The binding of Spike/RBD proteins of coronaviruses to receptors triggers MC degranulation
MCs express a diversity of membrane receptors that are capable of recognizing and responding to a multitude of stimuli (
20). To elucidate the role of viral receptors in MCs, we first examined their expression profiles. We have recently found the expression of ACE2 on the LAD2 mast cell line, which facilitates the productive infection of SARS-CoV-2 (
14). In this study, we extended the investigation to two additional human mast cell lines, namely LUVA and HMC-1. The analysis with flow cytometry exhibited that these cell lines expressed high levels of aminopeptidase N (APN) and ACE2, with a relatively lower expression of dipeptidyl peptidase 4 (DPP4) (
Fig. 1A ;
Fig. S1). Notably, the LAD2, LUVA, and HMC-1 mast cell lines were found to be susceptible to infection by human coronavirus 229E (HCoV-229E), which utilizes APN as its receptor, and human coronavirus NL63 (HCoV-NL63), which utilizes ACE2 as its receptor (
Fig. 1B).
We have previously demonstrated that the binding of the Spike/RBD protein of SARS-CoV-2 to ACE2 receptor could trigger rapid MC degranulation (
14). To further validate this observation and investigate the ubiquity for coronavirus-induced MC degranulation, we expanded the investigation to include recombinant Spike/RBD proteins sourced from multiple human coronaviruses. The binding of Spike/RBD proteins to cells was first investigated. Specifically, HMC-1 and LUVA cells were exposed to recombinant Spike/RBD proteins derived from HCoV-229E, HCoV-NL63, HCoV-OC43, HCoV-HKU1, Middle East respiratory syndrome coronavirus (MERS-CoV), SARS-CoV, and SARS-CoV-2 for 1 h at 4°C. The recombinant proteins were identified through immunostaining with a His-Tag antibody, and the receptors of APN, ACE2, and DPP4 were indicated with immunostaining with specific antibodies. Confocal microscopy was employed to observe the co-localization of these proteins with their respective receptors (
Fig. 1C). The binding interactions of Spike/RBD proteins derived from human coronaviruses OC43 and HKU1, which recognize O-acetylated sialic acid (O-ac Sia) as their receptor, were visualized through the overlay of bright-field images with immunostaining of the His-tag (
Fig. 1C).
MC degranulation was detected by measuring the secretion of granule content β-hexosaminidase (
21). The compound 48/80 (C48/80) that can stimulate MC degranulation was used as the control. The treatment with human coronavirus Spike/RBD proteins induced degranulation in both HMC-1 and LUVA mast cell lines (
Fig. 1D). We have previously demonstrated that LAD2 cells underwent rapid degranulation upon stimulation with Spike/RBD proteins or authentic viruses of SARS-CoV-2, HCoV-NL63, and HCoV-229E (
14). Here, we extended this observation by examining the degranulation response of LAD2 cells to a broader range of Spike/RBD proteins derived from SARS-CoV-2 and its diverse mutants, including S-R, beta, delta, and omicron variants. Additionally, we included Spike proteins from other coronaviruses, including SARS-CoV, BANAL-20-52, MERS-CoV, and MERS-CoV-2P (
Fig. 1D). Furthermore, we used the authentic viruses HCoV-NL63 and HCoV-229E to validate the triggered degranulation in HMC-1, LUVA, and LAD2 cells (
Fig. 1E), reinforcing the finding that viral Spike proteins play a crucial role in initiating this process.
To ascertain whether the mere binding of Spike proteins to receptors on MCs is sufficient to trigger degranulation (
14), these spike proteins derived from distinct coronaviruses that have the divergent binding affinity with their receptors were used. Specifically, modifications were introduced to the Spike protein of SARS-CoV-2, referred to as S-R, by covalently inserting a disulfide bond between residues 413 and 987. This engineered mutant, designated as S-R/x2, is stabilized in a prefusion conformation, which significantly attenuates its binding affinity for ACE2 (
22,
23). The bat coronavirus BANAL-20-52 and the MERS-CoV-2P strain (specifically the England1 variant) showed high affinity binding to their respective dimeric receptors, ACE2 and DPP4 (
24,
25). In contrast, the bat coronavirus BM48-31-CoV (GenBank
ON131096) and VsCoV-1 exhibited limited binding ability to their receptors (
26–28). Expectedly, the treatments with Spike/RBD proteins derived from S-R, BANAL-20-52, and MERS-CoV-2P, which maintain a high receptor binding affinity, resulted in the induction of MC degranulation (
Fig. 1F). Conversely, Spike/RBD proteins isolated from S-R/x2, BM48-31, and VsCoV-1, which exhibited limited receptor binding capability, failed to elicit such an induction (
Fig. 1F). Additionally, the S2 subunit from SARS-CoV-2 Spike protein was used as the control. These pre- and post-fusion S2 subunits do not contain the RBD domain and did not induce degranulation of LUVA and HMC-1 cells (
Fig. 1G).
Collectively, these findings underscore the capacity and ubiquity of coronavirus Spike/RBD proteins to provoke MC degranulation and implicate that the interaction between Spike proteins and receptors is the pivotal trigger of this process.
Transcriptome analysis reveals Spike/RBD protein-induced MC activation
To comprehensively elucidate the activation of MCs in response to stimuli with Spike/RBD proteins, we conducted a transcriptome analysis. LAD2 cells were exposed to Spike/RBD proteins derived from MERS-CoV, SARS-CoV, and SARS-CoV-2, as well as virions of HCoV-229E and HCoV-NL63, respectively, for 24 h. Transcriptome profiles of the LAD2 cells were analyzed using standardized protocols. Data from three independent repeats were summarized.
