SARS-CoV-2 N protein induces proinflammatory responses in cells and mice tissues
SARS-CoV-2 infection induces hyperinflammatory syndromes characterized by overexpression of pro-inflammatory factors5,24,25. Here, we initially analyzed and compared the expression statuses of cytokines and chemokines in mock-infected and SARS-CoV-2-infected macrophages and dendritic cells (DCs) based on the data set GSE155106 from the GEO database. The data showed that expression of proinflammatory cytokines (IL-1β, IL-6, TNF-α, IL-11, IL-10, and IL-27) and chemokines (CXCL8, CCL2, CXCL10, and IL-7) were significantly induced upon SARS-CoV-2 infection in macrophages (Fig. 1a) and DCs (Fig. 1b). As inflammasome-mediated mature IL-1β induces the release of cytoplasmic proteins and factors and plays an important role in initiating “cytokine storm”19, and thus, we evaluated the effect of SARS-CoV-2 proteins on the production of mature IL-1β. SARS-CoV-2 genes (N, M, E, 3a, 6, 7a, 8, and 10) encoding for three structural proteins (N, M, and E) and five accessory proteins (ORF3a, ORF6, ORF7a, ORF8, and ORF10) were synthesized based on the sequences of SARS-CoV-2 (GenBank accession number MN908947.3). The synthesized genes were cloned to pcDNA3.1(+) expression vector with HA-tag or Flag-tag. We showed that SARS-CoV-2 N, M, E, ORF3a, ORF6, ORF7a, and ORF8, except ORF10, were expressed in human embryonic kidney 293T (HEK293T) cells transfected with plasmids encoding the proteins (Fig. 1c).
The roles of SARS-CoV-2 proteins in regulating the secretion of mature IL-1β were assessed in a reconstructed NLRP3 inflammasome system as described previously15. HEK293T cells were co-transfected with plasmids encoding the three components (NLRP3, ASC, and pro-Casp-1) of the NLRP3 inflammasome and the substrate pro-IL-1β to generate a NLRP3 inflammasome system. The HEK293T-NLRP3 inflammasome system cells were subsequently transfected with plasmids encoding N, M, E, 3a, 6, 7a, 8, and 10, respectively. ELISA results showed that IL-1β secretion was significantly induced by N protein in this system (Fig. 1d). In addition, human monocytic cell lines (THP-1) were differentiated into macrophages. PMA-differentiated THP-1 macrophages were then transfected with plasmids encoding N, M, E, 3a, 6, 7a, 8, and 10, and stimulated with Nigericin (a specific activator for NLRP3 inflammasome). We noted that the secretion of mature IL-1β was significantly induced by Nigericin, and the Nigericin-induced mature IL-1β secretion was further remarkably enhanced by the N protein (Fig. 1e). Collectively, these results demonstrated that SARS-CoV-2 N protein plays an important role in the induction of mature IL-1β secretion.
Next, the effect of N protein on the regulation of proinflammatory cytokines and chemokines was assessed in PMA-differentiated THP-1 macrophages. THP-1 cells were infected with N-Lentivirus (Lentivirus carrying the N gene) and CT-Lentivirus (as a negative control) to generate two cell lines stable expressing N-Lentivirus and CT-Lentivirus, respectively (Fig. 1f). The stable cells were subsequently differentiated into macrophages. qRT-PCR analyses showed that the expression of IL-1β, IL-6, TNF-α, CXCL10, IL-11, IL-13, CCL2, and CXCL8 mRNAs was notably elevated by Lentivirus-N, whereas the expression of IL-18, IL-10, and VEGFA mRNAs was not affected by Lentivirus-N (Fig. 1g).
Moreover, the role of N protein in the regulation of proinflammatory cytokines and chemokines was evaluated in mice. C57BL/6 mice were tail vein injection with adenovirus associated virus (AAV)-Lung-EGFP (control), AAV-Lung-N, and AAV-Lung-N plus Ac-YVAD-cmk (a selective irreversible inhibitor of Caspase-1 that blocks IL-1β maturation), respectively. We noted that N protein was expressed in the lung of mice injected with AAV-Lung-N, but not detected in the liver or spleen of mice injected with AAV-Lung-N (Fig. 1h). qRT-PCR analyses showed that IL-1β, IL-6, TNF-α, CXLC10, CCL2, IL-11, IL-7, and CXCL8 mRNAs were notably induced in the lung of mice injected with AAV-Lung-N, but relatively unaffected in the lung of mice injected with AAV-Lung-CT (Fig. 1i). Interestingly, N-induced expression of IL-1β, IL-6, TNF-α, CXLC10, CCL2, IL-11, IL-7, and CXCL8 were significantly suppressed by Ac-YVAD-cmk (Fig. 1i). Taken together, these results demonstrated that SARS-CoV-2 N protein plays a critical role in the induction of proinflammatory cytokines, and suggested that N protein may involve in regulating the activation of the NLRP3 inflammasome.
