Phenotypic and transcriptomic characterisation of the human B cell response to a novel Ebola vaccine regimen (Ad26.ZEBOV, MVA-BN-Filo)

Ebola virus disease (EVD) outbreaks are increasing in frequency — threatening health and wellbeing of affected communities. Early and effective public health measures to manage outbreaks rely on health care and frontline workers; consequently, protecting these at high-risk groups is a key pillar of EVD vaccination strategies. IgG specific to the viral glycoprotein is used as the correlate of protection in recent vaccine licensure. Here, we use contemporary cellular and transcriptomic analyses to dissect B cell responses to the heterologous Ad26.ZEBOV, MVA-BN-Filo Ebola vaccine. We reveal robust plasma cell and persistent B cell memory responses following vaccination. Using machine-learning, trained on blood gene expression post-vaccination, we were able predict the magnitude of antibody responses. Moreover, we described a B cell receptor CDRH3 sequence which was exclusively seen post-vaccination that is similar to known Ebola GP-binding antibody sequences. Single cell analysis was used to further characterise changes to plasma cell frequency, subclass usage and CDRH3 properties, following vaccination. These results demonstrate that early immune responses —captured in blood using systems immunology approaches — are delineative and predictive of B cell responses to vaccination. Venous peripheral blood (up to 3ml) was collected directly in TempusTM RNA tube immediately pre-vaccination, then 10 days post-first vaccine and 7 days post-second vaccine. RNA was extractedusing theTempus™spin RNAisolationkit(Invitrogen, USA).Eluted RNA was quantifiedusingQubit™ RNA BR Assay Kit(Invitrogen, USA) and quality control was conducted usingthe Agilent RNA ScreenTape assay on the 4200 TapeStation System (Agilent, USA). RNA was ribodepleted and globin depleted usingRibo-ZeroTM Gold rRNA removal Kit (Illumina, USA). RNA was converted to cDNA, second-strand cDNA synthesis incorporated dUTP. The cDNA was end-repaired, A-tailed and adapter-ligated and prior to amplification, samples underwent uridine digestion. The prepared libraries were size selected, multiplexed and quality controlled before 75bp paired-end sequencing (HiSeq4000). Sequencing was conducted at the Wellcome Trust Centre for Human Genetics (Oxford, UK). The sequencing data were aligned against the whole human (Homo sapiens) genome build GRCh38 (https://ccb.jhu.edu/software/hisat2/index.shtml), using STAR (version 2.6.1d). Gene features were counted using HTSeq (version 0.11.1), using human gene annotation general transfer format version GRCh38.92 (www.ensembl.org).