Glycobiology, 2000, Vol. 10, No. 11 1201-1208
© 2000 Oxford University Press
Molecular modeling and mutagenesis studies of the N-terminal domains of galectin-3: evidence for participation with the C-terminal carbohydrate recognition domain in oligosaccharide binding
National Institute for Medical Research, Mill Hill, London NW7 1AA, UK
Received on March 20, 2000; revised on May 10, 2000; accepted on May 20, 2000.
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsReferences |
|---|
A model structure (Henrick,K., Bawumia,S., Barboni,E.A.M., Mehul,B. and Hughes,R.C. (1998) Glycobiology, 8, 45–57) of the carbohydrate recognition domain (CRD, amino acid residues 114–245) of hamster galectin-3 has been extended to include N-terminal domain amino acid residues 91–113 containing one of the nine proline-rich motifs present in full-length hamster galectin-3. The modeling predicts two configurations of the N-terminal tail: in one the tail turns toward the first (SI) and last (S12) ß-strands of the CRD and lies at the apolar dimer interface observed for galectins -1 and -2. In the second folding arrangement the N-terminal tail lies across the carbohydrate-binding pocket of the CRD where it could participate in sugar-binding: in particular tyrosine 102 and adjacent residues may interact with the partly solvent exposed nonreducing N-acetylgalactosamine and fucose substituents of the A-blood group structure GalNAc
1,3 [Fuc1,2]Galß1,4GlcNAc-R. Binding studies using surface plasmon resonance of a recombinant fragment 
1–93 protein containing residues 94–245 of hamster galectin-3 and a collagenase-derived fragment 1–103 containing residues 104–245, as well as alanine mutagenesis of residues 101–105 in 1–93 protein, support the prediction that Tyr102 and adjacent residues make significant contributions to oligosaccharide binding. Key words: galectin-3/N-terminal domains/carbohydrate binding
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Galectin-3, a member of a family of carbohydrate-binding proteins (Liu, 1993
; Barondes et al., 1994; Kasai and Hirabayashi, 1996; Hughes, 1997; Leffler, 1997; Perillo et al., 1998), is expressed in the cytoplasm and sometimes the nucleus (Wang et al., 1991, 1992, 1995) of immune and epithelial cells. Intracellularly, the lectin appears to regulate important pathways including splicing reactions and apoptosis (Dagher et al., 1995; Yang et al., 1996; Akahani et al., 1997). The lectin is also secreted from cells (Hughes, 1999) and can bind to cell surface glycoproteins such as integrin subunits (Dong and Hughes, 1997; Ochieng et al., 1998) or to extracellular matrix glycoproteins such as laminin (Woo et al., 1990; Ochieng et al., 1998; Sato and Hughes, 1992). Proposed extracellular roles for galectin-3 include modulation of cellular adhesion and signaling, activation of immune reactions and regulation of tumorigenesis (Liu, 1993; Hughes, 1997; Perillo et al., 1998; Rabinovich, 1999).
Galectin-3 is a multidomain molecule (Figure 1) containing a C-terminal carbohydrate-recognition domain (CRD) with close structural homology to the CRDs of other galectins (Lobsanov and Rini, 1997; Seetharaman et al., 1998; Henrick et al., 1998; Rini and Lobsanov, 1999). Galectin-3 uniquely contains an N-terminal domain of various lengths according to species, which is encoded within a single exon of the genomic sequence (Gritzmacher et al., 1992) and consists of repeats of proline-tyrosine-glycine rich motifs (Herrmann et al., 1993). There are nine such repeats in hamster galectin-3 (Figure 1). Although galectins in general bind to ß-galactosides, such as lactosamine (Galß1,4GlcNAc), galectin-3 shows additional specificity in binding to oligosaccharides bearing 2- or 3-O--substituents on the terminal galactose residue, such as NeuNAc2,3 lactosamine or the A–blood group structure GalNAc1,3 [Fuc1,2]Galß1,4GlcNAc (Leffler and Barondes, 1986; Sparrow et al., 1987; Sato and Hughes, 1992; Ahmed and Vasta, 1994; Feizi et al., 1994). Previous work (Henrick et al., 1998; Seetharaman et al., 1998) implicated several amino acid residues of the CRD of galectin-3, absent in other galectins, in binding to extended oligosaccharides. Mutations at one or more of these residues of hamster galectin-3 abolished the discrimination shown by the wild-type lectin for extended oligosaccharides compared with simple ß-galactosides (Henrick et al., 1998). Although many studies show the dominant importance of the C-terminal domain of galectin-3 in ligand binding, other results indicate that N-terminal domains may also play a role. Thus, full-length galectin-3 binds to multiglycosylated proteins with positive cooperativity whereas the CRD fragment lacking N-terminal domains does not (Hsu et al., 1992; Massa et al., 1993; Probstmeier et al., 1995). Positive binding cooperativity implies that at increasing concentrations the full-length galectin-3 associates into multimeric complexes, presumably involving interactions between the N-terminal tails. Secondly, monoclonal antibodies recognizing N-terminal epitopes were found to modulate binding of galectin-3 to glycoproteins, negatively or positively according to the N-terminal epitope recognized by the antibody (Liu et al., 1996; Barboni et al., 1999). However, up to now there has been no evidence for any direct role of non-CRD domains in galectin-3 binding to carbohydrates.