IL-1β maturation and IL-6 production are induced by SARS-CoV-2 N protein
Given the dominant effect of the inhibitor on the expression of the pro-inflammatory factors, we further determined the role of SARS-CoV-2 N protein in IL-1β maturation. HEK293T-NLRP3 inflammasome system cells were transfected with plasmid encoding SARS-CoV-2 N protein at different concentrations. The results showed that IL-1β secretion (Fig. 2a) and IL-1β (p17) cleavage as well as Casp-1 (p20) maturation (Fig. 2b) were induced by N protein in dose-dependent fashions.
Next, PMA-differentiated THP-1 macrophages were transfected with plasmids encoding SARS-CoV-2 N at different concentrations and treated with Nigericin. We noted that IL-1β secretion was induced by Nigericin and the Nigericin-induced IL-1β secretion was further promoted by N protein in a dose-dependent manner (Fig. 2c). Similarly, IL-1β (p17) cleavage and Casp-1 (p20) maturation were induced by Nigericin and Nigericin-induced IL-1β (p17) cleavage and Casp-1 (p20) maturation were further facilitated by N protein in concentration-dependent fashions (Fig. 2d). However, we found that the level of mature IL-18 was relatively unaffected by N protein (Fig. 2e), although its maturation and production relied on the same cleavage mechanism as IL-1β.
In addition, THP-1 cells stably infected with CT-Lentivirus and N-Lentivirus were differentiated into macrophages, which were then treated with LPS plus ATP or LPS plus Nigericin. The results indicated that IL-1β secretion was induced by LPS plus ATP or LPS plus Nigericin, and such inductions were further enhanced significantly by N protein (Fig. 2f). Consistently, IL-1β (p17) cleavage was also induced by LPS plus ATP or LPS plus Nigericin, and such inductions were further enhanced by N protein (Fig. 2g). We noted that the level of mature IL-18 was induced by LPS plus ATP or LPS plus Nigericin, but such inductions were relatively unaffected by N protein (Fig. 2h). Notably, the production of widely described proinflammatory cytokine IL-6 was induced by LPS plus ATP or LPS plus Nigericin, and such induction was also further enhanced by N protein (Fig. 2i). Moreover, IL-6 production was confirmed to be induced by N protein in a dose-dependent manner (Fig. 2j).
Finally, granulocyte macrophage-colony stimulating factor (GM-CSF) differentiated mice bone marrow-derived monocytes (BMDMs) were infected with CT-Lentivirus or N-Lentivirus and stimulated with LPS, LPS plus ATP, or LPS plus Nigericin. In the treated mouse BMDMs, IL-1β secretion was activated by LPS plus ATP or LPS plus Nigericin, and such activations were further facilitated notably by N protein (Fig. 2k). Similarly, IL-6 production was also induced by LPS plus ATP or LPS plus Nigericin, and such inductions were further promoted by N protein in treated mice BMDMs (Fig. 2l). Therefore, our results demonstrated that SARS-CoV-2 N protein induces IL-1β maturation and IL-6 production and suggested that N protein may play an important role in the activating the NLRP3 inflammasome.
SARS-CoV-2 N protein interacts with NLRP3 protein
The role of SARS-CoV-2 N protein in the regulation of NLRP3 inflammasome was then investigated. Initially, the interaction of N protein with the NLRP3 inflammasome components, NLRP3, ASC, and Caspase-1, was explored. Co-immunoprecipitation (Co-IP) assays showed that N protein only interacted with NLRP3 protein and failed to interact with ASC protein or Casp-1 protein in HEK293T and A549 cells (Fig. 3a, b). Reciprocal Co-IP assays further conformed that NLRP3 interacted with N in HEK293T and A549 cells (Fig. 3c, d). Importantly, we demonstrated that purified His-N protein could directly bind to NLRP3 protein (Fig. 3e). Moreover, confocal microscope analyses showed that in HEK293T and A549 cells, N protein and NLRP3 protein were co-localized in the cytoplasm (Fig. 3f, g), and N protein failed to interact with ASC (Fig. 3h, i). Taken together, these results demonstrated that N protein interacts with NLRP3 protein to regulate the NLRP3 inflammasome.
