1,4-Galactosyltransferase activity of the membrane β1,4-GalT7s

To address the in vivo consequences of the mutations on the enzyme function, the membrane-bound wild-type, A186D, L206P, L206A and R270C β1,4-GalT7s were overexpressed in CHO618 cells. The overexpression of each recombinant protein in membrane fractions was confirmed by Western blot using an anti-Myc antibody. All of the proteins were expressed and had the expected molecular mass of approx. 37 kDa (Figure 1). The expression levels of the wild-type, A186D, L206P L206A and R270C β1,4-GalT7s, as quantified by densitometry, were similar, allowing the determination and the comparison of specific activities at saturating concentrations of 4MU-Xyl and UDP-Gal (referred to as kobs and expressed as nmol of products formed/mg of total protein per min).

Immunoblot analysis of the membrane fraction of CHO618 cells overexpressing wild-type and mutant β1,4-GalT7

Figure 1
Immunoblot analysis of the membrane fraction of CHO618 cells overexpressing wild-type and mutant β1,4-GalT7

Membrane proteins (10 μg) containing wild-type (lane 1), A186D (lane 2), L206P (lane 3), R270C (lane 4) and L206A (lane 5) β1,4-GalT7s were loaded. The β1,4-GalT7 proteins were probed with primary antibody directed against the c-Myc tag. The molecular mass in kDa is indicated on the left-hand side.

Figure 1
Immunoblot analysis of the membrane fraction of CHO618 cells overexpressing wild-type and mutant β1,4-GalT7

Membrane proteins (10 μg) containing wild-type (lane 1), A186D (lane 2), L206P (lane 3), R270C (lane 4) and L206A (lane 5) β1,4-GalT7s were loaded. The β1,4-GalT7 proteins were probed with primary antibody directed against the c-Myc tag. The molecular mass in kDa is indicated on the left-hand side.

The ability of each membrane protein to catalyse the transfer of galactose from donor (UDP-Gal) to acceptor (pNP-Xyl) substrate analogues was determined. No activity was detected with the membrane fraction containing the L206P β1,4-GalT7; the membrane fractions containing the wild-type β1,4-GalT7, and also the A186D, L206A and R270C β1,4-GalT7s were active. As shown in Table 1, the Km values of UDP-Gal for the membrane A186D and R270C β1,4-GalT7s were 8- and 2-fold higher than that for the wild-type respectively, whereas no difference was observed for the L206A mutant. The Km values with pNP-Xyl for A186D, L206A and R270C β1,4-GalT7s were 4-, 2.6- and 6-fold higher than that for the wild-type respectively. Finally, the kobs values for A186D and R270C β1,4-GalT7s were 3- and 1.5-fold lower respectively compared with the wild-type, whereas no difference was observed for the L206A mutant.

Table 1
Apparent kinetic parameters of the membrane wild-type and mutated β1,4-GalT7s expressed in CHO618 cells towards donor (UDP-Gal) and acceptor (pNP-Xyl) substrates

To determine the β1,4-galactosyltransferase activity, standard reactions were performed in 100 mM cacodylate buffer, pH 7.0, containing 10 mM MnCl2 and 0.3–0.6 g/l of total proteins containing membrane wild-type or mutant β1,4-GalT7. The kobs values determined at saturating concentration of both substrates were evaluated after normalization of the β1,4-GalT7 expression level by immunoquantification by Western blotting as described in the Material and methods section. Results are the means±2 S.D. for three individual determinations. *P< 0.05 compared with β1,4-GalT7; –, activity not detected.

  Donor substrate Acceptor substrate 
  UDP-Gal pNP-Xyl 
Enzyme kobs (nmol/mg of total protein per min) Km (mM) kobs/Km (μl/mg of total protein per min) Km (mM) kobs/Km (μl/mg of total protein per min) 
β1,4-GalT7 8.0±1.5 0.29±0.01 27.6 0.85±0.07 9.4 
A186D 2.5±0.2 2.41±0.48* 1.0 3.76±0.57* 0.7 
L206P – – – – – 
L206A 7.7±0.6 0.31±0.03 24.8 2.29±0.01* 3.4 
R270C 5.5±0.9 0.49±0.12* 11.2 5.01±0.11* 1.1 
  Donor substrate Acceptor substrate 
  UDP-Gal pNP-Xyl 
Enzyme kobs (nmol/mg of total protein per min) Km (mM) kobs/Km (μl/mg of total protein per min) Km (mM) kobs/Km (μl/mg of total protein per min) 
β1,4-GalT7 8.0±1.5 0.29±0.01 27.6 0.85±0.07 9.4 
A186D 2.5±0.2 2.41±0.48* 1.0 3.76±0.57* 0.7 
L206P – – – – – 
L206A 7.7±0.6 0.31±0.03 24.8 2.29±0.01* 3.4 
R270C 5.5±0.9 0.49±0.12* 11.2 5.01±0.11* 1.1 

β1,4-Galactosyltransferase activity of the MBP-truncated β1,4-GalT7s

In order to determine the molecular basis of the reduced activity of the A186D and R270C β1,4-GalT7s, and of the loss of activity of the L206P β1,4-GalT7, the mutated β1,4-GalT7s were expressed and purified as soluble truncated [lacking amino acids 1–81 and designated WT (wild-type) or mutant aa82] MBP fusion proteins, as performed previously for the wild-type enzyme [23]. The purified MBP–R270Caa82 (Figure 2, lane 1) and MBP–WTaa82 (Figure 2, lane 4) migrated as a single band, with an apparent molecular mass of 70 kDa, consistent with the calculated molecular mass of truncated β1,4-GalT7 fused with MBP. Analysis of purified MBP–A186Daa82 (Figure 2, lane 3) and MBP–L206Paa82 (Figure 2, lane 2) revealed the presence of a major band (~65% and 90% of purity respectively) with the same electrophoretic mobility as the MBP–WTaa82, along with several other bands with lower molecular mass. These additional bands were probably due to proteolysis, as they were also recognized by the anti-MBP antibodies (results not shown).

Analysis of purified recombinant human wild-type and mutant MBP–β1,4-GalT7 proteins by SDS/PAGE

Figure 2
Analysis of purified recombinant human wild-type and mutant MBP–β1,4-GalT7 proteins by SDS/PAGE

Human wild-type and mutant β1,4-GalT7s were expressed in E. coli as proteins fused with MBP, and then were purified on an amylose column as described in the Material and methods section. The purified proteins (2 μg) MBP–R270Caa82 (lane 1), MBP–L206Paa82 (lane 2), MBP–A186Daa82 (lane 3), MBP–WTaa82 (lane 4), MBP–L206Paa1 (lane 6) and MBP–WTaa1 (lane 7) and protein markers (at 25, 37, 75, 100 and 150 kDa; lane 5) were analysed by SDS/PAGE (10% gels) and gels were stained with Coomassie Blue.

