New tools for deciphering the roles of tryptophan C-mannosylation

Abstract

Tryptophan C-mannosylation involves the covalent and irreversible attachment of α-D-mannose onto the C2 carbon of the indole ring of tryptophan forming an unusual C-glycosidic bond that is far more stable than an N- or O-glycosidic bond. C-mannosylation was first discovered on RNaseII in humans in 1994 and there are now over twenty human proteins for which there is evidence confirming the presence of this modification. It is found predominantly on thrombospondin type-1 repeat domains and type-I cytokine receptors. The glycosylation consensus sequence for C-mannosylation is WXXW.

Despite being conserved on many important proteins, the function and biological significance of C-mannosylation remains unknown. Addressing this problem was hampered by the fact that the genetic basis for C-mannosylation was unknown for a long time and because there are limited tools available for installing and detecting this protein modification. This has been remedied to some degree by the discovery that the dpy19 gene in C. elegans encodes a tryptophan C-mannosyltransferase. There are four homologues of dpy19 in humans and recent studies have found that at least two of these encode functional tryptophan C-mannosyltransferases.

I have embarked on the creation of a suite of tools for studying the roles of tryptophan mannosylation to progress this field. A simple Pichia pastoris protein production system was created to produce natively C-mannosylated proteins, with both C2-α-D-mannosyl-tryptophan and without the modification. This was accomplished by integrating C. elegans dpy19 under the strong constitutive GAPDH promoter. This yeast strain enabled the production of a wide variety of recombinant proteins with and without C2-α-D-mannosyl-tryptophan. Two to five-fold improvements in protein production were observed when the protein was C-mannosylated, implying that the modification may play a role in protein folding, trafficking and/or stability. Differential scanning fluorimetry subsequently revealed that C-mannosylation of proteins stabilises them against thermal denaturation. This same yeast system provided a convenient means to explore the substrate specificities of the DPY19 enzyme and probe what enzyme residues were important for enzyme activity.

I have also generated the first monoclonal antibodies for detecting C2-α-D-mannosyl-tryptophan. Using a C-mannosylated peptide derived from recombinant proteins produced in my engineered yeast, I was able to generate five novel monoclonal antibodies that only recognised proteins with the modification. These monoclonal antibodies were validated for use in Western blots and ELISAs and biophysically characterised using surface plasmon resonance techniques. Sequencing the hybridomas that made these antibodies revealed a case of immunoconvergence, where two mice converged on essentially the same solutions for recognising C2-α-D-mannosyl-tryptophan. I was able to crystallise a complex of one antibody’s Fab with the C-mannosylated peptide and determine its structure using X-ray diffraction techniques to reveal the key interactions used to recognise C2-α-D-mannosyl-tryptophan.

Using the engineered yeast, monoclonal antibodies and substrate mimics of C-mannosylation, we were able to discover the first chemical inhibitors of C-mannosylation that can be used to probe C-mannosylation biology.

Lastly, I genetically disrupted the dpy19 homologues in human cells and performed some comparative cell surface proteomic experiments to provide the first insights into how loss of C-mannosylation impacts the human proteome.