|
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Modeling of the N-terminal domains of galectin-3
No structure is available as yet for intact galectin-3. However, structural analysis of human galectin-3 CRD (Seetharaman et al., 1998
) used a fragment, derived by collagenase treatment of intact lectin, that contained additional residues N-terminal to the CRD proper (see Figure 1). The latter was shown to consist of 12 ß-strands with close packing similarities to the CRDs of other galectins (Figure 2A,B). Of the extra residues, electron density could be observed for the last five (L IVPY, LTVPY in hamster, see Figure 1) which indicated that the LIV triplet, together with the ring hydroxyl side-chains of residues Tyr118 and Tyr247, may obstruct the apolar interface of human galectin-3 CRD formed by hydrophobic amino acid residues in the first (S1) and last (S12) ß-strand. This surface appears (Lobsanov et al., 1993; Liao et al., 1994) to be the driving force for the formation of galectin-1 and -2 homodimers (Figure 2A). The crystal structure of human galectin-3 CRD also indicated that the Val residue in the conserved triplet VPY (see Figure 1) makes a cis–peptide linkage with the following Pro residue, which could account for the projection of Leu114 (Leu109 in hamster; Figure 1) and preceding residues into the S1/S12 interface. In our present study, we have extended a previously derived model structure of the CRD (residues 114–245) of hamster galectin-3 (Henrick et al., 1998) to include all amino acid residues from Ser 91 to Ile 245. In the modeling, N-terminal domain residues 91 to 113 (see Figure 1) were allowed to adopt a conformation lying at the S1/S12 interface of the CRD (Figure 2B). We found that the turn at the VPY triplet was such that the hydroxyl side-chain of Tyr113, equivalent to Tyr118 in human galectin-3, could reduce apolarity surrounding Val 111, and indicated several additional influences of N-terminal residues on the hydrophobic character of the S1/S12 interface (Figure 2D,E). In particular, Pro 105 and Tyr 102 point to Leu 115 and to a lesser extent Leu 117 on the S1 ß-strand and Leu 237 on the S 12 ß-strand.
|
We also considered the consequences if galectin-3 CRD subunits could dimerize, e.g., by adopting the same folding arrangement as that observed for galectin-1 (Liao et al., 1994
; see Figure 2A) and galectin-2 (Lobsanov et al., 1993). Clearly, were galectin-3 dimers to form in such a way then the N-terminal region has to adopt a second conformation. A trial model for this second mode of N-terminal folding generated a structure in which residues 91 to 111 fold away from the CRD, held in a dimer. With this different orientation of the putative VPY hinge triplet (see Figure 2C), one alternative is to loop the N-terminal tail away from the S1/S12-interface such that it lies across the carbohydrate-binding pocket. Further modeling of galectin 3 residues 91 to 245 in this configuration, with bound blood-group A-tetrasaccharide GalNAc1,3 [Fuc1,2] Galß1,4GlcNAc or NeuNAc2,3Galß1,4Glc(NAc), suggested a contribution to sugar binding by N-terminal residues (Figure 2F). Thus, Tyr 102–OH is predicted to form H-bonds to Fuc-O2, -O3, or, in the case of sialylated sugars, to NeuNAc–O1A. Hamster Tyr 102 is conserved in human and mouse but is Phe in rat and dog. Similarly, Thr 106–OG is predicted to H-bond to Tyr 113–OH which is conserved in all species. This positioning of Tyr 113 contributes to the positioning of the side-chain of the conserved CRD amino acid residue Glu 230 previously implicated in binding to NeuNAc–O10, and in interactions with the side chain of Ser232, one of two critical CRD residues together with Arg 139, in binding of extended oligosaccharides (Henrick et al., 1998).
Binding of 1–93 protein to immobilized laminin
In order to test the prediction that N-terminal tail residues may contribute to carbohydrate-binding, we examined first the binding of recombinant hamster galectin-3 fragment 1–93 (Figure 1) to immobilized laminin using surface plasmon resonance (SPR) measurements. Binding occurred in two distinct phases (Figure 3A); a relatively rapid increase in the measured response with apparent kass 30,000 M–1 s–1, followed by a slower increase with apparent kass 4400 M–1 s–1 (Table I). Dissociation also was biphasic: a fast rate with kdiss 0.3 s–1 and a much slower dissociation rate with kdiss 0.002 s–1. These values are similar to earlier determinations (Barboni et al., 1999).When the protein sample (30 µl) was injected over the sensor surface at faster flow-rates than the standard 10 µl/min, the amount of the slowly dissociating binding was greatly reduced and this fraction was increased at slower flow-rates (results not shown). These data suggested that the slowly associating 1–93 molecules became incorporated into a slowly dissociating lectin population bound to the substratum with a Ka value of about 2 x 106 M–1 (Table I). As described previously (Barboni et al., 1999), the collagenase–derived CRD fragment 1–103 (Figure 1), which lacks any proline-rich repeat, showed only fast association and dissociation rates giving a weak affinity constant Ka 0.5 x 105 M–1 (Table I), a value similar to that (Ka 1 x 105 M–1) derived for the fast association-fast dissociation binding of 1–93 protein (Table I).