Figure 2
Analysis of purified recombinant human wild-type and mutant MBP–β1,4-GalT7 proteins by SDS/PAGE

Human wild-type and mutant β1,4-GalT7s were expressed in E. coli as proteins fused with MBP, and then were purified on an amylose column as described in the Material and methods section. The purified proteins (2 μg) MBP–R270Caa82 (lane 1), MBP–L206Paa82 (lane 2), MBP–A186Daa82 (lane 3), MBP–WTaa82 (lane 4), MBP–L206Paa1 (lane 6) and MBP–WTaa1 (lane 7) and protein markers (at 25, 37, 75, 100 and 150 kDa; lane 5) were analysed by SDS/PAGE (10% gels) and gels were stained with Coomassie Blue.

As shown previously [23], MBP–WTaa82 can transfer the galactose residue from UDP-Gal to β-xylosides with Km values of 0.23 mM, 1.27 mM and 0.27 mM for UDP-Gal, pNP-Xyl and 4MU-Xyl respectively (Table 2) similar to those of the membrane wild-type β1,4-GalT7 (at 0.29, 0.85 and 0.24 mM respectively). MBP–A186Daa82 and MBP–R270Caa82 were also able to transfer galactose from UDP-Gal to both of the acceptor substrates pNP-Xyl and 4MU-Xyl. Structural characterization of the reaction products by NMR analysis showed that MBP–A186Daa82 and MBP–R270Caa82 formed the same β1→4 linkage as MBP–WTaa82 between the galactose from the UDP-Gal donor substrate and the xylose residue from the acceptor substrate in the reaction product (results not shown). The Km values of UDP-Gal and pNP-Xyl for MBP–A186Daa82 were ~9- and 3-fold higher respectively than those obtained for MBP–WTaa82; for MBP–R270Caa82Km values were ~3- and 6-fold higher (Table 2). These increases in Km values were similar to those observed for the corresponding membrane proteins (for the A186D mutant, ~8- and 5-fold higher towards UDP-Gal and pNP-Xyl respectively; for the R270C mutant, ~2- and 6-fold higher). On the other hand, the kcat values measured for the reaction catalysed by MBP–A186Daa82 and MBP–R270Caa82 were similar to MBP–WTaa82 (Table 2).

Table 2
Kinetic parameters of purified MBP–A186Daa82, MBP–L206Paa82 and MBP–R270Caa82 towards donor substrate (UDP-Gal) and acceptor substrates (4MU-Xyl and pNP-Xyl)

Standard reactions were performed in 100 mM cacodylate buffer (pH 7.0) containing 10 mM MnCl2 and 0.04–0.2 μM of purified truncated MBP-proteins. Each data point represents the mean ±2 SD of three or more individual determinations. *P< 0.05 compared with the results obtained with MBP–WTaa82. Because the purities of the MBP–A186Daa82 and MBP–L206Paa82 can be estimated to be 90 and 65%, the kcat values are underestimated and rather correspond to ~66 and 100 min−1 respectively.

  Donor substrate Acceptor substrate 
  TUDP-Gal 4MU-Xyl pNP-Xyl 
Enzyme kcat (min−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1
WTaa82 66.6±1.2 0.23±0.07 289 0.27±0.06 247 1.27±0.07 79 
A186Daa82 60.0±3.6 2.07±0.67* 29 2.26±0.80* 26 4.25±0.30* 18 
L206Paa82 63.6±1.8 0.65±0.18* 98 2.99±0.61* 21 5.54±0.60* 13 
R270Caa82 63.0±6.6 0.63±0.17* 100 3.26±0.53* 19 7.93±0.19* 7.8 
  Donor substrate Acceptor substrate 
  TUDP-Gal 4MU-Xyl pNP-Xyl 
Enzyme kcat (min−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1
WTaa82 66.6±1.2 0.23±0.07 289 0.27±0.06 247 1.27±0.07 79 
A186Daa82 60.0±3.6 2.07±0.67* 29 2.26±0.80* 26 4.25±0.30* 18 
L206Paa82 63.6±1.8 0.65±0.18* 98 2.99±0.61* 21 5.54±0.60* 13 
R270Caa82 63.0±6.6 0.63±0.17* 100 3.26±0.53* 19 7.93±0.19* 7.8 

MBP–L206Paa82, in contrast with membrane L206P β1,4-GalT7, which displayed no activity, was able to transfer the galactose residue from UDP-Gal to the acceptor substrate generating the same β1→4 linkage product as wild-type β1,4-GalT7 (results not shown). The Km values with UDP-Gal and pNP-Xyl were ~3- and 4-fold higher respectively than those obtained for MBP–WTaa82. The kcat value of MBP–L206Paa82 was similar to that of MBP–WTaa82 (at 60 min−1; Table 2). Because of the partial purity of MBP–L206Paa82, the kcat value determined for MBP–L206Paa82 was certainly underestimated and can be considered to correspond to a value of ~100 min−1 (assuming a purity of ~65%).

β1,4-Galactosyltransferase activity of the MBP–WTaa1 and MBP–L206Paa1 β1,4-GalT7

To better understand why the L206P substitution led to loss of activity for membrane L206P β1,4-GalT7 despite MBP-truncated L206P β1,4-GalT7 being active, the entire β1,4-GalT7 (amino acids 1–327, designated aa1) and the corresponding L206P β1,4-GalT7, both fused with MBP, were produced in E. coli and purified as described above. SDS/PAGE analysis of purified MBP–WTaa1 (Figure 2, lane7) and MBP–L206Paa1 (Figure 2, lane 6) revealed the presence of a major band with an apparent molecular mass of 85 kDa. As reported previously, many other bands with lower molecular mass were recognized by the anti-MBP antibodies, suggesting a significant proteolysis of these recombinant proteins.

MBP–WTaa1 was active, but showed a 40-fold decrease in kcat compared with MBP–WTaa82. (1.7 min−1 for MBP-WTaa1, compared with 66.6 min−1), whereas Km values of acceptor and donor substrates were increased 4- and 3-fold respectively. The reason why the activity decreases when the N-terminal amino acids 1–81 are inserted between MBP and the catalytic domain remains to be determined. In contrast with MBP–WTaa1 and MBP–L206Paa82, no detectable activity was observed for MBP–L206Paa1 (i.e. kcat < 0.05 min−1) (Table 3).

