SARS-CoV-2 ORF3a induces COVID-19-associated kidney injury through HMGB1-mediated cytokine production

Abstract

The primary challenge posed by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is COVID-19-related mortality, often exacerbated by additional medical complications, such as COVID-19-associated kidney injuries (CAKIs). Up to half of COVID-19 patients experience kidney complications, with those facing acute respiratory failure and kidney injury having the worst overall prognosis. Despite the significant impact of CAKI on COVID-19-related mortality and its enduring effects in long COVID, the underlying causes and molecular mechanisms of CAKI remain elusive. In this study, we identified a functional relationship between the expression of the SARS-CoV-2 ORF3a protein and inflammation-driven apoptotic death of renal tubular epithelial cells in patients with CAKI. We demonstrate in vitro that ORF3a independently induces renal cell-specific apoptotic cell death, as evidenced by the elevation of kidney injury molecule-1 (KIM-1) and the activation of NF-kB-mediated proinflammatory cytokine (TNFα and IL-6) production. By examining kidney tissues of SARS-CoV-2-infected K18-ACE2 transgenic mice, we observed a similar correlation between ORF3a-induced cytopathic changes and kidney injury. This correlation was further validated through reconstitution of the ORF3a effects via direct adenoviral injection into mouse kidneys. Through medicinal analysis, we identified a natural compound, glycyrrhizin (GL4419), which not only blocks viral replication in renal cells, but also mitigates ORF3a-induced renal cell death by inhibiting activation of a high mobility group box 1 (HMGB1) protein, leading to a reduction of KIM-1. Moreover, ORF3a interacts with HMGB1. Overproduction or downregulation of hmgb1 expression results in correlative changes in renal cellular KIM-1 response and respective cytokine production, implicating a crucial role of HMGB1 in ORF3a-inflicted kidney injuries. Our data suggest a direct functional link between ORF3a and kidney injury, highlighting ORF3a as a unique therapeutic target contributing to CAKI.

Importance: The major challenge of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection during the pandemic is COVID-19-related mortality, which has tragically claimed millions of lives. COVID-19-associated morbidity and mortality are often exacerbated by pre-existing medical conditions, such as chronic kidney diseases (CKDs), or the development of acute kidney injury (AKI) due to COVID-19, collectively known as COVID-19-associated kidney injuries (CAKIs). Patients who experience acute respiratory failure with CAKI have the poorest clinical outcomes, including increased mortality. Despite these alarming clinical findings, there is a critical gap in our understanding of the underlying causes of CAKI. Our study establishes a direct correlation between the expression of the SARS-CoV-2 viral ORF3a protein and kidney injury induced by ORF3a linking to CAKI. This functional relationship was initially observed in our clinical studies of COVID-19 patients with AKI and was further validated through animal and in vitro cellular studies, either by expressing ORF3a alone or in the context of viral infection. By elucidating this functional relationship and its underlying mechanistic pathways, our research deepens the understanding of COVID-19-associated kidney diseases and presents potential therapeutic avenues to address the healthcare challenges faced by individuals with underlying conditions.

Keywords: CAKI; HMGB1; K18-hACE2 mice; NF-kB; ORF3a; SARS-CoV-2; TNFα and IL-6; glycyrrhizin; viral infection.

Fig 1

The correlation between SARS-CoV-2 infection and ORF3a expression that contributes to inflammation-associated apoptosis observed in RPTECs of the kidney in COVID-19 patients with CAKI. (A) Immunostaining of ORF3a in COVID-19-negative controls (Ctr, a) and COVID-19-positive subjects (b) (19). Comparative staining of NF-kB (p65) with DAPI between control and COVID-19-positive tissues (c). Arrows show nuclear p65, an indication of NF-kB activation. Co-immunolabeling for ORF3a and TNFα within RPTEC (d). (B) Comparative staining of KIM-1 (a vs b) and cleaved caspase 3 between control and COVID-19-positive tissues. The kidney shown is a representative of two COVID-19-positive and two control cases. Scale bars, 100 µM.

Fig 2

ORF3a independently induces renal cell-specific injuries and cell death through activation of NF-kB-mediated cytokine production. (A) Distinctive subcellular localization of WT and Omicron T223I mutant ORF3a proteins in HK2 cells. The WT protein is predominantly localized on the lysosomes as indicated by anti-LAMP-1 antibody and as we reported previously (28, 33). The T223I mutant mostly localizes in the ER and Golgi complex that are indicated by anti-calnexin and anti-giantin antibodies, respectively. (B) Time-course expression of WT and Omicron ORF3a T223I mutant shows concurrent elevation of KIM-1 protein levels in renal epithelial 293T cells (A). 293T cells were transfected with a WT or T223I ORF3a-carrying pCAG plasmid. Transfected cells were collected over time and as indicated in hours of post-transfection (hpt). Numeric numbers 1, 2, and 3 on the top of the Western blot represent vector-only control, WT, and T223I mutant ORF3a, respectively. (C) Expression of WT and T223I mutant proteins triggers NF-kB-mediated cytokine production. (a) Immunostaining of a RelA/p65 subunit of NF-kB in ORF3a-expression HK2 cells shows that nuclear entry of p65 is specifically associated with ORF3a, an indication of ORF3a-mediated NF-kB activation. Cells were examined at 24 hpt. (b) ORF3a induces NF-kB activation as measured by an NF-kB promoter firefly luc assay (27), and upregulation of cytokines TNFα, IL-6, and IFNβ1 by RT-qPCR. (D) Both WT and T223I ORF3a induce apoptosis and necrosis as measured by a RealTime-Glo apoptosis and necrosis assay (Promega). The levels of all markers were measured by RT-qPCR. Statistical significance: *, P  <  0.05; **, P <  0.01; ***, P <  0.001.

