Adenovirus-vectored vaccines offer an effective platform for the delivery of viral antigen, but it is important for the generation of neutralizing antibodies that they produce appropriately processed and assembled viral antigen that mimics that observed on the SARS-CoV-2 virus.
The Spike (S) protein is responsible for mediating host-cell entry, with the S1 and S2 subunits facilitating ACE2 receptor binding and membrane fusion, respectively. The SARS-CoV-2, S gene encodes the extensively glycosylated trimeric Spike (Trimeric means that each spike is composed of three spike proteins.
Both Moderna’s and Pfizer’s mRNA vaccines encodes full length S with two mutations to stabilise the prefusion conformation and Sinovac’s CoronaVac inactivated virus vaccine presents the wild-type Spike on the viral surface although the majority of spikes are in the postfusion conformation. I explain the difference between pre-fusion and post-fusion Spike here >> https://coviddatareview.wordpress.com/2021/02/09/a-potential-reason-why-the-oxford-astrazeneca-vaccine-is-less-effective-at-providing-protection-against-the-south-african-b-1135-variant-than-some-of-the-other-vaccines/
One key aim for SARS-CoV-2 vaccine development is to elicit a robust immune response against the spike, and more specifically the Receptor Binding Domain (RBD). The RBD is the part of the Spike that is used by the virus to attach to the ACE2 receptors on body cells. To this end, many vaccine candidates include (two or more) stabilising mutations in the Spike (S) protein, such that the protein maintains the prefusion conformation and avoids shedding of Subunit 1 (S1).
The Oxford AZ adenovirus-vectored vaccine encodes the full-length wild-type SARS-CoV-2 Spike protein. The Oxford AZ vaccine has previously been shown to elicit not only strong neutralizing antibody responses, but also robust spike-specific T-cell responses. Although adenovirus-vectored vaccines are a promising way to deliver viral glycoprotein antigens, the processing and presentation of the SARS-CoV-2 spike is yet to be characterised. Understanding the molecular features of the expressed viral antigen (Spike) is important for the interpretation of the immune response to this vaccine.
The team at Oxford first detected the presence of the Spike glycoprotein at the cell surface of cells infected with the Oxford AZ vaccine vector. Sera from mice vaccinated with the Oxford AZ vaccine was used to detect the expression level of Spike at the cell surface, revealing that 60-70% of the cells expressed the Virus Spike.
They then examined the properties of the Spike expressed on the cell surface using ACE2 and human antibody which can only bind to specific regions of the Spike protein. ACE2 is the receptor found on body cells that the virus uses to gain entry. They found cells infected by Oxford AZ vaccine vector would bind to the ACE2. This indicates the infected cells possessed Spike protein in the desired pre-fusion state, which enabled the Spike RBD to attach to ACE2.
This observation is further supported by the binding of the human antibody. The human antibodies used included; Ab45 an antibody which binds to the Spike RBD, Ab71 which recognises the trimeric spike, Ab111 which recognises the N-terminal Domain (NTD) and Ab44 which recognises Subunit 2 (S2). All of these antibodies demonstrated considerable binding. These data confirm significant presence of the pre-fusion (trimer) spike protein at the cell surface. In the absence of a post-fusion specific anti-S2 antibody they were unable to quantify if some post-fusion Spike is present at the cell surface.
Whilst this observation that the majority of cells infected with Oxford AZ vaccine vector present pre-fusion (trimer) Spike on the cell surface it is interesting to note that a population may shed the S1 subunit. Whether this is a beneficial or detrimental feature with respect to the elicitation of immune responses during vaccination is unknown. Shedding of S1 subunits from viruses occurs during native infection, and the Oxford AZ vaccine derived S proteins mimics this native feature of the viral spike.
The surface expression of the Spike driven by Oxford AZ vaccine vector infection was confirmed by cytometry. They imaged the cells infected with the Oxford AZ vaccine vector. The tomograms revealed that surface of the cells is densely covered with protruding densities consistent with the size and shape of the prefusion conformation of SARS-CoV-2 Spike.
In order to investigate the presence of Spike protein in pre-fusion state, they employed template searching for pre- and post-fusion Spikes on cells infected with the Oxford AZ vaccine vector. This analysis revealed a presence of abundant pre-fusion Spikes on the cell surface, while little post-fusion Spike was detected.
The presentation of pre-fusion Spike was achieved through encoding SARS-CoV-2 Spike protein in the ChAdOx1 (Oxford AZ vaccine vector) backbone without the incorporation of stabilising mutations. Given that most neutralizing antibodies target epitopes (binding sites) displayed on the pre-fusion Spike, the analysis revealing the expression of trimeric SARS-CoV-2 Spike in the prefusion conformation strongly supports Oxford AZ vaccine as an effective vaccine strategy for the generation of a neutralizing immune response.
The SARS-CoV-2 S gene encodes 22 N-linked glycosylation sequons which span both the S1 and S2 subunits. These host-derived glycans mask immunogenic protein epitopes (binding sites) from the humoral immune system. (Humoral immune system = antibodies). This is a common strategy utilised by viruses to evade the immune system. It is therefore of considerable importance that Spike produced upon vaccination successfully recapitulates the glycosylation observed from infection by the actual SARS-CoV-2 virus. If this doesn’t occur antibodies may be produced against protein epitopes (binding-sites) that are hidden by glycans during natural infection.
Using the cryo-EM structure of the trimeric SARS-CoV-2 S protein, Oxford mapped the glycosylation status of the S1/S2 protein. Observation of Glycan sites on the Spike proteins on the surface of cells that were infected by Oxford vaccine vector provides additional evidence of the Spikes being in the pre-fusion state. Glycan sites observed included N234 and N165 which are known to have stabilising effects on the RBD.
Adenoviruses can infect many cell types in the body due to the widespread distribution of their cellular receptors such as coxsackie, adenovirus receptor (CAR) and CD46. Replication-deficient adenovirus vectored vaccines, such as the Oxford vaccine vector are administered intramuscularly and predominantly induce antigen expression in muscle cells, fibroblasts and professional antigen presenting cells (APCs). APCs are activated either through direct adenovirus infection or through cross priming by non-immune cells that have been infected with the vaccine vector.
Spike protein processing within the infected cell is impacted by the enzymes and surface receptors that are expressed. Presence of the cell entry receptor ACE2 and host-cell proteases including furin and TMPRSS2 can drive conformational rearrangements resulting in the undesired post-fusion Spike. Tissue distribution of ACE2 is primarily in the nasal mucosa and Gastro-Intestinal tracts, and TMPRSS2 is predominantly expressed in lungs and Gastro-Intestinal tracts, where cells are susceptible to SARS-CoV-2 infection. In contrast, the muscle, fibroblast and immune cell types targeted by intramuscularly administered Oxford Virus vector, have not been shown to express high levels of ACE2 or TMPRSS2, which should significantly lessen the likelihood of the Spike being rearranged into the undesired pre-fusion state.
Oxford conclude that their vaccine predominantly exposes the body to the desired prefusion Spike protein, complete with the important glycan modifications. There is however some evidence of S1 subunit shedding, which is also observed during natural infection with the SARS-CoV-2 virus. They note that the impact of this breakdown of the trimeric spike protein on immunogenicity and vaccine efficacy is yet to be determined.
They end by saying that they hope the data presented here will assist comparison across SARS-CoV-2 vaccination strategies and aid the development of next- generation vaccines.