Impact of BNT162b First Vaccination on the Immune Transcriptome of Elderly Patients Infected with the B.1.351 SARS-CoV-2 Variant

Fast-spreading variants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) energize the COVID-19 pandemic. The B.1.351 variant carrying the escape mutation E484K in the receptor binding domain is of particular concern due to reduced immunological protection following vaccination. Protection can manifest as early as 10 days following immunization with full protection two weeks following the second dose, but the course is not well-characterized for variants. Here, we investigated the immune transcriptome of six elderly individuals (average age 82 yr.) from an old people’s home, who contracted B.1.351, with four having received the first dose of BNT162b eight to 11 days prior to the onset of COVID-19 symptoms. The patients were hospitalized and received dexamethasone treatment. Immune transcriptomes were established from PBMCs approximately 10 and 35 days after the onset of COVID-19 symptomology. RNA-seq revealed a more intensive immune response in vaccinated patients as compared to unvaccinated ones. Specifically, transcription factors linked to the JAK/STAT pathway, interferon stimulated genes, and genes associated with innate antiviral immunity and COVID-19-SARS-CoV-2 infection were highly enriched in vaccinated patients. This rendered the transcriptomes of the older vaccinated group significantly different than older unvaccinated individuals infected at the same institution and more similar to the immune response of younger unvaccinated individuals (ages 48–62) following B.1.351 infection. All individuals in this study whether vaccinated or not were hospitalized due to B.1.351 infection and one vaccinated patient died illustrating that although an enhanced immune response was documented infection it was insufficient to protect from disease. This highlights the need for maintaining physical distancing and prevention measures throughout the time course of vaccination in older adults.


Introduction
The SARS-CoV-2 variant B.1.351 1 has multiple changes in the immunodominant spike protein that facilitates viral cell entry via the angiotensin-converting enzyme-2 (ACE2) receptor. The receptor-binding domain (RBD) mutation E484K provides tighter ACE2 binding and widespread escape from monoclonal antibody neutralization 2-5 , a concern for a limited protection by some vaccines 6 , and recent data from Qatar demonstrate a BNT162b vaccine 7 effectiveness of 75% with the B.1.351 variant 7 . Vaccine response as well as the immunological response to SARS-CoV-2 infection are time-dependent, with rst a rise and then a fall to normal levels of immune-related gene transcription 8 . Differences in sampling times post vaccination and/or infection strongly in uence results derived. Limited data is available on the comparative response of the immune transcriptome between infected individuals who are vaccinated versus unvaccinated following the rst dose. Since the immune response declines with age, it is most critical to explore this question in the older age group that is most affected by the COVID-19 pandemic.
Here, we conducted a study on elderly patients in an old-people's home that experienced an outbreak of the SARS-CoV-2 B.1.351 variant. A unique aspect of this single setting study is that the infected individuals were all exposed within the same timeframe and vaccinated individuals were all immunized within the same timeframe, controlling for environmental and kinetic variables.

Study design
To better understand the extent of protection of elderly individuals from COVID-19 after receiving the rst dose of BNT162b vaccine, we investigated a unique cohort of hospitalized patients from an old people's home that experienced an outbreak of the SARS-CoV-2 variant B.1.351 in January of 2021 with similar kinetics of vaccination, infection and disease development. Six elderly individuals (average age 82 yr.) developed COVID-19 symptoms whole viral genome sequencing con rmed the B.1.351 strain (Supplementary Table 1). Four patients had received the rst dose of the BNT162b (P zer-BioNTech) vaccine 11 days before the onset of COVID-19 symptoms and one patient eight days prior (Fig. 1a). All patients had underlying health conditions, they were hospitalized and cared for by the same physician and received dexamethasone treatment. Three additional hospitalized B.1.351-positive unvaccinated patients (average age 56) from other communities in Tyrol and South Tyrol (Italy) were included in the broader study ( Fig. 1a; Supplementary Table 1). Having this well-controlled cohort of individuals with equivalent ages from the same nursing home, provided the opportunity to explore the impact of the rst vaccination on the immune transcriptome in hospitalized COVID-19 patients.