|
|
The monophasic binding kinetics of
1–103, together with the similar behavior of two plant lectins (Barboni et al., 1999), on binding to immobilized laminin suggests a mass transfer influence (Karlsson and Falt, 1997; Myszka, 1997; Sinohara et al., 1997) on the more complex binding kinetics of intact galectin-3 (Barboni et al., 1999) or 1–93 under identical conditions is unlikely. The absence of a significant SPR response of intact galectin-3 when passed over immobilized BSA or deglycosylated laminin (Barboni et al., 1999), and the independence with regard to flow rate of the initial association rates of intact galectin-3 (Barboni et al., 1999) or 1–93 with immobilized laminin, support this conclusion (Karlsson and Falt, 1997). Furthermore, the shape of the response curve was essentially the same when intact galectin-3 (Barboni et al., 1999) or 1–93 (data not shown) was passed over separate sensor surfaces (BIAcore chips CM5, B1, and F1) with different matrix characteristics and derivatized with laminin covering a 10-fold range of densities. These control results argue against any significant effect of ligand density or sensor surface properties on analyte diffusion coefficients and mass transport effects influencing binding kinetics.
Effects of N-terminal mutations of 1–93 protein on binding to immobilized laminin
The binding to immobilized laminin of 1–93 mutants M1-3 (Figure 1) obtained by mutagenesis of N-terminal tail residues lying between Pro101 and Pro105 is shown in Figure 3A. The initial association and dissociation rates in each case were similar to those found for 1–93 protein (Table I). However, the second association event proceeded progressively more slowly for M1 and M2 and was undetectable for M3 over the standard 180 s injection time (Figure 3A, Table I). Correspondingly, no slowly-dissociating binding was detected with M3. A slowly dissociating fraction of M1 and M2 was found (Figure 3A), approximately equal (M1) or even greater (M2) in amount in comparison with 1–93 protein, but of weaker avidity with Ka < 0.2 x 106 M–1 (Table I). We also tested the effect of mutation at Pro112, lying within the putative hinge sequence VPY. The binding to immobilized laminin of M4 (Figure 1) proceeded with kinetics most similar to M1, and with similar weak avidity (Figure 3A, Table I).
Hapten inhibition of lectin binding to immobilized laminin
We examined next the inhibition of binding of 1–93 proteins to immobilized laminin by the A-blood group tetrasaccharide GalNAc 1,3 [Fuc 1,2] Galß1,4 GlcNAc. Lactose was used as reference sugar. By using SPR measurements we could examine the effect of the haptens on all kinetic components of binding. In previous solid state binding assays, the A-tetrasaccharide was shown to be 12–25 times more effective compared with lactose in inhibiting the binding of full-length galectin-3 of human, rat, or hamster origin to different glycoprotein substrates (Leffler and Barondes, 1986; Sparrow et al., 1987; Sato and Hughes, 1992; Ahmed and Vasta, 1994).
The fast association–fast dissociation binding (phase I, Figure 3B) of 1–93 to the sensor–surface was reduced to 50% by 25 µM A-tetrasaccharide and by 300 µM lactose (Figure 4, Table II), a relative value of 
12. The slow association–slow dissociation binding (phase II, Figure 3B) was more resistant to both A–tetrasaccharide and lactose: ID50 82 µM and 760 µM, respectively (Figure 4, Table II). Nevertheless, the A-tetrasaccharide remained almost 10 times more inhibitory than lactose. Single mutations of tyrosine residue 102 or proline 105 in M1 and M2, respectively, had two effects on binding to immobilized laminin. First, the inhibitory effect of A-tetrasaccharide (ID50) relative to lactose was decreased 3- to 4-fold (Figure 4, Table II). Secondly, the slow association–slow dissociation binding became more sensitive to lactose (Figure 4, Table II). Mutations of residues 101–105 in M3 also decreased the effectiveness of A-tetrasaccharide relative to lactose in inhibiting the monophasic binding to laminin (Figure 4, Table II). By contrast, the M4 protein containing a mutation of proline residue 112 retained about 8-fold sensitivity to A-tetrasaccharide relative to lactose, similar to that shown by 1–93 protein (Figure 4, Table II). However, unlike 1–93, both binding phases were equally sensitive. We also determined the relative inhibitory effects of A-tetrasaccharide and lactose on the binding to laminin of the collagenase-derived 1–103 protein, obtaining a value of 2.6 similar to M1-3 proteins (Figure 4, Table II).
