key: cord-0304799-mm8ms0hv
authors: Koutsakos, M.; Lee, W. S.; Reynaldi, A.; Tan, H.-X.; Gare, G.; Kinsella, P.; Liew, K. C.; Williamson, D. A.; Kent, H. E.; Stadler, E.; Cromer, D.; Khoury, D. S.; Wheatley, A. K.; Juno, J. A.; Davenport, M. P.; Kent, S. J.
title: Dynamics of immune recall following SARS-CoV-2 vaccination or breakthrough infection
date: 2021-12-23
journal: nan
DOI: 10.1101/2021.12.23.21268285
sha: f930cd91411a1e1e92011e30fa4b493b3378aa0d
doc_id: 304799
cord_uid: mm8ms0hv

Vaccination against SARS-CoV-2 results in protection from acquisition of infection as well as improved clinical outcomes even if infection occurs, likely reflecting a combination of residual vaccine-elicited immunity and the recall of immunological memory. Here, we define the early kinetics of spike-specific humoral and T cell immunity after vaccination of seropositive individuals, and after breakthrough infection in vaccinated individuals. Intensive and early longitudinal sampling reveals the timing and magnitude of recall, with the phenotypic activation of B cells preceding an increase in neutralizing antibody titres. In breakthrough infections, the delayed kinetics of humoral immune recall provides a mechanism for the lack of early control of viral replication but likely underpins accelerated viral clearance and the protective effects of vaccination against severe COVID-19.

Vaccines encoding the spike (S) antigen of SARS-CoV-2 are effective in reducing the risk of symptomatic SARS-CoV-2 infection, as well as progression to severe COVID-19 disease (1). Neutralizing antibodies are a correlate of protection (2, 3) and likely act to prevent infection by blocking viral attachment and entry. However, as antibody levels naturally wane (4), vaccine effectiveness drops (5) and the frequency of "breakthrough infections" among vaccinated individuals increases in the population.

The emergence of antigenic variants including Beta and Omicron have highlighted the potential for viral escape from neutralizing antibody recognition, which can considerably reduce vaccine effectiveness against acquisition of SARS-CoV-2 infection (6). Nevertheless, vaccine-elicited immunity continues to provide robust protection against severe disease outcomes, even in the face of viral variants (7). Viral growth rates and peak viral RNA levels in the upper respiratory tract are similar between vaccinated and unvaccinated infected individuals during the first week of infection (8-10). However, vaccinated individuals consistently display more rapid clearance of viral RNA than unvaccinated controls during the second week of infection (8, 9) . Importantly, there is a lower probability of culturing infectious virus from respiratory samples of infected vaccinated individuals (11). However, the immunological mechanisms that underpin accelerated viral clearance remain unclear.

The comparable viral levels within vaccinated and unvaccinated individuals in the first week of infection suggest that residual (post-vaccination, pre-infection) antibody or T cell immunity fails to limit early viral replication in the respiratory tract. However, the recall of SARS-CoV-2 specific antibodies, memory B and T cell responses following breakthrough infection could contribute to viral clearance and temper disease severity, as is thought to be the case for other respiratory viral infections (12, 13).

Understanding the mechanisms and effectiveness of recall responses in protecting from severe SARS-CoV-2 infection is critical to informing the optimal deployment of current vaccines and guiding the design of novel vaccines to maintain maximal protection against severe disease. To date, however, the precise kinetics of immune recall in the context of SARS-CoV-2 infection are poorly resolved.