The analysis with the volcano plot revealed that treatment with Spike/RBD proteins led to the downregulation of a relatively limited number of genes, whereas it induced the upregulation of hundreds of genes (
Fig. 2A). In contrast, the treatments with virions of HCoV-229E and HCoV-NL63 showed a profound effect on transcripts, causing thousands of differently expressed genes (DEGs) (
Fig. 2A). A comprehensive analysis was conducted to summarize the genes that underwent consistent alterations in response to treatments involving Spike/RBD proteins or virions. A total of 111 genes were observed to be upregulated, whereas only a single gene, CLEC12A (C-type lectin domain family 12 member A), exhibited downregulation (
Fig. 2B) (
Fig. S2,
Table S2).
To further elucidate the functional implications of these upregulated genes, a gene ontology (GO) functional enrichment analysis was performed. The genes that underwent consistent alterations in response to treatments involving Spike/RBD proteins (from MERS-CoV, SARS-CoV, and SARS-CoV-2) and virions (HCoV-229E and HCoV-NL63) were used for analysis. This analysis demonstrated an enrichment of gene sets that regulate immune response, cytokine production, and inflammatory responses among the upregulated genes (
Fig. 2C;
Fig. S2). The most prominently upregulated genes included CCL2, CCL3, CCL20, CCL3L1, CCL4, CXCL1, CXCL11, CXCL8, IL1β, IL18R1, IL32, IRAK2, IRF7, TNFAIP6, NFRSF9, and TNFSF15, etc. (
Tables S2 and S3 ).
To validate the transcriptome data and quantify the upregulation of inflammatory factors, we performed real-time (RT-) PCR analysis. The results confirmed the upregulation of IL-1β, IL-6, and TNF-α in virus-infected LAD2 cells (
Fig. 2D) or Spike/RBD protein-treated LUVA cells (
Fig. 2E). Notably, when LUVA cells were treated with Spike/RBD proteins exhibiting limited receptor binding affinity, the induction of IL-6, TNF-α, and IL-8 levels was compromised (
Fig. 2F), further underscoring the necessity of Spike protein binding to receptors for eliciting MC activation. In summary, these transcriptome data reveal Spike/RBD proteins-induced MC activation.
Spike/RBD protein induces activation of the cellular Src/PI3K/AKT signaling pathway
In the subsequent analysis, we delved into the cellular pathways that are activated in response to Spike/RBD protein-induced MC degranulation. Despite the expressions of numerous membrane receptors that mediate MC activation upon stimulation, the most extensively studied mechanism underlying MC activation pertains to IgE-mediated FcεRI activation in type I hypersensitivity allergic reactions (
1,
8). Upon binding of IgE to FcεRI, SFK member Lyn becomes engaged and undergoes phosphorylation for activation. This activation subsequently triggers a series of cytoplasmic cascades, ultimately culminating in the release and new synthesis of mediators within the cell (
1,
8,
19).
Similarly, the treatment with Spike/RBD proteins was found to activate the cellular Src/PI3K/AKT signaling pathway. The treatments with Spike/RBD proteins elicited auto-phosphorylation of endogenous Src kinase at the tyrosine residue 416 (Tyr-416) (
Fig. 3A). Further analysis of downstream cellular signaling cascades revealed an augmented phosphorylation of the p85 subunit of PI3K kinase, as well as Akt (
Fig. 3A). Notably, when the PI3K inhibitor LY294002 was incorporated into the treatment alongside Spike/RBD proteins, the resultant degranulation of LUVA was completely abrogated (
Fig. 3B). The LY294002 could also effectively block the degranulation of LUVA and HMC-1 cells induced by HCoV-229E and HCoV-NL63 virions (
Fig. 3C). These findings underscore the necessity of the intracellular Src/PI3K/Akt signaling pathway for Spike protein-induced MC degranulation.
Reducing the cytoplasmic Ca2+ or blocking the microtubule reorganization abolishes Spike protein-triggered MC degranulation
In the downstream signaling cascade triggered by FcεRI, both Ca²
+-dependent and -independent pathways can mediate the reorganization of microtubules and actin filaments, which are essential for the transport of secretory granules and the subsequent membrane fusion (
1,
8,
19). To investigate whether Spike protein-triggered MC degranulation is Ca
2+-dependent, the F03 intracellular calcium ion fluorescent probe was added during the treatment of MCs with Spike/RBD proteins. An obvious increase in cytoplasmic Ca²
+ concentration was observed after the treatments with spike/RBD proteins in LUVA cells (
Fig. 4A). The rapid increases of cytoplasmic Ca
2+ along with treatments were observed in both LUVA and LAD2 cells (
Fig. 4B).
To investigate the Ca²
+ dependency of this process, the Ca²
+ chelator BAPTA-AM was used. The Ca²
+ ion could be effectively chelated by BAPTA-AM (
Fig. 4B). Subsequently, the induction of degranulation in both LUVA and LAD2 cells was abolished in the presence of BAPTA-AM (
Fig. 4C), suggesting a crucial role of cytoplasmic Ca²
+ in this process. The addition of nocodazole, a compound known to block microtubule reorganization, significantly diminished the triggered degranulation in both LUVA and LAD2 cells (
Fig. 4D), demonstrating the dependence of microtubule reorganization for this process. Collectively, these data provide evidence that Spike protein-induced MC degranulation is dependent on both the elevation of cytoplasmic Ca²
+ and the sequential reorganization of microtubules.