Table 3
Kinetic parameters of the purified MBP–WTaa82, MBP–WTaa1, MBP–L206Paa82 and MBP–L206Paa1 expressed in E. coli towards UDP-Gal and towards 4MU-Xyl

Standard reactions were performed in 100 mM cacodylate buffer (pH 7.0) containing 10 mM MnCl2 and 0.04–1 μM of purified fused proteins. Kinetic parameters for 4MU-Xyl and UDP-Gal were obtained after the addition of increasing amounts of acceptor or donor substrates (0.02–10 mM of 4MU-Xyl in the presence of 10 mM of UDP-Gal; 0.025–10 mM of donor substrates in the presence of 10 mM 4MU-Xyl) to the standard reaction. Each data point represents the means ±2 S.D. of three individual determinations. nd, not determined.

  Donor substrate Acceptor substrate 
  UDP-Gal 4MU-Xyl 
Enzyme kcat (min−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1
WTaa82 66.6±1.2 0.23±0.07 289 0.27±0.06 247 
WTaa1 1.7±0.06 0.80±0.10 2.13 0.62±0.02 2.74 
L206Paa82 63.6±1.8 0.65±0.18 98 2.99±0.61 21 
L206Paa1 <0.05 nd nd nd nd 
  Donor substrate Acceptor substrate 
  UDP-Gal 4MU-Xyl 
Enzyme kcat (min−1Km (mM) kcat/Km (min−1·mM−1Km (mM) kcat/Km (min−1·mM−1
WTaa82 66.6±1.2 0.23±0.07 289 0.27±0.06 247 
WTaa1 1.7±0.06 0.80±0.10 2.13 0.62±0.02 2.74 
L206Paa82 63.6±1.8 0.65±0.18 98 2.99±0.61 21 
L206Paa1 <0.05 nd nd nd nd 

Thermodynamic characterization of the substrate binding to MBP–A186Daa82 and MBP–R270Caa82

As shown from the comparison of the Km values of the acceptor and donor substrates for the wild-type and mutant β1,4-GalT7s, the A186D and R270C mutations lead to an increase of the Km values for both substrates. It is therefore difficult to conclude with certainty whether Ala186 and Arg270 contribute to the donor or the acceptor sites, and thus to interpret the consequences of the substitutions at the structural level. Consequently, the thermodynamic parameters of the binding of donor and acceptor substrates to MBP–A186Daa82 and MBP–R270Caa82 were compared with MBP–WTaa82 by ITC. Exothermic binding profiles were obtained with all the purified proteins. Typical calorimetric profiles of the binding of substrate analogues to MBP–WTaa82 and MBP–A186Daa82 are presented in Figure 3. Similar profiles were obtained for the MBP–R270Caa82.

Calorimetric profiles of the donor and acceptor substrate analogues binding to MBP–A186Daa82, compared with MBP–WTaa82

Figure 3
Calorimetric profiles of the donor and acceptor substrate analogues binding to MBP–A186Daa82, compared with MBP–WTaa82

(A) Binding of UDP to MBP–WTaa82 and MBP–A186Daa82. The reaction cell contained MBP–WTaa82 (1.4 ml, 30 μM; black line) or MBP–A186Daa82 (1.4 ml, 50 μM; grey line) in 20 mM Mops buffer containing 150 mM NaCl and 5 mM MnCl2. The syringe contained 0.5 mM or 1 mM UDP in the same buffer, for MBP–WTaa82 and MBP–A186Daa82 respectively. Upper panel: raw calorimetric data obtained from the injection of UDP at 4-min intervals. Lower panel: the integrated binding isotherm with the experimental points and the corresponding best fit, providing values for thermodynamic parameters. (B) Binding of pNP-Xyl to MBP–WTaa82 and MBP–A186Daa82. The reaction cell contained MBP–WTaa82 (1.4 ml, 30 μM; black line) or MBP–A186Daa82 (1.4 ml, 25 μM; grey line) in 20 mM Mops buffer containing 150 mM NaCl, 5 mM MnCl2 and 5 mM UDP. The syringe contained 40 mM pNP-Xyl in the same buffer. Upper panel: raw calorimetric data obtained from the injection of pNP-Xyl at 4-min intervals. Lower panel: the integrated binding isotherm with the experimental points and the corresponding best fit, providing values for thermodynamic parameters.

Figure 3
Calorimetric profiles of the donor and acceptor substrate analogues binding to MBP–A186Daa82, compared with MBP–WTaa82

(A) Binding of UDP to MBP–WTaa82 and MBP–A186Daa82. The reaction cell contained MBP–WTaa82 (1.4 ml, 30 μM; black line) or MBP–A186Daa82 (1.4 ml, 50 μM; grey line) in 20 mM Mops buffer containing 150 mM NaCl and 5 mM MnCl2. The syringe contained 0.5 mM or 1 mM UDP in the same buffer, for MBP–WTaa82 and MBP–A186Daa82 respectively. Upper panel: raw calorimetric data obtained from the injection of UDP at 4-min intervals. Lower panel: the integrated binding isotherm with the experimental points and the corresponding best fit, providing values for thermodynamic parameters. (B) Binding of pNP-Xyl to MBP–WTaa82 and MBP–A186Daa82. The reaction cell contained MBP–WTaa82 (1.4 ml, 30 μM; black line) or MBP–A186Daa82 (1.4 ml, 25 μM; grey line) in 20 mM Mops buffer containing 150 mM NaCl, 5 mM MnCl2 and 5 mM UDP. The syringe contained 40 mM pNP-Xyl in the same buffer. Upper panel: raw calorimetric data obtained from the injection of pNP-Xyl at 4-min intervals. Lower panel: the integrated binding isotherm with the experimental points and the corresponding best fit, providing values for thermodynamic parameters.

In all the mutated enzymes, the stoichiometry of UDP binding was between 0.4 and 0.6, which is probably due to the possible dimeric state of the enzyme, as described previously [23]. The binding of donor substrate analogue to wild-type, A186D and R270C β1,4-GalT7s was driven by an enthalpy change, which was partially counterbalanced by an unfavourable entropy change (Table 4). Binding of UDP to MBP–R270Caa82 showed similar thermodynamic parameters as with MBP–WTaa82, suggesting that there was no modification of UDP binding by the R270C mutation (Table 4). On the contrary, binding of UDP to the enzyme was significantly modified by the replacement of the Ala186 residue by aspartate (Figure 3A), with a significant decrease in ΔH that was not fully offset by a decrease in ΔS. Therefore the dissociation constant was 4-fold higher than that of MBP–WTaa82 (Table 4).

