David V. Renouf

The association of an HPV16 oncogene variant with HLA-B7 has implications for vaccine design in cervical cancer.

HLA-restricted cytotoxic T-lymphocyte (CTL) recognition of human papillomavirus (HPV) oncogene products may be important in the control of the HPV infections associated with the development of cervical cancer. We have identified, in HLA-B7 individuals, a consistent variation in the HPV16 E6 oncoprotein sequence, which alters an HLA-B7 peptide binding epitope in a way likely to influence immune recognition by CTLs. These results illustrate a biologically relevant mechanism for escape from immune surveillance of HPV16 in HLA-B7 individuals. Thus, both HLA type and HPV16 strain variation need to be considered in the screening of at-risk individuals and for the rational design of anti-HPV vaccines.

O-linked protein glycosylation structure and function

There has been a recent resurgence of interest in the post-translational modification of serine and threonine hydroxyl groups by glycosylation, because the resulting O-linked oligosaccharide chains tend to be clustered over short stretches of peptide and hence they can present multivalent carbohydrate antigenic or functional determinants for antibody recognition, mammalian cell adhesion and microorganism binding. Co-operativity can greatly increase the affinity of interactions with antibodies or carbohydrate binding proteins. Thus, in addition to their known importance in bearing tumour associated antigens in the gastrointestinal and respiratory tracts, glycoproteins with O-linked chains have been implicated as ligands or co-receptors for selectins (mammalian carbohydrate binding proteins). Microorganisms may have adopted similar mechanisms for interactions with mammalian cells in infection, by having relatively low affinity ligands (adhesins) for carbohydrate binding, which may bind with higher affinity due to the multivalency of the host ligand and which are complemented by other virulence factors such as interactions with integrin-type molecules. In addition to specific adhesion signals from O-linked carbohydrate chains, multivalent O-glycosylation is involved in determining protein conformation and forming conjugate oligosaccharide-protein antigenic, and possible functional determinants.

Glycosylation of immunoglobulin light chains associated with amyloidosis

AL amyloidosis is a fatal disease caused by deposition of immunoglobulin light chains in a fibrillar form (AL) in various organs. By searching the Kabat database of immunoglobulin sequences using the KabatMan software, we have shown that there is a preponderance of the consensus glycosylation sequon (AsnXxxSer/Thr) in the framework regions of amyloid light chains. We have characterised by computer graphics simulations, NMR spectroscopy and carbohydrate biochemistry the structure and conformation of the oligosaccharide from amyloid protein AL MS (λI) and from the amyloid associated Bence Jones protein of patient MH (K1). These proteins have glycosylation in the hypervariable complementarity–determining region versus framework region, respectively. Both contained a 2–6 sialylated core fucosylated biantennary chain mostly with bisecting GlcNAc. Together our results suggest that light chain glycosylation may be one of several modifications which may render the protein more prone to amyloid formation.

A proposed molecular model for the carboxy terminus of HIV-1 gp120 showing structural features consistent with the presence of a T-cell alloepitope

A computer graphics molecular model of the C terminus of gp120 of HIV has been constructed using predicted secondary structure based on homologies with proteins for which X-ray crystallographic data have been published. The model shows sequences known to be important in CD4 binding in close proximity to regions with a high probability of forming alpha helical and beta strand motifs. The orientation adopted by these domains approximates to the known 3D structure of HLA-A2 alpha 2 chain without constraints based on HLA-A2 as a template being introduced. The model may therefore represent an energetically favourable conformation for a part of gp120 which mimics the binding domain for the T-cell receptor on MHC molecules. Recognition of gp120 as an alloepitope in high affinity association with CD4 would explain many of the sequelae of acquired immune deficiency on HIV infection.

Glycosylation and epitope mapping of the 5T4 glycoprotein oncofoetal antigen

The human 5T4 oncofoetal antigen is a focus for development of several antibody-directed therapies on the basis of the murine monoclonal antibody against 5T4 (mAb5T4), which recognizes a conformational epitope. 5T4 molecules are highly N-glycosylated transmembrane glycoproteins whose extracellular domain contains two regions of leucine-rich repeats (LRRs) and associated flanking regions, separated by an intervening hydrophilic sequence. Using a series of deletion and mutated cDNA constructs as well as chimaeras with the murine homologue, we have mapped the mAb5T4 epitope to the more membrane-proximal LRR2 or its flanking region. Analysis of the glycosylation of the seven consensus Asp-Xaa-Ser/Thr sites was consistent with all of the sites being glycosylated. A combination of two high-mannose chains (predominantly octasaccharide) and five mostly sialylated bi-, tri- and tetra-antennary complex chains with minor quantities of core fucose were detected. The two glycosylation sites, which are the most likely to have predominantly high-mannose chains, are in the only two regions that show significant differences between the human and the 81% identical mouse sequence. A site-directed mutation, which abolished glycosylation at one of these sites (position 192), did not alter antigenicity. The other, which is nearest to the N-terminus in the human, has an Asn-Leu-Thr to Asn-Leu-Leu conversion in the mouse, so cannot be glycosylated in the latter species. The large complex glycosylation at the other sites is likely to influence the antigenicity and tertiary structure generating the 5T4 epitope.

Glycosylation of prions and its effects on protein conformation relevant to amino acid mutations.