Immune transcriptome
Blood was drawn at ~10 and 30 days after the development of COVID-19 symptoms, RNA was prepared from buffy coats followed by RNA-seq analyses. The principal component analysis (PCA) showed a separation between RNA-seq samples from non-COVID controls 8 and the rst and second time points from COVID-19 patients on the rst principal component (PC1) (Fig. 1b). In total, 3410 genes were signi cantly differentially expressed between non-COVID controls and the nine COVID-19 patients infected with the B.1.351 variant. The expression of 2026 genes was elevated in COVID-19 patients compared to non-COVID controls ( Fig. 1c; Supplementary Table 2). As shown previously 9,10 , the signi cantly up-regulated genes are enriched in immune response pathways, including IL-JAK/STAT signaling and interferon alpha/gamma responses (Fig. 1d). Expression of the angiotensin-converting enzyme 2 (ACE2) receptor 11,12 was elevated in the patients (Fig. 1e) suggesting that that the classical or novel dACE2 promoter responded to immune activation 13,14 . Of note, the immune response was linked to viral defense mechanisms (Fig. 1f).

Elevated immune response in patients infected after the rst vaccination
Next, we focused on the immune transcriptome response in six hospitalized elderly patients (average age 82 yr.) from the same old people's home (Supplementary Table 1), four of which had received the rst dose of the BNT162b vaccine 11 days prior to developing COVID-19 symptoms. All six patients developed COVID-19 within two days of each other, were admitted to the same hospital, treated with Fortecortin (Dexamethasone) and were cared for by the same physician. First, we generated the transcriptomes from PBMCs isolated between 9 and 12 days after developing COVID-19 ( Fig. 1a; Supplementary Table 1). The PCA plot shows a separation between the vaccinated and unvaccinated groups on PC1 (Fig. 2a), which is further strengthened by hierarchical clustering of the 181 signi cantly differentially expressed genes between the six patients ( Fig. 2b; Supplementary Table 3 Table 3). Speci cally, the highly ranked ETV7, STAT1 and STAT2 ( Fig. 2f) are key transcription factors executing interferon responses and are top predictors for ACE2 expression in human airway epithelium 14,[16][17][18][19][20] . Notably, expression of IFITM3, a gene with a polymorphism that has been linked to the severity of COVID-19 21 , is highly induced in vaccinated patients ( Supplementary Fig. 1a). While three out of the four vaccinated patients had a highly activated immune transcriptome, patient ID-02 had a blunted immune response, more similar to that seen in the nonvaccinated patients ( Supplementary Fig. 1b). One explanation is the shorter time period between vaccination and the onset of symptoms.
Temporal progression of the immune response From our cohort of the four vaccinated patients, one died of COVID-19 and three were discharged from the hospital after a single stay of 10-14 days (Fig. 1a). From the two non-vaccinated patients, one was discharged after a single ten day stay and one was re-admitted. To dig deeper into the temporal immune response after discharge from the hospital we analyzed the immune transcriptome of the three vaccinated patients approximately three weeks after the rst transcriptome analysis (Figs. 1a and 3; Supplementary Table 4). Gene enrichment analyses of differentially expressed transcripts illustrate a greatly diminished immune response, including INFa/g signaling, in the recovered patients ( Fig. 3a; Supplementary Table 4). Expression of speci c ISGs (e.g., IFI35, IFIT5, IFITM3, STAT1, STAT2, OAS2) was sharply reduced (Fig. 3b), approaching levels seen in non-COVID samples.

Immune responses at different ages
Having analyzed the immune transcriptome from the elderly hospitalized population (average age 82 yr.) we addressed potential age differences and included RNA-seq transcriptomes from three hospitalized patients all below 62 years of age (average age 56 yr.) (Fig. 1a). The two groups were separated in the PCA plot ( Fig. 4a; Supplementary Table 5 Table 5). These results support the concept that age might be a de ning factor in the immune response in COVID-19 patients.