Table 4
Thermodynamic parameters for the binding of UDP to the MBP–WTaa82, MBP–A186Daa82 and MBP–R270Caa82

ITC experiments were carried out at 30 °C in 20 mM Mops buffer, pH 7.0, containing 150 mM NaCl, 5 mM MnCl2 and 5 mM UDP using a VP-ITC Microcalorimeter (Micro Cal). Results are the means±2 S.D. for three individual determinations. *P< 0.01 compared with the wild-type enzyme.

Enzyme ΔG (cal/mol) ΔH (kcal/mol) ΔS (cal/mol per K) TΔS (kcal/mol) Kd (mM) 
WTaa82 −6880±160 −22.9±4.8 −54 16.07 9.0 
A186Daa82 −6030±40  −8.1±1.3* −6.9* 2.08* 38* 
R270Caa82 −6580±70 −21.0±2.3 −48.5 14.44 15 
Enzyme ΔG (cal/mol) ΔH (kcal/mol) ΔS (cal/mol per K) TΔS (kcal/mol) Kd (mM) 
WTaa82 −6880±160 −22.9±4.8 −54 16.07 9.0 
A186Daa82 −6030±40  −8.1±1.3* −6.9* 2.08* 38* 
R270Caa82 −6580±70 −21.0±2.3 −48.5 14.44 15 

Similarly, the consequences of the A186D and R270C mutations for the binding of the acceptor substrate (pNP-Xyl) to the β1,4-GalT7 were analysed by ITC (Figure 3B). The pNP-Xyl substrate was used because of the limited solubility of 4MU-Xyl. Because UDP is required to promote the binding of the acceptor substrate to the enzyme, we explored the binding of pNP-Xyl in the presence of UDP and Mn2+, as described previously [23]. Since the Kd value obtained for the binding of pNP-Xyl to the wild-type protein was higher than the protein concentration, the c-value (c=[Prot]total/Kd) was lower than one. Under low c-value conditions, both ΔG and ΔH can be measured with confidence provided that at least 80% saturation is achieved and that the stoichiometry (n) is known, as described by Turnbull and Danaras [27]. For MBP–WTaa82 and MBP–A186Daa82, but not for MBP–R270aa82, it was possible to reach saturating ligand concentration. Therefore we used the stoichiometry value determined independently by the active-site titration experiments with UDP under higher c-value conditions for fitting the low c-value titrations. Consequently, a dissociation constant of 4.6 mM was measured for the binding of pNP-Xyl to MBP–WTaa82 (Table 5). Its binding to MBP–WTaa82 was driven by an enthalpy change (–11.8 kcal/mol; 1 kcal≈4.184 kJ) with an unfavourable entropy change. Binding of pNP-Xyl to MBP–A186Daa82 was relatively similar to that of MBP–WTaa82, with an estimated dissociation constant of approx. 5.8 mM (Figure 3B). The magnitude of ΔH was higher (by approx. 13 kcal/mol) than that obtained for MBP–WTaa82, yet fully offset by an equivalent increase in entropy, resulting in a similar affinity to the wild-type. In contrast with MBP–A186Daa82, the binding of pNP-Xyl to MBP–R270Caa82 was difficult to detect by ITC and the saturation was not reached, preventing an accurate determination of the thermodynamic parameters. Nevertheless, the estimated binding affinity seemed to be very weak (Kd>10 mM).

Table 5
Thermodynamic parameters for the binding of pNP-Xyl to the MBP–WTaa82, MBP–A186Daa82 and MBP–R270Caa82

ITC experiments were carried out at 30 °C in 20 mM Mops buffer, pH 7.0, containing 150 mM NaCl, 5 mM MnCl2 and 5 mM UDP using a VP-ITC Microcalorimeter (Micro Cal). Results are the means±2 S.D. for three individual determinations. nd, not determined.

Enzyme ΔG (cal/mol) ΔH (kcal/mol) ΔS (cal/mol per K) TΔS (kcal/mol) Kd (mM) 
WTaa82 −3180±10 −11.8±0.2 −29 8.60 4.6 
A186Daa82 −3040±80 −24.8±5.4 −73 21.83 5.8 
R270Caa82 nd nd nd nd >10 
Enzyme ΔG (cal/mol) ΔH (kcal/mol) ΔS (cal/mol per K) TΔS (kcal/mol) Kd (mM) 
WTaa82 −3180±10 −11.8±0.2 −29 8.60 4.6 
A186Daa82 −3040±80 −24.8±5.4 −73 21.83 5.8 
R270Caa82 nd nd nd nd >10 

Stimulation of the β1,4-GalT7-dependent GAG synthesis by glycosides

To identify the consequence of the mutations on GAG synthesis, we overexpressed membrane wild-type, A186D, L206P and R270C β1,4-GalT7s in CHO618 cells and evaluated the GAG synthesis level after transfection. Immunoblot analysis showed that all β1,4-GalT7s were produced at a similar level (Figure 1). Analysis of the GAG synthesis rate at 48 h after transfection led to a 4-fold increase of 35S-incorporation compared with mock-transfected cells (Figure 4), indicating that overexpression of β1,4-GalT7 significantly enhanced GAG synthesis. Surprisingly, overexpression of A186D or R270C β1,4-GalT7 stimulated GAG synthesis at the same level as wild-type β1,4-GalT7. Only L206P β1,4-GalT7 failed to increase GAG synthesis, and those cells had a rate of synthesis similar to that of mock-transfected cells (Figure 4).

Restoration of GAG synthesis in CHO618 cells expressing wild-type or mutated β1,4-GalT7

Figure 4
Restoration of GAG synthesis in CHO618 cells expressing wild-type or mutated β1,4-GalT7

CHO618 cells expressing wild-type (WT), A186D or L206P or R270C β1,4-GalT7, were labelled at 2 days after transfection, with Na2[35S]SO4 (10 μCi/ml) for 9 h. The labelled [35S]GAGs were isolated by successive potassium acetate and cetylpyridinium chloride precipitations. β1,4-GalT7 expression was analysed by Western blotting using anti-Myc antibodies, as described in the Materials and methods section. **, P < 0.01 compared with wild-type.

Figure 4
Restoration of GAG synthesis in CHO618 cells expressing wild-type or mutated β1,4-GalT7

CHO618 cells expressing wild-type (WT), A186D or L206P or R270C β1,4-GalT7, were labelled at 2 days after transfection, with Na2[35S]SO4 (10 μCi/ml) for 9 h. The labelled [35S]GAGs were isolated by successive potassium acetate and cetylpyridinium chloride precipitations. β1,4-GalT7 expression was analysed by Western blotting using anti-Myc antibodies, as described in the Materials and methods section. **, P < 0.01 compared with wild-type.