To understand the dynamics of recall of SARS-CoV-2 spike-specific immunity, we first analysed immune responses after vaccination of seropositive individuals. We recruited and longitudinally sampled a cohort of 25 individuals with previous PCRconfirmed SARS-CoV-2 infection and/or baseline spike protein seropositivity (seropositive group), and a comparator group of 8 seronegative individuals with no history of SARS-CoV-2 infection (naïve group) (Table S1 ). We undertook early longitudinal sampling from day 3 onwards after vaccination with either BNT162b2 or ChAdOx1 nCoV-19 vaccines. In seropositive individuals, S-and RBD-specific antibody titres began to increase 5 days after vaccination, with titres peaking between day 10-14 ( Fig 1A and B) . In naïve individuals, S and RBD antibody titres emerged later after the first dose (day 9 onwards) and remained at lower levels compared to immunized seropositive individuals, in line with other reports of primary immunization of immunologically naïve individuals (14) . We also assessed serological responses using a live virus neutralization assay (15). At the time of vaccination, only 47% of previously infected individuals had detectable plasma neutralization activity, which reflected residual activity following waning from peak neutralization titres seen in early convalescence ( Fig 1C) . Following the first vaccine dose, neutralizing titres increased from day 6, concomitant with the rise in S and RBD binding antibodies and peaked between day 10-14 ( Fig 1C) . In contrast, immunization of naïve individuals elicited much lower levels of neutralizing antibodies, which only emerged around day 12 postfirst dose. We applied a segmented linear model to estimate the initial period of delay, the rate of increase, and fold change over baseline in the recall of antibodies (Table   S2 ). We estimated that the initial delay phase before neutralizing antibody levels increased was 4.85 days, with a doubling time of 0.74 days thereafter and with a peak fold-change over baseline of 25.8.

Memory B cells constitute an important arm of durable vaccine-elicited immunity, rapidly responding to secondary antigen exposure via differentiation into antibody secreting cells (ASCs). To better understand memory B cell re-activation in vivo, we assessed changes in the frequency and phenotype of SARS-CoV-2 specific memory B cells in seropositive individuals (n=21) in response to immunization.

Antibody-secreting cells (ASCs; CD19 + CD20 lo CD71 + , commonly termed plasmablasts) expanded in peripheral blood, increasing from as early as day 3 based on flow cytometry (estimated by segmented linear modelling to occur as early as day 2.4), peaking between day 7 and 9 before contracting to near baseline levels from day 11 onwards (Fig 1D and E) . S-specific class-switched memory B cells (Spike + IgD -. CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint CD19 + cells), which unlike ASCs constitute a stable population of quiescent memory (4), were detectable in all seropositive individuals prior to vaccination (0.31-1.5% of IgD -B cells). Following vaccination, the frequency of S-specific class-switched B cells increased from day 7 onwards based on flow cytometry (estimated as early as day 6.5) and peaked by day 10 (Fig 1 F and G) . The activation state of S-specific B cells was assessed longitudinally using surface-expressed activation markers CD21 and CD71 (16). Consistent with expansion of S-specific B cells, CD21 downregulation and CD71 upregulation, both denoting cellular activation, were evident as early as day 3

and was maximal around day 9 (Fig 1H-I) .

Given the potential of T cells to contribute to the control of viral replication and the association of CD4 + T cell responses with the development of neutralizing antibodies (17), we assessed the recall of S-specific CD4 + and CD8 + T cells following vaccination of seropositive individuals. Using re-stimulation with recombinant S protein and an activation-induced marker (AIM) assay (Fig S1 and S2) , an increase in both Sspecific CD4 + memory T cells (CD4 + Tmem; CD45RA -CXCR5 -) (Fig 2A and B ) and circulating CD4 + T follicular helper cells (cTFH; CD45RA -CXCR5 + ) was evident from day 5 onwards, peaking around day 9 and declining thereafter (Fig 2C and D) . Recall of S-specific CD4 + T cell responses have also been previously quantified at an epitope-specific level in a subset of the vaccination cohort (n=10 individuals; (18)). Use of an HLA-DRB1*15/S751 tetramer to precisely enumerate antigen-specific T cell frequencies following vaccination provided similar results to the AIM assay, with recall evident from day 5 onward and peaking between days 8 and 10 (Fig 2E and F ). Similar kinetics were observed for S-specific CD8 + memory T cells (Fig 2G and H) , albeit at a lower magnitude than S-specific CD4 + T cell responses. Using the same modelling approach as above, the initial delay for T cell recall was ~4 days (Table S2 ). The peak levels (amongst available samples) of ASCs and S-specific cTFH cells were positively correlated with the peak binding and neutralizing antibodies, as well as with each other is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint infection is rarely known and is often referenced from the time of symptom onset (estimated to be a mean of 4.3 days after acquisition for Delta) (19). Nevertheless, we recruited 6 individuals with generally mild-moderate PCR-confirmed breakthrough SARS-CoV-2 infections that occurred 1-4 months after receiving a second dose of a COVID vaccine (Table S3) , during a time when Delta was the dominant circulating variant. Serial blood samples and nose swabs were obtained over 1-18 days after symptom onset and S-specific antibody and cellular immune responses analysed as before. An additional follow-up sample was available for four donors between day 26-39. For four individuals, the time of infection could be definitively established as 2-3 days prior to symptom onset as these subjects had a single defined exposure event in a low incidence environment. Neutralizing antibody titres, as well as S-and RBDbinding antibodies, remained at baseline levels for a remarkable 5-7 days after symptom onset (8-10 days post exposure) before rising steadily during the second and third weeks up to the last time point collected (Fig 3A-C) . Activation of S-specific memory B cells was evident from day 6 post symptom onset (Fig S3A) . The circulating frequencies of ASCs and S-specific memory B cells remained stable for the first week after symptom onset, before expanding around day 7-8 post symptom onset (Fig 3C   and D) . With the exception of subjects #5 and #6, whose precise exposure time was undefined, the recall kinetics of antibody and memory B cells following breakthrough infection appeared delayed when compared to vaccination of seropositive individuals.

