Our previous crystallographic studies on b4Gal-T1 and of the mutant Met340His-b4Gal-T1 in complex with chitobiose and various trisaccharides, together with the enzyme kinetic analysis and MD simulations defined the oligosaccharide binding site of b4Gal-T1 (see Project #s Z01 BC 009304). For a better understanding of the branch specificity of b4Gal-T1 towards the GlcNAc residues of N-glycans, the kinetic and crystallographic studies with the wild-type human b4Gal-T1 (h-b4Gal-T1) and the mutant Met340His-b4Gal-T1 (h-M340H-b4Gal-T1), in complex with a GlcNAc containing pentasaccharide and several GlcNAc containing trisaccharides present in N-glycans, showed that b4Gal-T1 preferentially interacts with the 1,2-1,6-arm trisaccharide rather than with the 1,2-1,3-arm or 1,4-1,3-arm of a bi- or tri-antennary oligosaccharide chain of N-glycan (see 2006 project # Z01 BC010041) In the present studies we are determining the transfer to a biantennary oligosaccharide of a glycoprotein e. IgG. Transfer preferences of b4Gal-T1 to the 1-3 or 1-6 arm of a biantennary glycan We showed, using synthetic trisaccharides as acceptor model systems, that the acceptor 1,2-1,3-arm (GlcNAc-b1,2-Man-a1,3-Man) and 1,4-1,3-arm (GlcNAc-b1,4-Man-a1,3-Man) trisaccharides have a 10- to 60-fold higher Km for b4Gal-T1 than the 1,2-1,6-arm trisaccharide (GlcNAc-b1,2-Man-a1,6-Man), latter shows substrate inhibition at concentrations that are much lower than for other acceptor substrates. We have established conditions under which the galactosylation of one or both antennas of an N-glycan of a glycoprotein, e.g., IgG, occur. To establish which antenna, 1-3-arm or 1-6-arm, or both, is galactosylated we have, in collaboration with Dr. Timothy Weybright (from Dr. Timothy Veenstras group, of the Mass Spectrometry Center, BPP, SAIC-Frederick, Inc.), analyzed the oligosaccharide products by MS/MS analysis and established that at high concentrations of IgG only 1-3-arm is galactosylated. We have followed the transfer of galactose to de-sialated, de-galactosylated monoclonal antibodies, Asialo-agalacto-IgG, which carry a single N-linked oligosaccharide chain on each heavy chain of the IgG at the Fc-region. MS analysis of the glycan chains after PNGaseF treatment of the monoclonal antibodies, bearing glycans of complex bi-antennary chains N-linked to Asn 297, show various glycosylated patterns; G0 glycoform having two terminal GlcNAc residues, and G1 and G2 glycoforms which bear oligosaccharides with one or two terminal galactose, respectively. As monitored by MS analysis, the conditions for the complete de-galactosylation (100 %) to G0 glycoform was established. After 6 hr transfer of galactose by the wild type 4Gal-T1 to G0 glycoform (MW 1485) of IgG, the MS analyses of the product after PNGaseF treatment of IgG, show that galactose is mainly transferred to one arm (G1 glycoform, MW 1647). Further MS/MS analysis of the bi-antennary glycan chain shows that the wild type 4Gal-T1 transfers galactose at a faster rate to the GlcNAc attached to the Man 1-3 arm. In contrast, the mutant 1,4Gal-T1-Y289L at 6 hours incubation transfers GalNAc mainly to both arms (G2 glycoform MW 1891). MS/MS analysis of the G1 glycoform after PNGasesF treatment of the monoclonal antibodies show that the mutant Y289L- b4Gal-T1 transfers GalNAc at a faster rate to the GlcNAc attached to the Man 1-6 arm. Crystal structure of the h-M340H-Gal-T1 in complex with the disaccharide GlcNAc-b3Gal-b-O-napthalenemethanol A class of synthetic disaccharides has been shown as potential inhibitors of tumor metastasis by Dr. Jeff Esko's lab in UC, San Diego. They are high affinity substrates for b4GalT1 and thus act as decoys for the synthesis of sialyl Lewis X (sLeX ), the cell adhesion epitope, which is expressed at elevated levels in metastatic cells. The most effective compound they have identified to date is GlcNAc-b3Gal-b-O-napthalenemethanol. The acetylated compound is taken up by the cells, O-deacetylated, and then the disaccharide decoys the synthesis of sLeX-containing glycans on cell surface glycoconjugates. In several model systems this results in an inhibition of tumor formation. As Dr. Esko's lab has observed, all the activity of the compound depends on the action of b4GalT(s) in the cell, since the first step in its utilization involves galactosylation. In their in vitro studies with GlcNAc-b3Gal-b-O-napthalenemethanol as an acceptor substrate for b4Gal-T1, they show a Km of 10 vM. This high affinity for the enzyme may be its unique mode of binding to b4Gal-T1 similar to the one we have observed with the 1,2-1,6-arm trisaccharide (see above). We have carried the crystal structure of h-M340H-Gal-T1 with GlcNAc-b3Gal-b-O-napthalenemethanol, provided by Dr. Esko's lab, and determined the mode of interaction between the disaccharide and the enzyme. The overall binding of GnGl-NP to the Met344His-Cys342Thr-Gal-T1 molecule is quite similar to the binding of the tri-saccharide GlcNAcb1-2Mana1-6Mana, observed earlier (see 2006 project # Z01 BC 010041), suggesting its Km for the enzyme will be similar to that of the 1-6 arm tri-saccharide, a Km of 60 vM. This tri-saccharide is derived from the 1-6 arm of the biantennary N-glycan, starting from the core mannose to the free GlcNAc at the non-reducing end. In the enzyme-bound 1-6 arm tri-saccharide, the middle mannose exhibits the least interactions with the protein molecule, while the terminal mannose (i.e., the core mannose residue of the biantennary N-glycans) makes extensive stacking interactions with the aromatic side chain of the Tyr282 residue.In contrast, the beta-linked Gal residue in the disaccharide GnGl-NP forms extensive stacking interactions with Tyr282, and the terminal aromatic naphthalene residue makes additional interactions, although weak, in the oligosaccharide binding site of the b4Gal-T1 molecule. Therefore, due to the additional interactions that arise from the naphthalene moiety, it is expected that GnGl-NP disaccharide will have a lower Km than that of the 1-6 arm tri-saccharide, thus making it the best known acceptor substrate with the highest affinity for b4Gal-T1. Furthermore, the b4Gal-T family members T5 and T6 have the conserved residue Tyr corresponding to the Phe356 residue of the b4Gal-T1. The Tyr residue in these family members is expected to make additional hydrogen bonding interactions via the side-chain hydroxyl group of the Tyr residue, which is expected to further lower the Km of the GnGl-NP disaccharide for b4Gal-T5 and b4Gal-T6. From the crystal structure analysis we are able to understand the high affinity of the disaccharide GnGl-NP for b4Gal-T1. Furthermore, the analysis suggests that future chemical modifications, such as incorporating the structural water molecule in the acceptor design, amino or methylated amino group at the second position of Gal or even appropriate substitutions of polar groups at the naphthalene ring, may improve the affinity of the acceptor substrate and lead to better design of the disaccharide inhibitors for the tumor metastasis.
|Schuyler, Adam D; Jernigan, Robert L; Qasba, Pradman K et al. (2009) Iterative cluster-NMA: A tool for generating conformational transitions in proteins. Proteins 74:760-76|