Sequence 260aa to 340aa of N protein involves in the NLRP3 inflammasome activation
The sequences of SARS-CoV-2 N protein involved in the interaction with NLRP3 protein were determined by evaluating progressive truncated mutants of N protein, N1–N8 (Fig. 4a). Co-IP results showed that NLRP3 interacted with N1(1–340), N5(90–420), N6(180–420), and N7(260–420), but failed to interact with N2(1–260), N3(1–180), N4(1–90), and N8(340–420) in HEK293T cells (Fig. 4b, c) and A549 cells (Fig. 4d), suggesting that the domain containing 260aa–340aa of N protein is involved in the interaction with NLRP3.
Next, the effect of N protein progressive truncations (N1–N8) on the activation of the NLRP3 inflammasome was evaluated. The HEK293T-NLRP3 inflammasome system cells were transfected with plasmids encoding N protein or its truncated mutants N1 to N8, respectively. ELISA assays showed that IL-1β secretion, IL-1β p17 cleavage, and Casp-1 p20 maturation in the cell supernatants were significantly induced by the N1, N5, N6, and N7, but not affected by N2, N3, N4, and N8 (Fig. 4e), indicating that the domain containing 260aa–340aa of N protein is involved in the activation of the NLRP3 inflammasome. Taken together, we demonstrated that the sequence 260aa–340aa of SARS-CoV-2 N interacts with NLRP3 to activate the NLRP3 inflammasome.
The interaction of NLRP3 with ASC is promoted by SARS-CoV-2 N protein
As SARS-CoV-2 N interacts with NLRP3, we would like to know the role of N in regulating the components of the NLRP3 inflammasome. Initially, the effect of N protein on the expression of NLRP3, Casp-1, and ASC was examined. The results showed that the protein levels of NLRP3, Caspase-1, and ASC were not influenced by N protein in HEK293T cells (Fig. 5a). However, N and ASC could interact with each other in the presence of NLRP3 in HEK293T cells (Fig. 5b, c) and A549 cells (Fig. 5d, e), suggesting that N, NLRP3, and ASC three proteins together might form a complex N-NLRP3-ASC. Interestingly, the interaction of NLRP3 and ASC was enhanced by N protein in dose-dependent manners in HEK293T cells (Fig. 5f) and A549 cells (Fig. 5g). Co-IP results also demonstrated that N protein interacted with endogenous ASC in the presence of NLRP3 and the interaction was promoted by stimulating with Nigericin in PMA-differentiated THP-1-N cells (Fig. 5h). Taken together, these results suggested that SARS-CoV-2 N promotes the interaction of NLRP3 with ASC.
The assemble of the NLRP3 inflammasome is facilitated by SARS-CoV-2 N protein
The role of SARS-CoV-2 N protein in regulating the assemble of the NLRP3 inflammasome was then explored. Localization of NLRP3 as speck structure in the cytoplasm is an indicator of inflammasome complex formation30. In HEK293T cells (Fig. 6a, top) and A549 cells (Fig. 6a, bottom), NLRP3 alone was diffusely distributed in the cytoplasm, whereas in the presence of SARS-CoV-2 N protein, NLRP3 co-localized with N and formed specks (Fig. 6a). We noted that N2 (1aa–260aa) (Fig. 6b) and N8 (340aa–420aa) (Fig. 6c) were distributed in both the nucleus and cytoplasm and failed to co-localize with NLRP3, and NLRP3 failed to form specks in the cytoplasm of HEK293T cells (Fig. 6b, c, top) and A549 cells (Fig. 6b, c, bottom). Confocal microscope showed that in PMA-differentiated THP-1 macrophages, N protein alone was diffusely distributed in the cytoplasm (Fig. 6d); endogenous NLRP3 protein alone also diffusely located in the cytoplasm (Fig. 6e, top); however, in the presence of N protein, endogenous NLRP3 protein aggregated and formed specks in the cytoplasm (Fig. 6e, bottom). Collectively, these data suggested that SARS-CoV-2 N protein may promote the formation of the NLRP3 inflammasome complex.