To further understand the consequences of the A186D and R270C mutations on GAG synthesis, we investigated the ability of 4MU-Xyl to prime the β1,4-GalT7-dependent GAG synthesis in CHO618 cells after transfection with the β1,4-GalT7s. At 2 days after transfection, cells were incubated for 9 h with 4MU-Xyl, and the ability of glycoside to prime GAG synthesis in transfected cells was evaluated. The kinetics of GAG synthesis are presented in Figure 5. Stimulation of GAG synthesis was not observed in CHO618 cells overexpressing L206P β1,4-GalT7, whatever the 4MU-Xyl concentration. In CHO618 cells overexpressing wild-type, A186D and R270C β1,4-GalT7s, the stimulation of GAG synthesis was a linear function of the 4MU-Xyl concentration. However, the GAG synthesis was stimulated less in CHO618 cells overexpressing the A186D and R270C β1,4-GalT7s. Therefore we determined the slope of the linear regression, which is representative of the stimulation of GAG synthesis. In cells overexpressing wild-type β1,4-GalT7, the slope was 0.04; with the A186D and R270C β1,4-GalT7s, the slopes markedly decreased to 0.03 and 0.02 respectively.

Stimulation kinetics of GAG synthesis by 4MU-Xyl in CHO618 cells expressing wild-type or mutated β1,4-GalT7

Figure 5
Stimulation kinetics of GAG synthesis by 4MU-Xyl in CHO618 cells expressing wild-type or mutated β1,4-GalT7

CHO618 cells expressing wild-type (WT), A186D or L206P or R270C β1,4-GalT7, then, 2 days after transfection, were incubated with various concentrations of 4MU-Xyl (0–1 μM) and labelled, at the same time, with Na2[35S]SO4 (10 μCi/ml) for 9 h. The labelled [35S]GAGs were isolated by successive potassium acetate and cetylpyridinium chloride precipitations.

Figure 5
Stimulation kinetics of GAG synthesis by 4MU-Xyl in CHO618 cells expressing wild-type or mutated β1,4-GalT7

CHO618 cells expressing wild-type (WT), A186D or L206P or R270C β1,4-GalT7, then, 2 days after transfection, were incubated with various concentrations of 4MU-Xyl (0–1 μM) and labelled, at the same time, with Na2[35S]SO4 (10 μCi/ml) for 9 h. The labelled [35S]GAGs were isolated by successive potassium acetate and cetylpyridinium chloride precipitations.

DISCUSSION

β1,4-GalT7 catalyses a limiting step in the biosynthesis of GAG chains [28,29]. Patients with mutations in the B4GALT7 gene can suffer from skin, cartilage and bone alterations [2]. Only three mutations of the B4GALT7 gene (A186D, L206P and R270C) have been identified to date [4,1921], leading to only two different genotypes (compound heterozygosity for A186D and L206P, and homozygosity for R270C) [4,19,20]. Surprisingly, whatever the genotype, patients with the progeroid variant of EDS showed the same premature aging phenotype. The aim of the present study was to elucidate the biochemical and structural consequences of the B4GALT7 mutations on the enzyme function using ITC as a method of choice for probing interactions at the substrate-binding sites of the enzyme [23,3032].

As the wild-type β1,4-GalT7 catalytic domain (from amino acid 82 to the end of the protein) fused with the MBP is highly soluble and shows Km values identical with those of the wild-type membrane form [23], we decided to express the catalytic domain of the mutated proteins fused with MBP. However, before exploring the physicochemical properties of the MBP-fused truncated enzymes, we systematically compared the ability to transfer the galactose residue from UDP-Gal to an acceptor substrate analogue with that of the corresponding enzyme expressed in Golgi membranes of eukaryotic cells.

The MBP–A186Daa82 is able to transfer the galactose residue from UDP-Gal to the acceptor substrate by generating the same β1→4 linkage product as wild-type β1,4-GalT7, as described previously [23]. However, the catalytic efficiencies of MBP–A186Daa82 with the UDP-Gal donor and 4MU-Xyl and pNP-Xyl acceptors are 10-, 10- and 4-fold lower respectively than those for MBP–WTaa82, due to higher Km values. Almeida et al. [19] have also reported an increase in the Km value for the donor substrate (4-fold higher), whereas no significant change was observed for the binding of the acceptor substrate. The ITC results clearly demonstrate that MBP–A186Daa82 was able to bind donor and acceptor substrate analogues. However, the Kd value of UDP for MBP–A186Daa82 was 4-fold higher than that of the wild-type enzyme, whereas that of pNP-Xyl was similar to wild-type. These results suggest that Ala186 contributes directly or, more probably, indirectly to the binding donor site [22]. Interestingly, the ΔH for the binding of UDP to MBP–A186Daa82 was significantly reduced, by approx. 15 kcal/mol compared with wild-type. In our previous recent study, a major contribution of the β-phosphate, and more moderately the α-phosphate, of the donor substrate to the binding to β1,4-GalT7 was demonstrated [23]. Therefore the carboxy side chain of Asp186 in MBP–A186Daa82 could induce repulsive interactions with the phosphate groups of UDP, and prevent the formation of a network of interactions between the donor substrate analogue and the protein. However, Ala186, which is located in the hydrophobic core of the enzyme and surrounded by bulky aromatic residues [22,33], is not likely to be essential for binding UDP in the wild-type enzyme. This is also in accordance with the fact that Ala186 corresponds to Ser274 in β1,4-GalT1, which is an uncharged hydrophilic residue, capable of forming hydrogen bonds, instead of being hydrophobic [19].

Little biochemical information is currently available on the R270C β1,4-GalT7 mutation. This mutation was reported previously to decrease, by approx. 3-fold, the β1,4-galactosyltransferase activity in β1,4-GalT7-R270C fibroblasts compared with control fibroblasts [6]. In the present study we determined that MBP–R270Caa82 was active, but the kcat/Km ratio towards 4MU-Xyl was particularly affected (11-fold lower compared with the wild-type enzyme) due to a higher Km value; in contrast the kcat/Km with UDP-Gal was similar to the wild-type enzyme. The ITC results clearly demonstrate that Arg270 does not interact directly with the donor substrate analogue, as UDP binding was unaffected with MBP–R270Caa82. In contrast, the binding of the acceptor substrate analogue to MBP–R270Caa82 was particularly modified and the binding isotherm did not reach saturation, preventing accurate determination of thermodynamic parameters of binding. The drastic loss of pNP-Xyl binding with MBP–R270Caa82 suggests that this positively charged residue is critical for acceptor substrate binding. In the absence of a three-dimensional structure of the β1,4-GalT7–acceptor–substrate complex, it is difficult to explain the requirement of Arg270 for the binding of acceptor substrate. Sequence comparison of the human β1,4-galactosyltransferase family members showed sequence variations in their oligosaccharide-binding site, which would account for their respective preferences for different oligosaccharides [34,35]. We suggest that Arg270, which is a lysine residue in the Drosophila enzyme [22], and is located on a flexible loop, may interact directly with the acceptor substrate by a salt bridge or by hydrogen bonding interactions, or may be a structural element in the organization of the acceptor active site.

