Engineering glycosylation in HIV-1 vaccine design

Abstract

The development of a protective HIV-1 vaccine remains a major challenge to the scientific community. The nature of the virus necessitates an immune response that is able to overcome the significant genetic diversity of HIV-1. The elicitation of broadly neutralizing antibodies offer a route to overcome this diversity. As the only viral antigen expressed on the virion surface, all bnAbs target the HIV-1 Envelope spike, and as such all candidate immunogens are based around this protein. The envelope glycoprotein is one of the most densely glycosylated proteins in nature and this dense glycan shield provides protection to conserved regions of the protein from the immune system and this is reflected by the conservation of key glycan attachment sequons despite a background of formidable sequence diversity. Broadly neutralizing antibodies exploit this conservation to allow for their broad recognition through recognizing both the protein and the glycans of Env. As of yet, recombinant immunogens have been unable to elicit bnAbs in animal models and humans. This is due to two main facets. One is that the pathway of somatic hypermutations that are reuqired to generate bnAbs are rare, and in order to guide a glycan-binding antibody to maturation there are self-reactivity controls that need to be overcome which particularly hinder glycan-binding bnAb development. Also, off-target responses distract the immune system and result from key differences in Env glycosylation between the virus and recombinant proteins. Engineering the glycosylation of Env, therefore, represents a potential mechanism to improve recombinant immunogens towards eliciting bnAbs. There are numerous intrinsic and extrinsic properties that influence Env glycosylation, including structural constraints and enzymatic availability, that can be harnessed to alter Env glycosylation. In this thesis, multiple approaches are investigated with the goal of editing the Env glycan shield towards one that is favourable for the elicitation of bnAbs. Due to the heterogeneity of glycosylation, bespoke workflows are needed to analyse glycosylation, which is required to validate the effectiveness of different engineering approaches. To achieve this, a methodology based around liquid chromatography-mass spectrometry (LC-MS) was used to comprehensively determine the site-specific glycosylation of Env. First, the impact of glycan additions and deletions on the overall processing of the glycan shield was determined. Such additions and deletions are commonplace in immunogens aimed at eliciting bnAbs. This revealed that additions/deletions are well tolerated but their induction influences the processing of neighbouring glycan sites. These observations were consistent across multiple Env strains and immunogen platforms. Next, the impact of altering the availability of glycan processing enzymes on Env glycosylation was explored, demonstrating that co-expressing key enzymes in the pathway was successful in engineering the glycan shield, but these effects were unpredictable. Both of these approaches resulted in widespread alterations in the glycan shield. To limit glycan engineering to specific epitopes, the targeted epitopes of bnAbs were exploited to control glycosylation by co-transfecting Env with bnAbs, which in turn alter the availability of Env to glycan processing enzymes during glycosylation in the ER/Golgi. Finally, to alter the glycosylation of Env towards a more viral-like configuration, the gene codon usage of recombinant Env was altered to that of the native virus, which contrasted the codon optimized variants typically in use. This resulted in a decreased rate of translation, which, in turn, altered the glycan shield of recombinant Env to a more viral like state. This thesis demonstrates the wide arsenal of approaches that can be used to change Env immunogen glycosylation, however it is difficult to predict the outcomes of such engineering without comprehensively studying the resultant immunogens. By investigating a broad range of approaches and reporting their successes and caveats it is possible for these methods to be integrated into immunogen design approaches with specific epitopes in mind. These tools may prove valuable in the design of an effective HIV-1 vaccine with an appropriate glycan profile.

More information

Identifiers

ORCID for Alessio D'Addabbo: orcid.org/0000-0003-2542-6465

ORCID for Max Crispin: orcid.org/0000-0002-1072-2694

Catalogue record

Export record