Surprisingly, frequencies of S-specific CD4 + and CD8 + T cells remained largely unchanged following breakthrough infection in 5 of 6 individuals, with only a single subject (#6) displaying a >5-fold increase (Fig 3E-G) . These results are in stark contrast to the rapid and clear recall of CD4 + and CD8 + T cell immunity following vaccination of seropositive subjects shown in Fig 2. To compare the dynamics of immune recall following vaccination and breakthrough infection, we applied the same model to parameterize the kinetics of recall. We considered the delay from exposure to symptom onset (which was known in 4/6 subjects and conservatively assumed this to also be 3 days prior to symptom onset in the other 2 subjects). The estimated time to initial increase in neutralizing and binding antibody levels was 7.22 and 8.21 days after exposure respectively, and 9.3 days for S-specific B cells (Table S4) , all of which were delayed compared to vaccination (increase estimated prior to day 7 for all three measures, Fig 4A) . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint To investigate the relationship between recall immunity and viral control, we analysed viral load kinetics by qPCR of N (Fig 4B) , RDRP and S (Fig S4B-E) genes in serial nasopharyngeal swabs from 4 of the individuals with breakthrough infection.

This indicated a peak of viral replication (amongst available timepoints) on day 7-8 after infection (day 4-5 after symptom onset), followed by rapid viral clearance thereafter. Neutralizing and binding antibody titres were negatively correlated with viral loads (Fig 4C and D, Fig S4C-G) . Comparison of viral load kinetics with the recall of antibodies, indicated that the peak of viral load preceded the rise in neutralizing antibodies and that recall of humoral immunity coincided with a decrease in viral load ( Fig 4E) . Where the timing of exposure was defined, the delay between infection and antibody recall was ~7-8 days (including 2-3 prior and 5-6 days after symptom onset). The longer delay in recall of immunity following infection compared to vaccination may reflect differences in antigen accumulation between respiratory acquisition of infection is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint Encouragingly, we find breakthrough infection of vaccinated individuals drives re-expansion of humoral immune memory with augmented neutralizing antibodies, albeit with some delay. This suggests that recall of immunity may mitigate disease severity of breakthrough infections with antigenically distinct variants including Omicron, while also boosting population level immunity against SARS-CoV-2, potentially further restricting the healthcare burden inflicted by the pandemic and smoothing a pathway towards endemicity.

We thank the participants for the generous involvement and provision of samples. We is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint

The authors declare no competing interests.

All data are available in the main text or the supplementary materials.

. CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint

The study protocols were approved by the University of Melbourne Human Research Ethics Committee (2021-21198-15398-3, 2056689) , and all associated procedures were carried out in accordance with approved guidelines. All participants provided written informed consent in accordance with the Declaration of Helsinki.