The replacement of Leu206 with a proline residue was reported previously as inactivating the enzyme [19,20]. In these studies, the L206P β1,4-GalT7 was expressed either as a truncated protein, lacking the first 53 N-terminal amino acids, and fused to Protein A [20] or as a secreted form lacking the first 63 N-terminal amino acids. Similar results were found in the present study when the full-length L206P protein was expressed either as a membrane protein or as a soluble protein fused with the MBP. As can be determined from the amino acid alignment, Leu206 is conserved in all mammalian β1,4-GalT7s and in the related Caenorhabditis elegans sqv-3 [36], whereas it is a valine residue in the Drosophila enzyme [22]. Leu206 or its equivalent valine is located in the hydrophobic core of the protein [22,37]. Therefore, owing to the location of Leu206 and the fact that the L206A mutant shows kinetic parameters similar to the wild-type, Leu206 probably has no role in the catalytic efficiency of the enzyme. As suggested by Almeida et al. [19] and Okajima et al. [20], a proline substitution at position 206 could drastically change the structure of the protein, and therefore would be responsible for the loss of activity. In the present study, however, when the L206P β1,4-GalT7 was expressed as an MBP-fusion protein, with a truncation of the first 81 N-terminal amino acids, a catalytic activity as efficient as that of its wild-type counterpart was observed. Such a result indicates that the replacement of Leu206 by a proline residue does not modify the fold of the catalytic domain, at least under a β1,4-GalT7 form in which the N-terminal cytoplasmic domain, the transmembrane segment and the stem region, which precedes the catalytic domain, have been deleted. Therefore our results, combined with those of Okajima et al. [20], which showed that L206P β1,4-GalT7 is inactive when fused to Protein A and in which only the stem is present, suggest that it is the combination of both the proline residue at position 206 and at least the stem that leads to loss of activity, and that this is probably due to a modification of the conformation of the protein and, in turn, of the active site.

Similar clinical symptoms and aberrant GAG substitution of PGs were observed in patients with compound heterozygosity for the A186D and L206P mutations, and patients with homozygosity for the R270C mutation [5,6,16,19]. Nevertheless, the molecular consequences of the three reported mutations of the β1,4-GalT7 identified in EDS patients on the enzyme function are clearly different. L206P β1,4-GalT7 is inactive due to altered conformation or altered substrate binding, and the L206P β1,4-GalT7 in CHO618 cells is unable to synthesize GAG chains. In contrast, A186D and R270C β1,4-GalT7s display activity. However, the A186D and R270C mutations lead to a decrease of the β1,4-GalT7 efficiency due to a lowered binding of donor substrate and acceptor substrate respectively. In the present study, the measurement of 35S incorporation into GAG, reflecting the biosynthesis rate, was surprisingly similar in wild-type, A186D and R270C β1,4-GalT7 when the enzymes were overexpressed in CHO618 cells. Therefore an evaluation of the ability of each mutated protein to prime the GAG biosynthesis ex vivo was carried out. Since the requirement for xylosylated core proteins can be bypassed by using synthetic β-D-xylosides containing hydrophobic aglycones [3840], we used 4MU-Xyl as a primer for GAG formation. Priming by β-D-xylosides depended linearly on the 4MU-Xyl concentration, suggesting that the acceptor substrate (i.e. the xylosylated-core protein) is a limiting factor of GAG synthesis. In cells overexpressing wild-type protein, stimulation of GAG synthesis was 10-fold higher when 1 μM 4MU-Xyl was used compared with the condition without glycoside; it was 9- and 6-fold higher in the A186D- and R270C-overexpressing cells respectively. The ability of the A186D β1,4-GalT7 to prime GAG synthesis was thus weakly affected compared with that of wild-type enzyme. If we consider that the availability of acceptor substrate is a limiting factor for GAG synthesis, this result is consistent with the ability of the A186D β1,4-GalT7 to bind the acceptor substrate. CHO618 cells stably expressing the A186D β1,4-GalT7 have also been described as being able to decorate human decorin with GAG chains; however, with a considerably decreased efficiency [5]. Because of the limited consequences of the A186D mutation on GAG synthesis, it would be interesting to know the phenotype of individuals homozygous for A186D in that respect; only the phenotype of individuals with compound heterozygosity for A186D and L206P is known. Because L206P β1,4-GalT7 is inactive, this phenotype is certainly essentially associated with the L206P mutation of the β1,4-GalT7. Moreover, it is likely that homozygosity for L206P is not a viable genotype. Indeed, the transfer of the first galactose on the tetrasaccharide linker region of the GAG chains is strictly catalysed by the β1,4-GalT7 and cannot be bypassed by another glycosyltransferase [24]. In contrast with A186D β1,4-GalT7, the ability of R270C β1,4-GalT7 to prime GAG synthesis was significantly affected compared with the wild-type enzyme. The lower priming efficacy with 4MU-Xyl for the R270C β1,4-GalT7 when overexpressed in CHO618 cells may be due to the replacement of Arg270 by a cysteine residue that modifies the binding affinity of the acceptor substrate. The reduced ability of the R270C β1,4-GalT7 to prime GAG synthesis can explain the defective glycosylation of decorin and biglycan in individuals homozygotic for the R270C mutation [6,16]. On the other hand, in β1,4-GalT7-R270C fibroblasts, a reduced epimerization of CS/DS GAG chains [6] and a modified pattern of HS chains [16] was observed. Thus there is evidence that the β1,4-GalT7 genetic defects, along with a decrease in the GAG biosynthesis rate, also affect the quality and diversity of the chains, via epimerization and sulfation reactions, and hence the biological function of these substances.

We would like to give our thanks to A. Kriznik (UMR CNRS-Nancy Université 7214, Vandoeuvre-Les-Nancy, France) and O. Fabre (UMR CNRS-INPL 7568, Vandoeuvre-Les-Nancy, France) for ITC and NMR technical support respectively.

