Broadly neutralising antibodies after vaccination

This fellowship studies the response to vaccination in humans. The SARS-CoV-2 pandemic has re-emphasised the importance of developing efficient vaccines against existing, and new, infectious diseases. Existing SARS-CoV-2 vaccines offer protection from disease in several ways. The most important way is through training the immune system to make neutralising antibodies, which bind the virus and prevent it from entering cells. This is how the monoclonal antibodies used to treat COVID-19 work - by coating the virus and preventing it from infecting nearby cells - and these treatments can clear SARS-CoV-2 in patients who otherwise lack an immune system.

Some individuals respond to the SARS-CoV-2 vaccines by producing antibodies that only bind the virus, and do not neutralise it. These individuals are probably still vulnerable to infection, even if the other parts of their vaccine-trained immune system can prevent severe disease. The ultimate goal for vaccination is to prevent any infection, even if mild, so as to also help control transmission. To achieve this, vaccines need to induce neutralising antibody.

Variants of concern (VOC) of SARS-CoV-2, make this more challenging. We find that after two vaccinations some individuals make antibodies that only neutralise some of the VOCs. For example, nearly everyone needs three doses of vaccine to neutralise Omicron, whereas many healthy individuals could neutralise Delta after two doses. A small number of individuals could neutralise Omicron after two doses only. Why might this be? What is it that controls the breadth of VOC-neutralising antibodies? Are there ways to influence to breadth of neutralisation, so we can make vaccination even more effective, or streamline the need to repeat doses for new VOCs?

This fellowship seeks to address these questions. I will study the breadth of neutralisation of different VOCs in several cohorts of healthy individuals, and their patterns over time. These cohorts are the SIREN study of over 140,000 healthcare workers around the UK, it's detailed sub-cohort studies PITCH and VIBRANT, and the Legacy study of over 500 healthcare and laboratory workers in northwest London. We will also study the breadth and trajectories of neutralising antibody in UK haemodialysis patients, and I have shown this patient group are poor responders to SARS-CoV-2 vaccines and are at an excess risk of hospitalisation or death from COVID-19. This work will take place in the Francis Crick Institute and UCL. We will use the Crick's high throughput live virus microneutralisation assay to define these antibody patterns.

Having defined individuals with different breadths of neutralisation, I will study the B cells in their blood that bind to the virus' spike protein, the target of current vaccines. Looking at the antibody these B cells makes allows me to study how carefully these antibodies have been selected by the immune system. Do different people's immune systems make selection decisions about antibody? Can we unpick those decisions, by looking at the B cells that have been selected by that process? Does the type of the first two vaccines matter?

Next I will look for mechanisms that control this process, as we may want to try to dampen or enhance them at the time of vaccination, or during antibody mediated autoimmune diseases. Genetic and environmental factors will perturb these mechanisms. In this fellowship, I will examine underlying genetic factors (as the genetic results inform which environmental factors to prioritise for study). The mechanisms are likely to reflect important parts of the decision making process for antibody generation in the human immune system, so could underpin new approaches to vaccination, and antibody-mediated disease treatments.

In this fellowship, I will use high-throughput live virus microneutralisation assay to define a novel immune phenotype of the breadth SARS-CoV-2 VOC neutralisation. After two doses vaccine, most individuals neutralised ancestral SARS-CoV-2, and neutralised Alpha, Beta and Delta in a stepwise deterioration. Relatively few individuals neutralised ancestral, D614G, Alpha, Beta and Delta. The breadth phenotype is made stark by the emergence of Omicron, where a fraction, but not all healthy individuals neutralise Omicron BA.1 after two doses, and most require 3 doses to neutralise Omicron BA.1 or BA.2. That further encounters with the same ancestral (vaccine) spike alters the breadth of response is unexpected, as the breadth only weakly correlates with (ancestral) S1 binding antibodies by ELISA, or with full-length, glycosylated (ancestral) S by flow cytometry.

I propose that the breadth of VOC neutralisation reflects the initiation of somatic hypermutation and the B cell pruning that occurs during affinity maturation. Super-neutralisers, who neutralise all of the tested VOCs (ancestral, D614G, Alpha, Beta, Delta and Omicron: BA.1 and BA.2), might select very high affinity antibody that is heavily somatically hypermutated, or select particular epitopes for neutralisation that are conserved between all VOCs - for example the epitope targeted by sotrovimab. Understanding the diversification and selection of these antibodies is important for the further development of vaccines to any pathogen requiring neutralising antibodies, and allows a new phenotypic description of the germinal centre and extra-follicular antibody responses for QTL mapping.