Intravesical BCG in bladder cancer induces innate immune responses against SARS-CoV-2

Abstract

BCG is the most efficient adjuvant therapy for high-risk, non-muscle-invasive bladder cancer (NMIBC). Both innate and adaptive immune responses have been implicated in BCG-mediated effects. BCG vaccination can boost innate immune responses via trained immunity (TI), resulting in an increased resistance to respiratory viral infections. Here we evaluated for the first time whether intravesical application of BCG triggers increased immunity against SARS-CoV-2 in patients with high-risk NMIBC. Serum and peripheral blood mononuclear cells (PBMCs) from heparinized whole blood samples of 11 unvaccinated SARS-CoV-2-naïve high-risk NMIBC patients were collected at baseline and during BCG treatment in a pre-COVID-19 era. To examine B-cell or T cell-dependent adaptive immunity against SARS-CoV-2, sera were tested for the presence of SARS-CoV-2 neutralizing antibodies. Using a SARS-CoV-2 peptide pool, virus-specific T cells were quantified via IFNγ ELISpot assays. To analyze innate immune responses, mRNA and protein expression levels of pro- and anti-inflammatory cytokines were measured after a 24-hour stimulation of PBMCs with either BCG or SARS-CoV-2 wildtype. ATAC- sequencing was performed to identify a potential epigenetic reprogramming in immune cells. We neither identified SARS-CoV-2 neutralizing antibodies nor SARS-CoV-2- reactive T cells, indicating that intravesical BCG did not induce adaptive immunity against SARS-CoV-2. However, a significant increase in mRNA as well as protein expression of IL-1β, IL-6 and TNFα, which are key cytokines of trained immunity, could be observed after at least four intravesical BCG instillations. Genomic regions in the proximity of TI genes (TLR2, IGF1R, AKT1, MTOR, MAPK14, HSP90AA1) were more accessible during BCG compared to baseline. Although intravesical BCG did not induce adaptive immune responses, repetitive intravesical instillations of BCG induced circulating innate immune cells that produce TI cytokines also in response to SARS-CoV-2.

Keywords: BCG; COVID-19; SARS-CoV-2; bladder cancer; trained immunity; viral infections.

Figure 1

Flow chart of the study. In this study we used blood samples that have been collected from (very) high-risk NMIBC patients undergoing intravesical BCG therapy in the pre- COVID-19 era (2014-2015) using four time points (baseline, 1-2 weeks, 3-4 weeks and 6-12 weeks during BCG).

Figure 2

The graph shows antibody titers against wildtype SARS-CoV-2 spike RBD in BAU/ml (binding antibody units per ml) from each patient and time interval (black dots). As comparison, antibody titers from SARS-CoV-2 recovered individuals 1 month post infection are shown (green dots, n=24). Fraction of positive titers are indicated as percentages above each group. Titers >7.1 BAU/ml were defined as positive according to the manufacturer’s instructions. Time points: 1-2 weeks/early, 3-4 weeks/mid and 6-12 week-interval/late during BCG. Data is presented as geometric mean (horizontal line) with 95% confidence interval (whiskers).

Figure 3

Virus neutralization assays and SARS-CoV-2/BCG specific IFNγ T cell responses. (A) Serum neutralization against SARS-CoV-2 wildtype in eight BCG-treated high-risk NMIBC patients (baseline and during BCG). (B) Serum neutralization in the internal control group (COVID-19 unvaccinated and 3x vaccinated patients, n=8). (C) IFNγ reactive T cells 24h post BCG/SARS-CoV-2 peptide pool (n=6). PBMCs (0.5x10⁵) were treated with CEF/CEFTA as positive control. Data represent mean with SD; *p <.05, **p <.01; ***p <.001.

Figure 4

Relative quantification of cytokine mRNA expression by multiplex qPCR. (A) Pro- inflammatory cytokines IL-1β, IL-6 and TNFα, (B) cytokines associated with Th1 response IL- 12A, IL-18 and IFNγ and (C) anti-inflammatory cytokine IL-10 after a 24h stimulation of PBMCs with either BCG or SARS-CoV-2 wildtype are shown. PBMCs were taken before the start of BCG induction (baseline) and at early, mid and late time intervals during BCG treatment. Data is presented as fold-change expression relative to untreated control group (mean with SEM and mean values annotated above each individual bar; *p <.05, **p <.01; ***p <.001). Time points: baseline (before BCG induction), 1-2 weeks/early, 3-4 weeks/mid and 6-12 week-interval/late during BCG.

Figure 5

Cytokine quantification using Luminex xMAP technology. (A) Pro-inflammatory cytokines IL-1β, IL-6 and TNFα, (B) cytokines associated with Th1 response IL-18 and IFNγ and (C) anti-inflammatory cytokine IL-10 were measured in cell culture supernatant of PMBCs after 24h stimulation with either BCG or SARS-CoV-2 wildtype are presented. PBMCs were prepared before the start of BCG induction (baseline) and at early, mid and late time intervals during BCG treatment. Data represent means with SEM and mean values annotated above each individual bar (*p <.05, **p <.01; ***p <.001). Time points: baseline (before BCG induction), 1-2 weeks/early, 3-4 weeks/mid and 6-12 week-interval/late during BCG. IL-12A data are not shown as their levels were not detectable and samples for repetition were no more available.

Figure 6

IL-1β, IL-6 and TNFα mRNA expression after 24h of LPS (100 ng/ml) treated PBMCs from patients before and during various time-points of BCG treatment. Data is presented as foldchange expression relative to untreated control group (mean with SEM). Time points: 1-2 weeks/early, 3-4 weeks/mid and 6-12 week-interval/late during BCG.

Figure 7

ATAC-seq analysis. (A) Venn diagram showing 6 genes (AKT1, HSP90AA1, IGF1R, MAPK14, MTOR, TLR2) overlapping between 64 trained immunity genes and 1697 genes with differentially accessible regions in their promoter or introns. (B) Differential affected immune system related processes (during BCG versus baseline) using ClueGO (C) Representative sequencing tracks for the TLR2 locus showing significantly higher ATAC-seq peaks at the promoter during BCG compared with baseline. Modifications of H3K4me3 were shown at the same region near the transcription start site of TLR2.

Figure 8

Schematic overview of BCG-induced trained immunity (TI). Innate immune cells such as monocytes, NK cells, dendritic cells, neutrophils, macrophages and their progenitor cells can be trained through repetitive BCG administration, which induces epigenetic and metabolic reprogramming of these cells. BCG is binding to the toll-like receptors (TLR)2 and TLR4 and thus leads to downstream epigenetic modifications at H3K4me3 through both the Akt/mTOR and NOD2 pathways. As a consequence, this specific modification of H3K4me3 facilitates the enhanced production of IL-1β, IL-6 and TNFα, known as the key cytokines of trained immunity (–28). During a secondary bacterial or viral stimulation trained innate immune cells display an increased capacity to produce pro-inflammatory cytokines, resulting in improved protection also against unrelated pathogens. Importantly, in our study intravesical instillations as a route of BCG administration were suitable to induce TI, reflected by increased production of TI cytokines after at least four BCG instillations at mRNA and protein levels. BCG administration also resulted in epigenetic modifications of TI-related metabolic genes and of the gene encoding TLR2, which is a specific receptor for BCG. Most importantly, we demonstrated for the first time that high-risk NMIBC patients undergoing intravesical BCG showed an increased SARS-CoV-2 innate immune responsiveness indicative of TI.