A cohort of subjects with either a prior positive nasal PCR for SARS-CoV-2 or a positive ELISA for SARS-CoV-2 S and RBD protein were recruited to provide blood samples following vaccination against SARS-CoV-2. Contemporaneous controls who had not previously experienced any symptoms of COVID-19 and who were confirmed to be seronegative were also recruited to provide blood samples prior to and following vaccination for SARS-CoV-2 (Table S1) . A cohort of previously vaccinated participants with a nasal PCR-confirmed breakthrough COVID-19 were recruited through contacts with the investigators and invited to provide serial blood samples (Table S1 ). For all participants, whole blood was collected with sodium heparin anticoagulant. Plasma was collected and stored at −80 °C, and PBMCs were isolated via Ficoll-Paque separation, cryopreserved in 10% DMSO/FCS and stored in liquid nitrogen.

Antibody binding to SARS-CoV-2 S or RBD proteins was tested by ELISA. The expression of recombinant S and RBD has been described previously (29) . For ELISA, 96-well Maxisorp plates (Thermo Fisher) were coated overnight at 4°C with 2 μg/ml recombinant S or RBD proteins. After blocking with 1% FCS in phosphatebuffered saline (PBS), duplicate wells of serially diluted plasma were added and incubated for 2 h at room temperature. Plates were washed in PBS-T (0.05% Tween-20 in PBS) and PBS before incubation with 1:20,000 dilution of HRP-conjugated antihuman IgG (Sigma) for 1 h at room temperature. Plates were washed and developed using TMB substrate (Sigma), stopped using sulphuric acid and read at 450 nm. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint Endpoint titers were calculated as the reciprocal serum dilution giving signal 2× background using a fitted curve (4 parameter log regression).

Plasma neutralization activity against SARS-CoV-2 was measured using a microneutralization assay as previously described (15). Wildtype SARS-CoV-2 (CoV/Australia/VIC/01/2020) isolate was passaged in Vero cells and stored at -80ºC. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. 

Analysis of SARS-CoV-2 specific T cells was performed as previously described (29). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. 

Three microlitres of cDNA was added to a commercial real-time PCR master mix Thermal cycling and rRT-PCR analyses for all assays were performed on the ABI 7500 FAST real-time PCR system (Applied Biosystems, USA) with the following thermal cycling profile: 95C for 2 min, followed by 45 PCR cycles of 95C for 5 s and 60C for 25 s for N gene and 95C for 2 min, followed by 45 PCR cycles of 95C for 5 s and 55C for 25 s for RdRP/Helicase gene and S gene.

We used a segmented model to estimate the activation time and growth rate of various immune responses after vaccination and breakthrough infection. The model of the immune response for subject at time can be written as:

. CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ;

( ) = { ( + ); ≥ 1 + 1 ( + ) ( + )( −( 1 + 1 )) ; 1 + 1 ≤ < 2 + 2 ( + ) ( + )(( 2 + 2 )−( 1 + 1 )) × −( + )( −( 2 + 2 )) ; ≥ 2 + 2 .

The model has 5 parameters; , , 1 , , and 2 . For a period before 1 , we assumed a constant baseline value for the immune response. After the activation time 1 , the immune response will grow at a rate of until 2 . From 2 , the immune response will decay at a rate of . For each subject , the parameters were taken from a normal distribution, with each parameter having its own mean (fixed effect). A diagonal random effect structure was used, where we assumed there was no correlation within the random effects. The model was fitted to the log-transformed data values, with a constant error model distributed around zero with a standard deviation . To account for the values less than the limit of detection, a censored mixed effect regression was used to fit the model. Values less 20, 100, and 0.0001 were censored for the neutralization, IgG bindings, and T cell data respectively. Model fitting was performed using MonolixR2019b. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 23, 2021. ; https://doi.org/10.1101/2021.12.23.21268285 doi: medRxiv preprint coefficients (rs) and p values are indicated on the figure, n=25 paired samples from 4 subjects.

. CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

Effectiveness of BNT162b2 and mRNA-1273 covid-19 vaccines against symptomatic SARS-CoV-2 infection and severe covid-19 outcomes in Ontario, Canada: test negative design study

Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection

Immune correlates analysis of the mRNA-1273 COVID-19 vaccine efficacy clinical trial