Discussion
COVID-19 is a rapidly evolving eld in which real-world data, especially those concerning the now prevalent variants of concern (VOC), is critical for everyday management of the COVID-19 disease by public health o cials, government agencies and institutions. There have been no studies of whole transcriptome response following rst vaccine in infected versus uninfected individuals.
It is known that the rst vaccine dose is not fully protective 22  Our study illustrates that whole transcriptome study from buffy coat is feasible and yields actionable data when performed in well-controlled clinical settings. It demonstrates a technique that is feasible to expand and reproduce across settings. While our study was limited to infection ~11 days following rst vaccination, extensions could compare the immune transcriptome after the second vaccination and investigate individuals who may have waning antibody response ten or more months following either infection and/or full immunization. As a caveat, it can be challenging to reliably compare RNA-seq data generated on different platforms and the disease severity at the time of the analysis greatly impacts the transcriptome 10 .
While there was a measurable immune transcriptome response to rst vaccination, it was insu cient to prevent signi cant clinical disease and protective measures against spread should be fully employed in individuals who have received only one dose. Moreover, SARS-CoV-2 variants of concern (VOC) such as B.1.351 studied here demonstrate resistance to neutralization by antibodies generated from some current vaccines 11 . Further research in vaccinology will continue to address mechanisms to improve the protective immunological response to SARS-CoV-2 and VOC through boosters and vaccines engineered speci cally for VOC. Serial immune transcriptome studies should be included in addition to antibody tests for a fuller understanding of the spectrum of immune response in real-world situations as more immunized individuals encounter SARS-CoV-2 VOC.

Limitations
The ndings in this report are subject to at least three limita tions. We have investigated the transcriptome from patients infected with B.1.351 but not with other variants. Our study focused on elderly (average age 82 yr.) vaccinated and unvaccinated patients with limited comparison to younger individuals (48-62 yr).
We have investigated the effect of the BNT162b vaccine but not of any other vaccine. The study group was small; however, the parallel timing of vaccination, infection and treatment in the elderly group enabled the study to be well controlled for environmental factors.

SARS-CoV-2 virus sequencing
RNA was extracted from patient's blood using a Maxwell RSC simply RNA Blood puri cation kit according to the manufacturer's instructions (Promega, USA). Library preparation and sequencing was performed as described 26 . In short, cDNA was obtained by using reverse transcriptase with random priming. Following cDNA synthesis, primers based on sequences from the ARTICnetwork were used to generate 400 bp amplicons in two different PCR pools. After merging of pools and ampli cation, libraries were constructed using QIASeq FX DNA Library UDI Kit following the manufacturer's instructions (Qiagen GmbH, North Rhine-Westphalia, Germany).

Extraction of the buffy coat and puri cation of RNA
Whole blood was collected, and total RNA was extracted from the buffy coat and puri ed using the Maxwell RSC simply RNA Blood Kit (Promega) according to the manufacturer's instructions. The concentration and quality of RNA were assessed by an Agilent Bioanalyzer 2100 (Agilent Technologies, CA). mRNA sequencing (mRNA-seq) and data analysis.
The Poly-A containing mRNA was puri ed by poly-T oligo hybridization from 1 mg of total RNA and cDNA was synthesized using SuperScript III (Invitrogen, MA). Libraries for sequencing were prepared according to the manufacturer's instructions with TruSeq Stranded mRNA Library Prep Kit (Illumina, CA, RS-20020595) and paired-end sequencing was done with a NovaSeq 6000 instrument (Illumina).
The raw data were subjected to QC analyses using the FastQC tool (version 0.11.9) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). mRNA-seq read quality control was done using Trimmomatic 27 (version 0.36) and STAR RNA-seq 28 (version STAR 2.5.4a) using 150 bp paired-end mode was used to align the reads (hg19). HTSeq 29 (version 0.9.1) was to retrieve the raw counts and subsequently, R (https://www.R-project.org/), Bioconductor 30 and DESeq2 31 were used. Additionally, the RUVSeq 32 package was applied to remove confounding factors. The data were pre-ltered keeping only genes with at least ten reads in total. The visualization was done using dplyr (https://CRAN.Rproject.org/package=dplyr) and ggplot2 33 . Genes were categorized as signi cant differentially expressed with an adjusted p-value (pAdj) below 0.05 and a fold change > 2 for up-regulated genes and a fold change of < -2 for down-regulated ones and then conducted gene enrichment analysis (https://www.gseamsigdb.org/gsea/msigdb).

Statistical analysis
Data were presented as the means ± s.e.m. (standard error of the mean) of all experiments with n = number of biological replicates. For comparison of RNA expression levels between two groups, data were presented as standard deviation in each group and were evaluated with a two-way ANOVA followed by Tukey's multiple comparisons test or a two-tailed unpaired t-test with Welch's correction using GraphPad PRISM (version 9.0). A value of *P < 0.05, **P < 0.001, ***P < 0.0001, ****P < 0.00001 was considered statistically signi cant.

Declarations
Ethics approval

Data availability
The RNA-seq data of patients will be uploaded in GEO before publishing the manuscript.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.