|
|
Affinity of laminin binding to immobilized lectins
The affinity of galectin-3 interactions with laminin was also determined in the reverse orientation with either
1–93 protein, M3 protein or 1–103 protein immobilized on the sensor surface and laminin in solution. Binding of laminin to each surface was readily detectable, whereas no binding was detected using laminin previously digested with endo-ß-galactosidase (Figure 5). Thus, the polylactosamine glycans of laminin (Arumugham et al., 1986; Fujiwara et al., 1988; Knibbs et al., 1989) are required for high-affinity binding of truncated galectin-3 mutants, as they are for the intact lectin (Barboni et al., 1999). However, unlike binding of mutant proteins to immobilized laminin, laminin binding to immobilized lectins gave rate values (kass 130–200,000 M–1 s–1 and kdiss 0.02 s–1) that varied little between the different surfaces, as did the calculated Ka values of 5–10 x 106 M–1 (Table III). These data indicate that when laminin binds from solution to surfaces carrying bound lectin, the N-terminal tails have no detectable influence on interactions mediated by the CRDs.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
At present there is no direct evidence that either of the structures (Figure 2D,F) suggested by molecular modeling of N-terminal tail residues of hamster galectin-3 exist. However, these predictions invite useful speculation, some of which has been approached experimentally here. Our results support the prediction that the N-terminal domains of galectin-3 contribute significantly to carbohydrate-binding. Previous studies suggested that CRD residues Arg139 and Ser232 play a major part in binding to sugars substituted in
-linkage to the 2- or 3-hydroxyl at the nonreducing end of lactose or lactosamine (Henrick et al., 1998). Mutations of either or both residues effectively reduced or abolished the preference of galectin-3 for the extended oligosaccharides compared with unsubstituted disaccharides. As judged by hapten inhibition assays (Figure 4, Table II), the effects of mutation of N-terminal residues 101–105 of 1–93 protein are quantitatively less but nevertheless significant. Furthermore, the collagenase-derived 1–103 protein lacking these residues behaved in hapten-inhibition experiments similarly to the mutant proteins (Figure 4, Table II). We considered the possibility that loss by collagenase treatment or mutagenesis of these residues, which precede the CRD (Figure 1), induces nonspecific changes in CRD structure leading to a changed carbohydrate-binding specificity. However, this is unlikely since M4 protein carrying a mutation at the even more proximal residue Pro112 (see Figure 1) retained the binding preference of 1–93 protein (Figure 4, Table II).
The putative conformational transitions of the N-terminal tail (Figure 2D,F), and the demonstrated contribution of N-terminal residues to sugar-binding, may be relevant in interpreting the interactions of 1–93 protein with immobilized laminin (Figure 3A). We propose as a working hypothesis that the following series of events may occur. In one type of interaction (Figure 6a), 1–93 in conformation A binds to laminin carbohydrate in a manner not involving N-terminal tail residues. An additional lectin molecule may complex with laminin-bound protein, possibly through interaction between N-terminal tails (Barboni et al., 1999) as proposed for the full-length lectin (Massa et al., 1993); however, the interaction is still monovalent (Figure 6b). In a second mode of interaction (Figure 6a), 1–93 in conformation B is bound to laminin carbohydrate in a manner assisted by N-terminal residues. In some way, yet to be determined but possibly by exposure of hydrophobic residues, substrate-bound subunits may complex by CRD–CRD association, leading to multivalent binding (Figure 6c). Additional monomers may be incorporated into the substrate-bound complex through N-terminal tail interactions (Figure 6d). Perhaps an A to B transition is promoted also in these subunits by their exposure to a hydrophobic interaction site presented by adjacent substrate-bound 1–93 molecules, leading to further CRD-CRD associations. Such multivalent interactions with substrate would be more stable than monovalent binding, perhaps accounting for the increased resistance to sugar haptens observed for the slowly associating–slowly dissociating fraction of 1–93 protein bound to a laminin substrate (Figure 4, Table II). The Ka value (about 2 x 106 M–1; Table I) for this fraction is similar to values obtained for binding of soluble laminin to lectin-coated surfaces (5–10 x 106 M–1; Table III). Laminin, a highly glycosylated molecule, would be expected to bind from solution to the lectin-coated surfaces multivalently.
|
Our results indicate that mutation of key residues in the N-terminal tail destabilizes substrate-bound lectin complexes, producing interactions of lower avidity. Perhaps, when bound to a substratum, conformation B (Figure 6a) is stabilized by N-terminal tail interactions with a high-affinity carbohydrate ligand such as laminin and mutation of residues 101–105 in M1-3 proteins may shift the equilibrium towards conformation A, with reduced capacity for CRD–CRD interactions and multivalent binding. Notably, many of the large number of polylactosamine chains in laminin are capped by an
-1,3-galactosyl residue (Arumugham et al., 1986; Fujiwara et al., 1988; Knibbs et al., 1989). Mutation within the putative hinge residues 111–113 as in M4 may affect the stability of a B state, with similar consequences.
Previous studies showed that intact galectin-3 binds to laminin and other glycoproteins with positive cooperativity (Hsu et al., 1992; Massa et al., 1993; Probstmeier et al., 1995) indicating a concentration-dependent self-assembly of lectin subunits. One study showed that galectin-3 binding to laminin at high or low concentrations was equally sensitive to lactose inhibition (Massa et al., 1993), suggesting that binding valency did not increase at high concentrations and multimerization occurred mainly through N-terminal tail interactions. However, these experiments were relatively short-term, lasting only 75 min, and additional CRD–CRD interactions akin to those proposed for 1–93 protein (Figure 6c) might take place slowly during binding of intact galectin-3 to immobilized laminin. Perhaps, conformational flexibility of the N-terminal tail is increased in 1–93 protein, promoting interactions between CRD subunits and leading more rapidly to stable substrate-bound aggregates. Interestingly, a 22 kDa fragment of human galectin-3 obtained by limited proteolysis of the N-terminal tail with metalloproteinases, bound more tightly to laminin than the undigested lectin (Ochieng et al., 1998).
