MAIT and other innate-like T cells integrate adaptive immune responses to modulate interval-dependent reactogenicity to mRNA vaccines [MAIT]

Adenoviral (Ad) vectors and mRNA vaccines exhibit distinct patterns of immune responses, with reactogenicity influencing their use during and beyond the COVID-19 pandemic. However, despite the clinical significance, the underpinning mechanisms remain unclear. Innate and innate-like lymphocytes, such as mucosal-associated invariant T cells, are highly sensitive to cytokines and can enhance both innate and adaptive vaccine responses. We compared longitudinal human immune responses and reactogenicity following homologous ChAdOx1 nCoV-19 and BNT162b2 vaccination to define the interactions between innate-like lymphocytes and adaptive immunity – and their in vivo consequences. Specifically, Ad vector priming elicited robust early type I interferon (IFN)-mediated activation of innate-like T cells, augmenting T cell responses (innate to adaptive signalling), which diminished upon boosting in the presence of anti-vector immunity. In contrast, mRNA vaccine responses in innate-like cells were markedly enhanced after boosting. This was initiated by IFN-? signalling from spike-specific memory T cells and amplified by IFN-?R+ innate-like lymphocyte networks (adaptive to innate signalling). Importantly, extending the interval between doses reduced inflammatory responses to mRNA vaccination. In an independent clinical trial, spike-specific T cells predicted severe reactogenicity to mRNA vaccine boosting regardless of the dosing interval and vaccine type. These findings reveal a close integration of innate-like and adaptive responses to novel vaccines, including an IFN-?-mediated function of innate-like T cells in orchestrating critical early responses to mRNA vaccines which may have significant implications for optimising future vaccine regimens. Overall design: To investigate longitudinal early immune responses to SARS-CoV-2 vaccines, we recruited 56 healthy adult volunteers and healthcare workers (HCW) prior to their initial homologous prime-boost vaccination with BNT162b2 or ChAdOx1 nCoV-19 (also known as AZD1222; hereafter referred to as ChAdOx1-S). Samples were collected at early time points following vaccination. The vaccination regimens included: BNT162b2 with a short-interval boost (n = 17; median interval, 21 days; IQR, 21-24; range, 17-30), BNT162b2 with a long-interval boost (n = 19; median interval, 70 days; IQR, 62-78; range, 42-120), and ChAdOx1-S with a long-interval boost (n = 20; median interval, 76 days; IQR, 70-79; range, 60-98). The cohort had a female bias (64%), with a median age of 35 years (IQR, 30-43; range 21-65), and a median body mass index (BMI) of 23.4 kg/m2 (IQR, 20.3-26.0). Some participants had a prior history of SARS-CoV-2 infection, evidenced by detectable anti-spike IgG antibodies at baseline. Prior infection rates were balanced among the vaccine groups, and unless otherwise stated, analyses focused on infection-naïve participants. Peripheral blood samples were collected in EDTA, Tempus RNA, and serum separator tubes (SST) at baseline (V1), one day post-prime (V1+1), pre-boost (V2), and one day post-boost (V2+1). Adaptive immune responses were assessed four weeks post-prime (V1+28), eight weeks post-prime (V1+70) for the long-interval groups, and four weeks post-boost (V2+28), measuring SARS-CoV-2 specific binding antibodies, neutralising antibodies, and spike-specific T cell IFN-? ELISpot responses. Due to differing dosing intervals, V1+28 corresponds to V2 for the short-interval BNT162b2 group, while V1+70 corresponds to V2 for the long-interval BNT162b2 and ChAdOx1-S groups. EDTA samples were processed to obtain plasma and peripheral blood mononuclear cells (PBMCs). Fresh PBMCs were used to phenotype cellular activation, while frozen PBMCs were used for ELISpot assays. RNA-sequencing (RNA-seq) was performed on samples collected in Tempus RNA tubes or from PBMCs, as appropriate. Sorting of human mucosal-associated invariant T (MAIT) cells: Thawed PBMCs (2 × 106/sample) were stained with MR1 tetramers (40 min, room temperature), washed twice, then incubated with surface antibodies in FACS buffer (20 min, 4°C). Dead cells were stained using SYTOX Green (Thermo Fisher Scientific; 1:6000 dilution). Viable MAIT cells (CD3+MR1/5-OP-RU+Va7.2-TCR+) resuspended in PBS + 0.05% BSA were sorted on a BD FACSAria III (BD Biosciences) using an 85 µm nozzle directly into 1 ml of TRIzol (Ambion, #15596026), immediately snap frozen on dry ice, and stored at -80°C until RNA extraction.