Abbreviations

     
  • CS

    chondroitin sulfate

  •  
  • DS

    dermatan sulfate

  •  
  • ECM

    extracellular matrix

  •  
  • EDS

    Ehlers–Danlos syndrome

  •  
  • GAG

    glycosaminoglycan

  •  
  • β1,4-GalT7

    β1,4-galactosyltransferase 7

  •  
  • HS

    heparan sulfate

  •  
  • ITC

    isothermal titration calorimetry

  •  
  • MBP

    maltose-binding protein

  •  
  • 4MU-Xyl

    4-methylumbelliferyl-β-D-xylopyranoside

  •  
  • PG

    proteoglycan

  •  
  • pNP-Xyl

    4-nitrophenyl-β-D-xylopyranoside

AUTHOR CONTRIBUTION

Sophie Rahuel-Clermont, Franck Daligault and Virginie Lattard designed the research. Sophie Rahuel-Clermont, Franck Daligault, Marie-Helene Piet, Sandrine Gulberti and Virginie Lattard performed the research. Sophie Rahuel-Clermont, Franck Daligault, Guy Branlant and Virginie Lattard analysed the results. Sophie Rahuel-Clermont, Franck Daligault, Patrick Netter, Guy Branlant, Jacques Magdalou and Virginie Lattard wrote the paper.

FUNDING

This work was supported by the Agence Nationale pour la Recherche [grant numer NT05–3_42251]; Region Lorraine; Communauté Urbaine du Grand Nancy; and Conseil Général de Meurthe et Moselle. ITC experiments on the VP-ITC apparatus were conducted in the ‘Service Commun de Biophysicochimie des Interactions Moléculaires’ of the University Henri Poincaré, Nancy Université, France.