The three-dimensional coordinates from a nuclear magnetic resonance (NMR)-averaged structure containing residues 121-226 of mouse prion were used as the starting geometry for MD of prion either with or without glycan in both mutant and wild-type forms. The following mutants were studied: Asp-178 to Asn, Thr-183 to Ala, Phe-198 to Ser, Glu-200 to Lys, and Gln-217 to Arg. NMR data vs structural models were compared to observe any major differences. Simulations of the change in protein structure with and without glycan were performed, as they cannot be tested by NMR analysis. Several mutants were expressed and analyzed for altered glycosylation and the results interpreted in terms of molecular modeling. N-linked glycosylation is likely to play an important role in prion biology as shown by visualization of glycoprotein conformation.

Analysis of the Glycosylation Patterns of the Extracellular Domain of the Epidermal Growth Factor Receptor Expressed in Chinese Hamster Ovary Fibroblasts

The extracellular domain (621 N-terminal amino acids) of the p170 epidermal growth factor (EGF) receptor has eleven consensus N-linked glycosylation sites. When expressed in Chinese hamster ovary cells this was glycosylated with a combination of high mannose and complex chains. The latter chains were shown by chromatographic separation and mass spectrometric analysis of tryptic digests to be clustered in the EGF-binding domain. Treatment with the endoglycosidase, peptide-N-glycosidase F (PNGase F), reduced the molecular weight from 110 kDa to 75 kDa. Released oligosaccharides were characterised at high sensitivity by high pH anion exchange chromatography with pulsed amperometric detection and gas-liquid chromatography/mass spectrometry. The data were consistent with the complex chains being trisialylated tetra-antennary oligosaccharides fucosylated on the reducing terminal GlcNAc. The large hydrodynamic mass of these oligosaccharides could influence ligand binding, an effect which is likely to vary with the difference in consensus glycosylation sites of proteins related to p170 i.e. p185erbB2/neu, p180erbB3 and p180erbB4.

Conformational studies of the backbone (poly-N-acetyllactosamine) and the core region sequences of O-linked carbohydrate chains.

The AMBER forcefield with added carbohydrate specific parameters for potential types, charges and anomericity has been used to explore the conformational space sampled by oligosaccharides of repeating (Gal beta 1-4GlcNAc beta 1-3)n sequence and those having alpha chain terminating substituents on a Gal beta 1-3GalNAc alpha, O-glycosylated core. The 17 lowest energy forms of the 16-mer N-acetyllactosaminic sequence included fully extended units and spring-like coils with dielectric constant 80 to simulate a water environment. More restricted stereochemistry was exhibited by the glycosylation adjacent to the core as has been previously predicted by n.m.r. studies.

Molecular Modelling of Glycoproteins by Homology with Non-Glycosylated Protein Domains, Computer Simulated Glycosylation and Molecular Dynamics

OBJECTIVES This study aims to visualise glycoproteins by computer graphics molecular modelling in order to research the dynamics of the oligosaccharide chains, determine their affects on protein conformation, antigenicity and function and to characterise oligosaccharide recognition determinants. With respect to the last, the modelling included the sialylpolylactosamine of thymocyte Thy-l and the sialyl Le(x)/Le(a) determinant present on brain Thy-l. METHODS The following techniques were used: 1) database searching for homologies with non-glycosylated protein domains; 2) protein modelling on the basis of homology and secondary structure prediction techniques; 3) oligosaccharide construction using a simulated annealing approach utilising the AMBER forcefield with appropriate parameters in the Biosyn software environment; 4) creation of glycoprotein conjugates for further investigation by energy minimisation and molecular dynamics. RESULTS This approach was successful in providing models of Thy-l and the carboxy terminus 27.5 kD of HIV-1 gp 120, by homology with immunoglobulin light chain folds and in one case (Thy-l) adding the oligosaccharide chains, phosphatidylinositol glycan anchor and lipid membrane and, in the other, adding additional highly glycosylated domains, and domains which folded by molecular dynamics. Significant affects on protein conformation were shown in the presence or absence of the lipid anchor and simulated membrane, but not by the N-linked oligosaccharide chains. CONCLUSIONS The highly glycosylated molecules Thy-l and gp120, which are not expected to crystallise in their native state, were modelled by computer graphics simulated annealing or molecular dynamics from which interactions could be predicted which agree with experimental data on antibody binding and in vitro activity.

Specificity studies of an antibody developed against a mucin-type glycoprotein

The specificity of a new anti-epiglycanin antibody (AE-3) which recognizes a mucin-type glycoprotein, the Human Carcinoma Antigen, found in the blood of patients with carcinomas, was studied. Information regarding the chemical nature of the antibody binding site was obtained by altering the structure of epiglycanin by chemical or enzymic means and testing the product in a competitive binding assay for inhibition of the binding of AE-3 to epiglycanin. The need for a high molecular weight antigen containing clustered T disaccharide, Galβ1-3GalNAc, was demonstrated. The specificity was further explored by inhibition studies with glycopeptides having one to three mono- to disaccharides. The results were interpreted using computer graphics molecular modeling which predicted the specific recognition of hydroxyl groups on oligosaccharides on adjacent amino acids. Thus T antigen O-linked glycopeptide tumour markers can be designed to be distinguished by antibodies by the amount of clustering of their oligosaccharides.