The present data add to the evidence accumulating for roles of the N-terminal domains in galectin-3 interactions and functions. These domains are sites for phosphorylation, which appears to be involved in nuclear retention of the protein in proliferating fibroblasts (Cowles et al., 1990; Hamann et al., 1991). They also carry determinants necessary for the secretion of galectin-3 from transfected cells through nonclassical secretory pathways independent of endoplasmic reticulum–Golgi compartments (Mehul and Hughes, 1997; Gong et al., 1999; Hughes, 1999; Menon and Hughes, 1999). Extracellularly, the N-terminal domains are substrates for cross-linking by tissue-type transglutaminases (Mehul et al., 1995), generating covalent oligomers of galectin-3 that stimulate spreading of human melanoma cells on laminin substrata (van den Brule et al., 1998). Since in vivo, galectin-3 is likely to be operating at limiting concentrations compared with a large repertoire of glycoconjugates bearing glycans of different affinities, mechanisms for the assembly of lectin conformers capable of stable and multivalent binding to relevant high-affinity cell–surface and matrix ligands may be vital for the functioning of galectin-3 in biological processes.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Restriction enzymes were from Boehringer Mannheim. Murine EHS tumor laminin, the A-tetrasaccharide GalNAc
1,3 [Fuc1,2] Galß1,4 Glc, lactose–agarose and other reagents were obtained from Sigma (Poole, Dorset). Partial deglycosylation of laminin with endo-ß-galactosidase from Bacteriodes fragilis was carried out as described previously (Barboni et al., 1999). Recombinant fragments of hamster galectin-3, as well as CRD fragment 1–103 (Figure 1) obtained after digestion of full-length lectin with bacterial collagenase, were purified as described previously (Mehul et al., 1994).
Expression plasmids
Hamster galectin-3 was truncated at position 94 to produce the 1-93 fragment (Figure 1), using the full-length galectin-3 cDNA as template and primers SB2 5'–TAT-GGA-TCC-TTA-GAT-CAT-GGT-GGG-TG-3' and SB3 5'-TAT-CCA-TGG-GAG-CCT-ATC-CTG-CTG-C-3', containing, respectively, a BamHI site coinciding with a start codon ATG and an NCo1 site (Barboni et al., 1999). The PCR was performed with the ExpandTM High Fidelity PCR system (Boehringer Mannheim) as follows: one cycle at 94°C for 2 min, 10 cycles each at 94°C for 20 s, at 68°C for 20 s and at 72°C for 45 s; 20 cycles each at 94°C for 20 s, at 68°C for 20 s and at 72°C for 2 min with extension time increasing 10 s/cycle, followed by 1 cycle at 72°C for 7 min. The PCR product was digested with BamHI and NCo1 and ligated into a PTMN vector (Deng et al., 1990) digested with BamHI and Nco1. Mutation of 1-93 from tyrosine to alanine at position 102 (M1, Figure 1) was carried out by PCR as above using 1-93 cDNA as template with primers SB6 5'-TA-TCC-ATG-GCA-GCC-TAT-CCT-GCT-GCT-GGC-CCC-GCT-GG-3' and SB2. Further alanine mutations were carried out at positions 105 (M2, Figure 1) using 1-93 cDNA as template and primers SB5 5'-TA-TCC-ATG-GGA-GCC-TAT-CCT-GCT-GCT-GGC-GCC-TAT-GGC-GCT-ACC-GGA-GCA-TTG-ACA-GTG-3' and SB2, or positions 101–105 (M3, Figure 1) using 1-93 cDNA as template and primers SB4 5'-TA-TCC-ATG-GGA-GCC-TAT-CCT-GCT-GCT-GGC-GCC-GCT-GCC-GCC-GCT-ACC-GGA-GCA-TTG-ACA-GTG–3' and SB2. A mutation at position 112 (M4, Figure 1) from proline to alanine was derived with the ExpandTM Long PCR system (Boehringer Mannheim) using 1-93 cDNA as template and primers SB12 5'-GCA-TTG-ACA-GTG-GCC-TAT-AAG-C-3' and SB11 5'-TCC-GGT-AGG-GGC-GCC-ATA-G-3'. The PCR product was incubated with Dpn1 and pfu DNA polymerase to digest template DNA and blunt end the PCR product, which after gel purification was ligated into PTMN. Purification of plasmids was carried out using a Hybaid kit and nucleotide sequences verified by dideoxy-chain termination method using a Sequenase kit, version 2. 0 (U.S Biochemical Company). Expression of recombinant lectins in plasmid-transformed E.coli strain BL21 DE3 pLys S was induced for 4 h at 37°C by addition of 0.2 mM isopropylthio-ß-galactoside (IPTG). Cells were pelleted, washed with cold PBS, resuspended in lysis buffer (1 M Tris pH 7.4, 5mM EDTA, 10mM 2-mercaptoethanol, 2 µg/ml aprotinin, 100 µM leupeptin, 1 µM pepstatin A, 1mM PMSF, 0. 2%w/v NaN3) and sonicated at 120 W for four 20 s cycles at 4°C. Bacterial suspensions were centrifuged at 28,000 r.p.m. and 4°C for 45 min and the clarified lysates eluted from a lactose-Agarose affinity column with 150 mM lactose, 50 mM TRIS pH 7.2, 5 mM EDTA, 2 mM 2-mercaptoethanol, 0.02% (w/v)NaN3 and 0.1 mM PMSF. After gel filtration on Biogel P60 to remove lactose (Barboni et al., 1999), lectins were stored frozen at 1 mg/ml.