Fig 3

Correlation of ORF3a in kidney tissues of SARS-CoV-2-infected K18-ACE2 transgenic mice with ORF3a-induced cytopathicity and kidney injury, and reconstitution of the ORF3a effect by viral injection of Ad5-ORF3a. (A) SARS-CoV-2 infection in K18-hACE2 tg mice shows a correlation of ORF3a with inflammation-associated renal apoptosis in kidney. Immunostaining of ORF3a in a kidney cross-section of mock control (a, left) and a COVID-19 (C-19) mouse (a, right). Arrows point on p65 in “pink“ nuclei (b) indicating nuclear entry of the “red“ p65, indicative of NF-kB activation. Co-immunolabeling of ORF3a with TNFα (c), elevated IL-6 (d), cleaved Casp3 (e), and KIM-1 (f) between mock and C-19. (B) Abundant ORF3a protein production is observed in the left kidney of C57BL/6 mice injected with Ad5-DsRed-ORF3a (Ad5-ORF3a) compared with the Ad5-DsRed control (a). The ORF3a protein is predominantly localized in renal tubule cells (b), coinciding with elevated KIM-1 protein, a kidney-specific injury marker (c). Note that both Ad5 constructs carry a DsRed expression cassette in the E3 region of the adenoviral backbone. The expression cassette of ORF3a is driven by a CMV promoter located in the E1 region of the adenoviral genome. Scale bars, 100 µm.

Fig 4

Glycyrrhizin (GL4419) blocks viral replication and mitigates ORF3a-induced renal cell damage through HMGB1. (A) The two chemical structures shown below glycyrrhizin are examples of glycyrrhizin derivatives we tested, including enoxolone (NSC-35347) and a dimer of glucuronic acid (NSC-2559). A list of glycyrrhizin-related compounds is listed in Table S1. (B) Treatment of GL4419 suppresses T223I ORF3a-induced cell death, displaying minimal cytotoxicity with a selectivity index of 22.11. (C) ORF3a mitigates ORF3a-induced apoptosis and necrosis. Tests were carried out in 293T cells at 24 hpt. Cell viability was measured by MTT, cell death by trypan blue, and apoptosis and necrosis by a RealTime-Glo Apoptosis and necrosis assay (Promega). The selectivity index was calculated by the ratio of EC₅₀ over CC₅₀ based on the MTT assay. (D-a). GL4419 inhibits viral replication, with an IC₅₀ of 50.05 ± 5.22 µM, assessed in a stable BHK21 cell-based SARS-CoV-2 NanoLuc replicon system using a Nano-Glo Luciferase Assay (Promega). (D-b). GL4419 inhibits SARS-CoV-2 replication in VeroE6 cells, exhibiting an EC₅₀ of 271.8 ± 54.57. The WT USA-WA1/2020 virus was used for infection with titer of 100 TCID₅₀, and the cells were collected at 48 hpi. (E) Treatment of ORF3a-expressing 293T cells with GL4419 (300 µM) reduces KIM-1 protein production possibly through protein degradation of ORF3a. 293T cells transfected with WT and T223I ORF3a were collected at 24 hpt. GL4419 dissolved in DMSO in the concentration of 300 µM was used. DMSO only was used as a negative control. western blotting results are shown in panel a with quantification shown in panel b. Results of RT-qPCR of transfected cells were collected at 48 hpt (c). (F) treatment of WT and T223I ORF3a-expressing HK2 cells with GL4419 block activation of NF-kB (indicated by the RelA/p65 subunit of NF-kB) and HMGB1. Expression of ORF3a without GL4419 treatment (mock with DMSO added) activates HMGB1 (a) and NF-kB (b) by promoting contrasting nuclear entry and nuclear exit, respectively. In contrast, adding GL4419 (300 µM) results in the retainment of HMGB1 in the nucleus (a) and NF-kB in the cytoplasm (b).

Fig 5

ORF3a interacts with HMGB1 and modulates renal cellular KIM-1 response and cytokine production. (A) Interaction of ORF3a with HMGB1 demonstrated by reciprocal co-IP. 293T cells were transfected with WT or T223I mutant ORF3a-carrying plasmid DNA. At 24 hpt, cells were lysed for co-IP analyses. Anti-HA pulldown confirmed the presence of ORF3a and HMGB1, verified by anti-HA and anti-HMGB1 antibodies. Reciprocal IP confirmed ORF3a and HMGB1 interaction. Rabbit IgG (rIgG) and mouse IgG (mIgG) served as negative controls. (B) Overproduction of HMGB1 enhances KIM-1 response and cytokine production. 293T cells were co-transfected with HMGB1 and WT or T223I ORF3a plasmids. mRNA levels of HMGB1, ORF3a, HAVCR1 (KIM-1), and cytokines (TNFα, IL-6, and IFNβ1) were quantified by RT-qPCR at indicated times. Significance: *, P  <  0.05; **, P  <  0.01; ***, P  <  0.001. (C) Downregulation of HMGB1 by siRNA reduces KIM-1 response and cytokine production. At 24 hpt, 293T cells were transfected with siRNA targeting HMGB1 or control siRNA, followed by transfection with WT or T223I ORF3a plasmids. Cells were harvested at 24 and 72 hpt. Western blot and RT-qPCR analyses showed successful HMGB1 downregulation, reduced KIM-1, and cytokine levels, more pronounced at 72 hpt.