References

References
1
Beighton
P.
De Paepe
A.
Steinmann
B.
Tsipouras
P.
Wenstrup
R. J.
Ehlers–Danlos syndromes: revised nosology, Villefranche, 1997. Ehlers-Danlos National Foundation (USA) and Ehlers-Danlos Support Group (UK)
Am. J. Med. Genet.
1998
, vol. 
77
 (pg. 
31
-
37
)
2
Kresse
H.
Rosthoj
S.
Quentin
E.
Hollmann
J.
Glössl
J.
Okada
S.
Tonnesen
T.
Glycosaminoglycan-free small proteoglycan core protein is secreted by fibroblasts from a patient with a syndrome resembling progeroid
Am. J. Hum. Genet.
1987
, vol. 
41
 (pg. 
436
-
453
)
3
Quentin
E.
Gladen
A.
Roden
L.
Kresse
H.
A genetic defect in the biosynthesis of dermatan sulfate proteoglycan: galactosyltransferase I deficiency in fibroblasts from a patient with a progeroid syndrome
Proc. Natl. Acad. Sci. U.S.A.
1990
, vol. 
87
 (pg. 
1342
-
1346
)
4
Faiyaz-Ul-Haque
M.
Zaidi
S. H.
Al-Ali
M.
Al-Mureikhi
M. S.
Kennedy
S.
Al-Thani
G.
Tsui
L. C.
Teebi
A. S.
A novel missense mutation in the galactosyltransferase-I (B4GALT7) gene in a family exhibiting facioskeletal anomalies and Ehlers–Danlos syndrome resembling the progeroid type
Am. J. Med. Genet.
2004
, vol. 
128A
 (pg. 
39
-
45
)
5
Götte
M.
Kresse
H.
Defective glycosaminoglycan substitution of decorin in a patient with progeroid syndrome is a direct consequence of two point mutations in the galactosyltransferase I (β4GalT-7) gene
Biochem. Genet.
2005
, vol. 
43
 (pg. 
65
-
77
)
6
Seidler
D. G.
Fayaz-Ul-Haque
M.
Hansen
U.
Yip
G. W.
Zaidi
S. H. E.
Teebi
A. S.
Kiesel
L.
Götte
M.
Defective glycosylation of decorin and biglycan, altered collagen structure, and abnormal phenotype of the skin fibroblasts of an Ehlers–Danlos syndrome patient carrying the novel Arg270Cys substitution in galactosyltransferase I (β4GalT-7)
J. Mol. Med.
2006
, vol. 
84
 (pg. 
583
-
594
)
7
Ruoslahti
E.
Yamaguchi
Y.
Proteoglycans as modulators of growth factor activities
Cell
1991
, vol. 
64
 (pg. 
867
-
869
)
8
Bandtlow
C. E.
Zimmermann
D. R.
Proteoglycans in the developing brain: new conceptual insights for old proteins
Physiol. Rev.
2000
, vol. 
80
 (pg. 
1267
-
1290
)
9
De Cat
B.
David
G.
Developmental roles of the glypicans
Semin. Cell Dev. Biol.
2001
, vol. 
12
 (pg. 
117
-
125
)
10
Knudson
C. B.
Knudson
W.
Cartilage proteoglycans
Semin. Cell Dev. Biol.
2001
, vol. 
12
 (pg. 
69
-
78
)
11
Rapraeger
A. C.
Molecular interactions of syndecans during development
Semin. Cell Dev. Biol.
2001
, vol. 
12
 (pg. 
107
-
116
)
12
Poole
A. R.
Proteoglycans in health and disease: structures and functions
Biochem. J.
1986
, vol. 
236
 (pg. 
1
-
14
)
13
Ruoslahti
E.
Proteoglycans in cell regulation
J. Biol. Chem.
1989
, vol. 
264
 (pg. 
13369
-
13372
)
14
Salmivirta
M.
Lidholt
K.
Lindahl
U.
Heparan sulfate: a piece of information
FASEB J.
1996
, vol. 
10
 (pg. 
1270
-
1279
)
15
Taipale
J.
Keski-Oja
J.
Growth factors in the extracellular matrix
FASEB J.
1997
, vol. 
11
 (pg. 
51
-
59
)
16
Götte
M.
Spillmann
D.
Yip
G. W.
Versteeg
E.
Echtermeyer
F. G.
Van Kuppevelt
T. H.
Kiesel
L.
Changes in heparan sulfate are associated with delayed wound repair, altered cell migration, adhesion and contractility in the galactosyltransferase I (β4GalT-7) deficient form of Ehlers–Danlos syndrome
Hum. Mol. Genet.
2008
, vol. 
17
 (pg. 
996
-
1009
)
17
Esko
J. D.
Zhang
L.
Influence of core protein sequence on glycosamino-glycan assembly
Curr. Opin. Struct. Biol.
1996
, vol. 
6
 (pg. 
663
-
670
)
18
Roden
L.
Lennarz
W. G.
The Biochemistry of Glycoproteins and Proteoglycans
1980
New York
Plenum
(pg. 
269
-
314
)
19
Almeida
R.
Levery
S. B.
Mandel
U.
Kresse
H.
Schwientek
T
Bennett
E. P.
Clausen
H.
Cloning and expression of a proteoglycan UDP-galactose–β-xylose β1,4-galactosyltransferase I: a seventh member of the human β4-galactosyltransferase gene family
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
26165
-
26171
)
20
Okajima
T.
Fukumoto
S.
Furukawa
K.
Urano
T.
Furukawa
K.
Molecular basis for the progeroid variant of Ehlers–Danlos syndrome: identification and characterization of two mutations in galactosyltransferase I gene
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
28841
-
28844
)
21
Furukawa
K.
Okajima
T.
Galactosyltransferase I is a gene responsible for progeroid variant of Ehlers–Danlos syndrome: molecular cloning and identification of mutations
Biochim. Biophys. Acta
2002
, vol. 
1573
 (pg. 
377
-
381
)
22
Ramakrishnan
B.
Qasba
P. K.
Crystal structure of the catalytic domain of Drosophila β1,4-galactosyltransferase-7
J. Biol. Chem.
2010
, vol. 
285
 (pg. 
15619
-
15626
)
23
Daligault
F.
Rahuel-Clermont
S.
Gulberti
S.
Branlant
G.
Netter
P.
Magdalou
J.
Lattard
V.
Thermodynamic insights into the structural basis governing the donor substrate recognition by human β1,4-galactosyltransferase 7
Biochem. J.
2009
, vol. 
418
 (pg. 
605
-
614
)
24
Esko
J. D.
Weinke
J. L.
Taylor
W. H.
Ekborg
G.
Roden
L.
Anatharamaiah
G.
Gawish
A.
Inhibition of chondroitin and heparan sulfate biosynthesis in Chinese hamster ovary cell mutants defective in galactosyltransferase I
J. Biol. Chem.
1987
, vol. 
262
 (pg. 
12189
-
12195
)
25
Bradford
M. M.
A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding
Anal. Biochem.
1976
, vol. 
72
 (pg. 
248
-
254
)
26
Ha
Y. W.
Jeon
B. T.
Moon
S. H.
Toyoda
H.
Toida
T.
Lindhardt
R. J.
Kim
Y. S.
Characterization of heparan sulfate from the unossified antler of Cervus elaphus
Carbohydr. Res.
2005
, vol. 
340
 (pg. 
411
-
416
)
27
Turnbull
W. B.
Danaras
A. H.
On the value of c: can low affinity systems be studied by isothermal titration calorimetry
J. Am. Chem. Soc.
2003
, vol. 
125
 (pg. 
14859
-
14866
)
28
Nakamura
Y.
Haines
N.
Chen
J.
Okajima
T.
Furukawa
K.
Urano
T.
Stanley
P.
Irvine
K. D.
Furukawa
K.
Identification of a Drosophila gene encoding xylosylprotein β4-galactosyltransferase
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
46280
-
46288
)
29
Takemae
H.
Ueda
R.
Okubo
R.
Nakato
H.
Izumi
S.
Saigo
K.
Nishihara
S.
Proteoglycan UDP-galactose–β-xylose β-1,4-galactosyltransferase I is essential for viability in Drosophila melanogaster
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
15571
-
15578
)
30
Dam
T. K.
Brewer
C. F.
Thermodynamic studies of lectin–carbohydrate interactions by isothermal titration calorimetry
Chem. Rev.
2002
, vol. 
102
 (pg. 
387
-
429
)
31
Chervenak
M. C.
Toone
E. J.
Calorimetric analysis of the binding of lectins with overlapping carbohydrate-binding ligand specificities
Biochemistry
1995
, vol. 
34
 (pg. 
5685
-
5695
)
32
Xie
H.
Bolam
D. N.
Nagy
T.
Szabo
L.
Cooper
A.
Simpson
P. J.
Lakey
J. H.
Williamson
M. P.
Gilbert
H. J.
Role of hydrogen bonding in the interaction between a xylan binding module and xylan
Biochemistry
2001
, vol. 
40
 (pg. 
5700
-
5707
)
33
Ramakrishnan
B.
Boeggeman
E.
Ramasamy
V.
Qasba
P. K.
Structure and catalytic cycle of β-1,4-galactosyltransferase
Curr. Opin. Struct. Biol.
2004
, vol. 
14
 (pg. 
593
-
600
)
34
Hennet
T.
The galactosyltransferase family
Cell. Mol. Life Sci.
2002
, vol. 
59
 (pg. 
1081
-
1095
)
35
Amado
M.
Almeida
R.
Schwientek
T.
Clausen
H.
Identification and characterization of large galactosyltransferase gene families: galactosyltransferases for all functions
Biochim. Biophys. Acta
1999
, vol. 
1473
 (pg. 
35
-
53
)
36
Herman
T.
Horvitz
H. R.
Three proteins involved in Caenorhabditis elegans vulval invagination are similar to components of a glycosylation pathway
Proc. Natl. Acad. Sci. U.S.A.
1999
, vol. 
96
 (pg. 
974
-
979
)
37
Almeida
R.
Amado
M.
David
L.
Levery
S. B.
Holmes
E. H.
Merkx
G.
Van Kessel
A. G.
Ryggaard
E.
Hassan
H.
Bennett
E.
Clause
H.
A family of human β4-galactosyltransferases: cloning and expression of two novel UDP-galactose–β-Nacetylglucosamine β1, 4-galactosyltransferases, β4Gal-T2 and β4Gal-T3
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
31979
-
31992
)
38
Okayama
M.
Kimata
K.
Suzuki
S.
The influence of p-nitrophenyl β-D-xyloside on the synthesis of proteochondroitin sulfate by slices of embryonic chick cartilage
J. Biochem. (Tokyo)
1973
, vol. 
74
 (pg. 
1069
-
1073
)
39
Schwartz
N. B.
Galligani
L.
Ho
P. L.
Dorfman
A.
Stimulation of synthesis of free chondroitin sulfate chains by β-D-xylosides in cultured cells
Proc. Natl. Acad. Sci. U.S.A.
1974
, vol. 
71
 (pg. 
4047
-
4051
)
40
Lugemwa
F. N.
Sarkar
A. K.
Esko
J. D.
Unusual β-D-xylosides that prime glycosaminoglycans in animal cells
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
19159
-
19165
)