Surface plasmon resonance
Surface plasmon resonance experiments were performed at 25°C on a BIAcore biosensor (BIAcore AB, Stevenage, UK) using the multichannel command and with PBS-5 mM EDTA-0.02% sodium azide as running buffer (Barboni et al., 1999). Proteins, either laminin or galectin-3 fragments, at 100 µg/ml in each case were covalently coupled to CM5 BIAcore sensor chips using the BIAcore Amine Coupling kit and following the manufacturers instructions. Control chip surfaces were prepared with either BSA or deglycosylated laminin. Protein solutions (30 µl, up to 10 µM) were injected across a control or experimental chip surface at 10 µl/min. Regeneration of the sensor surface was carried out with a 30 s pulse of 100 mM lactose in the above buffer. Data transformation was prepared with BIAcore 2.1 evaluation software. Association rate (kass; M–1 s–1) was computed from the binding data at optimal concentration using non-linear fitting statistics. The dissociation rate (kdiss; s–1) was measured starting 5–10 s after the sample was replaced with buffer. The association constant (Ka) was then calculated (Ka = kass/kdiss). For oligosaccharide inhibition studies, samples (30 µl) of lectin solutions at a fixed concentration (1–5 µM in different experiments) were mixed with increasing concentrations of competing sugar and injected as described above. Inhibition of lectin binding to immobilized laminin was derived from the percentage of plateau response units (RU) reached by the sugar-incubated sample compared with the control sample and expressed as the concentration of sugar producing 50% inhibition (ID50).
Molecular modeling
The coordinates of hamster galectin-3 residues 111–243, derived previously (Henrick et al., 1998), were used as a starting point for modeling the extra N-terminal residues 94–110. Modeling and oligosaccharide docking procedures were carried out as described previously (Henrick et al., 1998).
|
Acknowledgements |
|---|
Dr. Barboni was supported by a PRIN 97 Programme of the Italian Ministry of University, Scientific and Technological Research.
|
Footnotes |
|---|
1 Permanent address: Departments of Cellular Biotechnology and Haematology, University of Rome "La Sapienza," Rome 00161, Italy
2 Present address: European Bioinformatics Institute, EMBL Outstation, Hinxton, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK
3 To whom correspondence should be addressed
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Ahmed,H. and Vasta,G.R. (1994) Galectins: conservation of functionally and structurally relevant amino acid residues defines two types of carbohydrate recognition domains. Glycobiology, 4, 545–549.
Akahani,S., Nangia-Makker,P., Inohara,H., Kim,H.R. and Raz,A. (1997) Galectin-3: a novel antiapoptotic molecule with a functional BH1 (NWGR) domain of Bcl-2 family. Cancer Res., 57, 5272–5276.
Arumugham,R.G., Hsieh,T.C.Y., Tanzer,M.L. and Laine,R.A. (1986) Structure of the asparagine–linked sugars of laminin. Biochim. Biophys. Acta, 883, 112–126.[Medline]
Barboni,E.A.M., Bawumia,S. and Hughes,R.C. (1999) Kinetic measurements of binding of galectin-3 to a laminin substratum. Glycoconjugate J., 16, 365–372.[Web of Science][Medline]
Barondes,S.H., Cooper,D.N.W., Gitt,M.A. and Leffler,H. (1994) Galectins. Structure and function of a large family of animal lectins. J. Biol. Chem., 269, 20807–20810.
Cowles,E.A., Agrwal,N., Anderson,R.L. and Wang,J.L, (1990) Carbohydrate binding protein 35. Isoelectric points of the polypeptide and a phosphorylated derivative. J. Biol. Chem., 265, 17706–17712.
Dagher,S.F., Wang,J.L. and Patterson,R.J. (1995) Identification of galectin-3 as a factor in pre-mRNA splicing. Proc. Natl. Acad. Sci. USA, 92, 1213–1217.
Deng,T., Noel,J.P. and Tsai,M.D. (1990) A novel expression vector for high–level synthesis and secretion of foreign proteins in Escherichia coli: overproduction of bovine pancreatic phospholipase A2. Gene (Amsterdam), 93, 229–234.[Web of Science][Medline]
Dong,S. and Hughes,R.C. (1997) Macrophage surface glycoproteins binding to galectin 3 (Mac2 antigen). Glycocon. J., 14, 267–274.[Web of Science][Medline]
Feizi,T., Solomon,J.C., Yuen,C.T., Jeng,K.C.G., Frigeri,L.G., Hsu,D.K. and Liu,F.T. (1994) The adhesive specificity of the soluble human lectin, IgE-binding protein, toward lipid-linked oligosaccharides–presence of the blood group-A, group-B, group-B-like and group-H monosaccharides confers a binding-activity to tetrasaccharide (lacto-N-tetraose and lacto-N-neotetraose) backbones. Biochemistry, 33, 6342–6349.[Medline]
Fujiwara,S., Shinkai,H., Deutzmann,R., Paulsson,M.and Timpl,R. (1988) Structure and distribution of N–linked oligosaccharide chains on various domains of mouse tumor laminin Biochem. J., 252, 453–461.[Web of Science][Medline]
Gong,H.C., Honjo,Y., Nangia-Makker,P., Hogan,V., Mazurak,N., Bresalier,R.S. and Raz,A. (1999) The NH2 terminus of galectin-3 governs cellular compartmentalization and functions in cancer cells. Cancer Res., 59, 6239–6245.