In the process of inflammasome assembly, ASC protein is aggregated to form oligomers, which is required for Casp-1 activation31. Notably, in PMA-differentiated THP-1 macrophages (Fig. 6f) and GM-CSF differentiated BMDMs (Fig. 6g), oligomerization of endogenous ASC protein was stimulated by Nigericin, and Nigericin-induced ASC oligomerization was further enhanced in the presence of N protein (Fig. 6f, g). In addition, HEK293T-NLRP3 inflammasome system cells were transfected with plasmids encoding SARS-CoV-2-N protein and its truncated mutants, respectively. Notably, ASC oligomerization was activated by N1, N5, N6, and N7 proteins, but not influenced by N2, N3, N4, and N8 proteins (Fig. 6h), demonstrating that the sequence 260aa–340aa of N protein interacts with NLRP3 to promote ASC oligomerization.
Moreover, we explored the effect of N protein on the assembly of the NLRP3 inflammasome complex. HEK293T cells (Fig. 6i) and A549 cells (Fig. 6j) were co-transfected with plasmids as indicated. In the presence of GFP and NLRP3, GFP was distributed in both cytoplasm and nucleus, NLRP3 was diffusely distributed in the cytoplasm (Fig. 6i a–e and Fig. 6j a–e). In the presence of N and NLRP3, NLRP3 was co-localized with N and formed speck structures (Fig. 6i f–j and Fig. 6j f–j). In the presence of GFP and ASC, GFP was distributed in both cytoplasm and nucleus, ASC was distributed in nucleus and cytoplasm and formed small ring structures (Fig. 6i k–o and Fig. 6j k–o). When N and ASC presented together, N and ASC were not colocalized (Fig. 6i p–t and Fig. 6j p–t). Notably, in the presence of GFP, ASC, and NLRP3, NLRP3 and ASC were co-localized in the cytoplasm to form “ring-like” structures (Fig. 6i u–y and Fig. 6j u–y). Importantly, when N, ASC, and NLRP3 expressed together, the three protein were obviously co-localized to form “sphere-like” structures (Fig. 6i z–ad and Fig. 6j z–ad). Taken together, these data demonstrated that SARS-CoV-2 N protein promotes the assemble of the NLRP3 inflammasome complex.
SARS-CoV-2 N protein induces mice lung injury via activating the NLRP3 inflammasome
The biological role of SARS-CoV-2 N protein in activating the NLRP3 inflammasome was then assessed. As the main clinical symptom of COVID-19 is acute lung injury5, and the “cytokine storm” is considered to be an important cause of acute lung injury25. Therefore, we explored the effect of N protein on lung injury using an AAV-lung-N C57BL/6 mouse model. C57BL/6 mice were subjected to tail vein injection with AAV-Lung-EGFP or AAV-Lung-N. The results indicated that the protein levels of IL-1β and IL-6 in the sera were significantly higher in AAV-N-infected mice than that in AAV-CT-infected mice (Fig. 7a, b). Additionally, C57BL/6 mice injected with AAV-Lung-EGFP or AAV-Lung-N were subjected to intraperitoneal injection with or without Ac-YVAD-cmk. We noted that the protein levels of IL-1β and IL-6 were much higher in the sera of AAV-N-infected mice as compared to AAV-CT-infected mice, however, AAV-N-mediated inductions of IL-1β and IL-6 were repressed by Ac-YVAD-cmk (Fig. 7c, d). Notably, immunohistochemical fluorescence analyses showed that IL-1β and IL-6 proteins were highly expressed in the lungs of mice carrying AAV-Lung-N, whereas such activations were attenuated by Ac-YVAD-cmk (Fig. 7e, f). Interestingly, hematoxylin & eosin staining analyses indicated that inflammatory lesions and tissue injuries were obvious in the lungs of mice carrying AAV-Lung-N, but such pathological changes were repressed by the administration of Ac-YVAD-cmk (Fig. 7g). Collectively, these results suggested that N protein induces a systemic inflammation and lung injury in mice via activating the NLRP3 inflammasome.
C57BL/6 mice were tail vein injected with AAV-Lung-EGFP or AAV-Lung-N, treated with PBS or LPS, and intraperitoneally injected with or without Ac-YVAD-cmk, as indicated. Interestingly, the survival rates were reduced in mice treated with LPS; relatively unchanged between mice treated with LPS plus injected with AAV-Lung-EGFP and mice treated with LPS; further downregulated significantly in mice treated with LPS plus injected with AAV-Lung-N; and recovered in mice treated with LPS plus injected with AAV-Lung-N and treated with Ac-YVAD-cmk (Fig. 7h). N protein specifically expressing in lung significantly shortened survival time of mice (Fig. 7h). Taken together, we reveal a pathological mechanism of COVID-19, in which SARS-CoV-2 N protein promotes inflammation and lung injury via promoting the activation of the NLRP3 inflammasome (Fig. 8).