Gritzmacher,C.A., Mehl,V.S. and Liu,F.T. (1992) Genomic cloning of the gene for an IgE-binding lectin reveals unusual utilization of 5'-untranslated regions. Biochemistry, 31, 9533–9538.[Medline]
Hamann,K.K., Cowles,E.A., Wang,J.L. and Anderson,R.L. (1991) Expression of carbohydrate binding protein 35 in human fibroblasts: variations in the levels of mRNA, protein and isoelectric species as a function of replicative competence. Exp. Cell Res., 196, 82–91.[Web of Science][Medline]
Henrick,K., Bawumia,S., Barboni,E.A.M., Mehul,B. and Hughes,RC (1998). Evidence for subsites in the galectins involved in sugar binding at the non-reducing end of the central galactose of oligosaccharide ligands: sequence analysis, homology modeling and mutagenesis studies of hamster galectin-3. Glycobiology, 8, 45–57.
Herrmann,J., Turck,C.W., Atchison,R.E., Huflejt,M.E., Poulter,L., Gitt,M.A., Burlingame,A.L., Barondes,S.H. and Leffler,H. (1993) Primary structure of the soluble lactose binding lectin L-29 from rat and dog and interactions of its non-collagenous proline, glycine, tyrosine-rich sequence with bacterial and tissue collagenase._J. Biol. Chem., 268, 26704–26711.
Hsu,D.K., Zuberi,R.I. and Liu,F.T. (1992) Biochemical and biophysical characterization of human recombinant IgE–binding protein, an S-type animal lectin. J. Biol. Chem., 267, 14167–14174.
Hughes,R.C. (1997)The galectin family of mammalian carbohydrate-binding molecules. Biochem. Trans., 25, 1194–1198.
Hughes,R.C. (1999) Secretion of the galectin family of mammalian carbohydrate-binding proteins. Biochim. Biophys. Acta, 1473, 172–185.[Medline]
Karlsson,R. and Falt,A. (1997) Experimental design for kinetic analysis of protein–protein interactions with surface plasmon resonance biosensors. J. Immunol. Methods, 200, 121–133.[Web of Science][Medline]
Kasai,K.-I. and Hirabayashi,J. (1996) Galectins: a family of animal lectins that decipher glycocodes. J. Biochem., 119, 1–8.
Knibbs,R.N., Perini,F. and Goldstein,I.J. (1989) Structure of the major concanavalin A reactive oligosaccharides of the extracellular matrix component laminin. Biochemistry, 28, 6379–6392.[Medline]
Kuklinski,S. and Probstmeir,R. (1998) Homophilic binding properties of galectin-3: involvement of the carbohydrate recognition domain J. Neurochem., 70, 814–823.[Web of Science][Medline]
Leffler,H. (1997) Introduction to galectins. Trends Glycosci. Glycotech., 9, 9–19.
Leffler,H. and Barondes,S.H. (1986) Specificity of binding of three soluble rat lung lectins to substituted and unsubstituted mammalian ß-galactosides. J. Biol. Chem., 261, 10119–10126.
Liao,D.-I., Kapadia,G., Ahmed,H., Vasta,G.R. and Herzberg,O. (1994) Structure of S-lectin, a developmentally regulated vertebrate ß-galactoside binding protein. Proc. Natl. Acad. Sci. USA, 91, 1428–1432.
Liu,F.T. (1993) S-Type mammalian lectins in allergic inflammation. Immunol. Today, 14, 486–490.[Web of Science][Medline]
Liu,F.-T., Hsu,D.K., Zuberi,R.I., Hill,P.N., Shenhav,A., Kuwabara,I. and Chen,S.S. (1996) Modulation of funtional properties of galectin-3 by monoclonal antibodies binding to the non-lectin domains Biochemistry, 35, 6073–6079.[Medline]
Lobsanov,Y., Gitt,M.A., Leffler,H., Barondes,S.A. and Rini,J.M. (1993) X-Ray crystal structure of the human dimeric S-lac lectin, L-14-II, in complex with lactose at 2.9Å resolution. J. Biol. Chem., 268, 27034–27038.
Lobsanov,Y.D and Rini,J.M. (1997) Galectin structure. Trends Glycosci. Glycotech., 9, 145–154.
Massa,S.M., Cooper,D.N.W., Leffler,H. and Barondes,S.H. (1993) L-29 an endogenous lectin, binds to glycoconjugate ligands with positive cooperativity. Biochemistry, 32, 260–267.[Medline]
Mehul,B., Bawumia,S., Martin,S.R. and Hughes,R.C. (1994) Structure of baby hamster kidney carbohydrate binding protein CBP30, an S-type animal lectin. J. Biol. Chem., 269, 18250–18258.
Mehul,B., Bawumia,S. and Hughes,R. (1995) Cross-linking of galectin-3, a galactose–binding protein of mammalian tissues, by tissue-type transglutaminase. FEBS Lett., 360, 160–164.
Mehul,B. and Hughes,R.C. (1997) Plasma membrane targetting, vesicular budding and release of galectin-3 from the cytoplasm of mammalian cells during secretion. J. Cell Sci., 110, 1169–1178.[Abstract]
Menon,R.P and Hughes,R.C. (1999) Determinants in the N–terminal domains of galectin-3 for secretion by a novel pathway circumventing the endoplasmic reticulum-Golgi complex. Eur. J. Biochem., 264, 569–576.[Web of Science][Medline]
Myszka,D.G. (1997) Kinetic analysis of macromolecular interactions using surface plasmon resonance. Current Opin. Biotechnol., 8, 50–57.[Web of Science][Medline]
Ochieng,J. and Warfield,P. (1995) Galectin 3 binding potential of mouse tumor EHS and placental laminins. Biochem. Biophys. Res. Commun., 217, 402–406.[Web of Science][Medline]
Ochieng,J., Green,B., Evans,S.,James,O. and Warfield,P. (1998) Modulation of the biological functions of galectin-3 by matrix metalloproteinases. Biochim. Biophys. Acta, 1379, 97–106.[Medline]
Perillo,N.L., Marcus,M.E. and Baum,L.G. (1998) Galectins: versatile modulators of cell adhesion, cell proliferation and cell death. J. Mol. Med., 76, 402–412.
Probstmeier,R., Montag,D. and Schachner,M. (1995) Galectin-3, a ß-galactoside–binding animal lectin, binds to neural recognition molecules. J. Neurochem., 64, 2465–2474.[Web of Science][Medline]
Rabinovich,G.A. (1999) Galectins: an evolutionarily conserved family of animal lectins with multifunctional properties: a trip from the gene to clinical therapy. Cell Death Different., 6, 711–721.[Web of Science][Medline]
Rini,J.M. and Lobsanov,Y.D. (1999) New animal lectin structures. Curr. Opinion Struct. Biol., 9, 578–584.
Sato,S. and Hughes,R.C. (1992) Binding specificity of a baby hamster kidney lectin for H Type I and II chains, polylactosamine glycans and appropriately glycosylated forms of laminin and fibronectin. J. Biol. Chem., 267, 6983–6990.
Seetharaman,J., Kanigsberg,A., Slaaby,R., Leffler,H., Barondes,S.H. and Rini,J.M. (1998) X-Ray crystal structure of the human galectin-3 carbohydrate recognition domain at 2.1A resolution. J. Biol. Chem., 272, 13047–13052.
Sinohara,Y., Hasegawa,Y., Kaku,H. and Shibuya,N. (1997) Elucidation of the mechanism enhancing the avidity of lectin with oligosaccharides on the solid phase surface. Glycobiology, 7, 1201–1208.
Sparrow,C.P., Leffler,H. and Barondes,S.H. (1987) Multiple soluble ß-galactoside binding lectins from human lung. J. Biol. Chem., 262, 7383–7390.
van den Brule,F.A., Liu,F.T. and Castronovo, V (1998) Transglutaminase-mediated oligomerization of galectin-3 midulates human melanoma cell interactions with laminin. Cell Adhesion Commun., 5, 425–435.[Medline]
Wang,J.L., Laing,J.G. and Anderson,R.L. (1991) Lectins in the cell nucleus. Glycobiology, 1, 243–252.
Wang,J.L., Werner,E.A., Laing,J.G. and Patterson,R.J. (1992)Nuclear and cytoplasmic localization of a lectin-ribonucleoprotein complex. Biochem. Trans., 20, 269–274.
Wang,L., Inohara,H., Pienta,K.J. and Raz,A. (1995) Galectin-3 is a nuclear matrix protein which binds RNA. Biochem. Biophys. Res. Commun., 217, 292–303.[Web of Science][Medline]
Woo,H.J., Shaw,L.M., Messier,J.M. and Mercurio,A.M. (1990) The major non-integrin laminin binding protein of macrophages is identical to carbohydrate-binding protein 35 (Mac2). J. Biol. Chem., 265, 7097–7099.
Yang,R.Y., Hsu,D.K., Liu,F.T. (1996) Expression of galectin-3 modulates T-cell growth and apoptosis. Proc. Natl. Acad. Sci. USA, 93, 6737–6742.
Yang,R.Y., Hill,P.N., Hsu,D.K. and Liu,F.T. (1998) Role of the carboxyl–terminal lectin domain in self-association of galectin-3. Biochemistry, 37, 4086–4092.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Kang, K. Moon, E. Lee, Y. Lee, E. Lee, C Ahn, and Y. Song Renal expression of galectin-3 in systemic lupus erythematosus patients with nephritis Lupus, January 1, 2009; 18(1): 22 - 28. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Demers, K. Biron-Pain, J. Hebert, A. Lamarre, T. Magnaldo, and Y. St-Pierre Galectin-7 in Lymphoma: Elevated Expression in Human Lymphoid Malignancies and Decreased Lymphoma Dissemination by Antisense Strategies in Experimental Model Cancer Res., March 15, 2007; 67(6): 2824 - 2829. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Jin, A. Greenwald, M. D. Peterson, and T. K. Waddell Human Monocytes Recognize Porcine Endothelium via the Interaction of Galectin 3 and {alpha}-GAL J. Immunol., July 15, 2006; 177(2): 1289 - 1295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Lagana, J. G. Goetz, P. Cheung, A. Raz, J. W. Dennis, and I. R. Nabi Galectin Binding to Mgat5-Modified N-Glycans Regulates Fibronectin Matrix Remodeling in Tumor Cells Mol. Cell. Biol., April 15, 2006; 26(8): 3181 - 3193. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||