SARS-CoV-2 infection causes periodontal fibrotic pathogenesis through deregulating mitochondrial beta-oxidation

Since its outbreak, the coronavirus disease 2019 (COVID-19) pandemic has infected over 761 million people globally and has been a major challenge to human health. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2 virus) has multiple transmission routes such as through respiratory fluids, therefore is highly infectious [1]. Increasing numbers of long COVID and repeated infection cases have been reported [2–4]. COVID-19 has been initially connected with acute inflammatory disease, causing progressive pulmonary fibrosis [5], with epithelial cell death considered a driver of pulmonary fibrosis [6]. Evidence suggests COVID-19 affects other organs such as heart, skin, kidneys, and brain [7]. The human oral cavity and saliva are an important reservoir of the SARS-CoV-2 virus [8], and saliva has been used for diagnosis [9]. Periodontal tissues are highly vulnerable to infectious diseases and COVID-19 has been potentially connected with poor periodontal health [10]. Severe oral mucosa lesions including detachment of oral epithelium with inflammation and fibrosis have been reported in COVID-19 patients [11,12]. The periodontal ligament (PDL) is vulnerable to extrinsic pathogens; ACE2 expression is elevated in periodontal fibroblasts under gingivitis and periodontitis and can be induced by LPS and inflammatory cytokines [14,15]. However, molecular mechanisms underlying potential pathological consequences of SARS-CoV-2 infection on human oral health are not systematically investigated. SARS-CoV-2 has four structural proteins: envelope, membrane, nucleocapsid and spike. Mutations predominantly occur in the spike, particularly the receptor-binding domain, while other components remain stable [16–18]. A key step of infection is spike RBD binding to ACE2 with TMPRSS2 acting as co-receptor to trigger internalization [19,20]. Most research and vaccines have focused on the spike protein, leaving the pathological functions of the other structural proteins less understood. This study uses human periodontal tissue and fibroblasts to dissect the direct pathological effects of SARS-CoV-2 structural components on these cells.

• Cell culture: Human periodontal ligament fibroblasts (HPLF; ScienCell Cat. 2630) cultured in DMEM/F12 (Gibco 31331-028) with 20% FBS (Sigma F7524) and 1% penicillin-streptomycin (Hyclone SV30079.01). Human gingival epithelial cells (HGEPp; CELLnTEC Cat. HGEPp) cultured in CnT-57 medium. Passage 3–6 used.
• Lentiviral infection/overexpression: SARS-CoV-2 structural protein plasmids (envelope, membrane, nucleocapsid; Addgene) packaged in HEK293T/293FT using JetOPTIMUS. HPLFs infected with viral supernatants plus 10 µg/ml polybrene. Samples collected at 6 h or 48 h. For ACE2 overexpression (Addgene), similar protocol with selection using 1 µg/ml puromycin for 7–10 days.
• Recombinant spike protein treatment: His-tagged spike protein (Bio-Techne 10549-CV) applied at 500 ng/ml or 5 µg/ml for 6 h or 48 h.
• Mitochondrial β-oxidation inhibition: Etomoxir (Sigma E1905) applied at defined concentrations for 6 h or 48 h to inhibit CPT1-mediated fatty acid import into mitochondria.
• Human periodontal tissue 3D equivalents: 5×10^6 HPLFs embedded in rat tail collagen gel (collagen:DMEM:FBS 9:1:1), neutralized with NaOH, cast into 0.4 µm inserts, overlaid with 5×10^4 HGEPp, cultured and airlifted; harvested at day 14 for sectioning.
• Hydrogel-based 3D matrix assay: 1×10^6 HPLFs mixed with CELLINK SKIN+ bioink, crosslinked, cultured with media changes, harvested at day 7 for sectioning.
• Immunostaining assays: Immunofluorescence for ACE2/TMPRSS2 expression in tissue, 3D equivalents and HPLFs; BrdU incorporation for proliferation; TUNEL assay for apoptosis; senescence-associated β-galactosidase staining for senescence.
• Flow cytometry: HPLFs (± ACE2 overexpression) incubated with His-tagged spike, stained with APC-conjugated anti-His Tag antibody; data acquired on BD FACSAria II; analysis in FlowJo.
• Western blotting: Proteins separated on 4–12% Bis-Tris gels (Thermo Fisher), transferred to PVDF; detection of ACE2, TMPRSS2, Collagen I, MMP1, HADHA; Lamin B1 or GAPDH used as loading controls; quantified using LI-COR system.
• Proteomics: In-gel digestion and LC-MS/MS as per appendix; differential protein expression analyzed; dataset deposited to PRIDE (PXD041281).
• Real-time RT-PCR: SYBR Green on Roche LightCycler 480; 36β4 as housekeeping gene; analysis by 2^-ΔΔCT with technical triplicates.
• Seahorse fatty acid oxidation stress test: 3×10^3 HPLFs/well in XFe96 plates; post-infection (6 h, 48 h) washed and incubated in assay medium; palmitate oxidation measured with/without etomoxir; data analyzed with Agilent Seahorse Analytics/Wave.
• Statistics: Prism (GraphPad) used; unpaired t-tests; Dunnett’s corrections in figure legends; significance thresholds *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

• ACE2 and TMPRSS2 expression: Immunofluorescence showed robust expression of ACE2 and TMPRSS2 in gingival epithelium and periodontal ligament (PDL) cells; confirmed in cultured HPLFs by immunofluorescence and western blot. 3D human gingiva equivalents also expressed both receptors.
• Spike binding/affinity: His-tagged spike protein bound to HPLFs under basal conditions and more prominently after ACE2 overexpression; validated by flow cytometry.
• Fibrotic phenotypes induced by structural proteins: Lentiviral expression of SARS-CoV-2 envelope and membrane proteins increased HPLF proliferation (BrdU) at 6 h (membrane) and 48 h (envelope and membrane), while recombinant spike did not alter proliferation at 500 ng/ml or 5 µg/ml. Envelope and membrane also increased apoptosis (TUNEL) at 48 h and elevated cellular senescence at 6 h and 48 h.
• Extracellular matrix remodeling: All structural proteins increased Collagen I protein; MMP1 was generally reduced, with spike showing an increase at one dose condition. RT-qPCR indicated transcriptional upregulation of Collagen I and downregulation of MMP1. In 3D hydrogel culture, Collagen I deposition markedly increased, especially with envelope and membrane; MMP1 reduced with membrane and nucleocapsid.
• Mitochondrial fatty acid β-oxidation downregulation: Proteomics identified significant decreases (≤0.5-fold) in mitochondrial/FAO-related proteins after envelope/membrane treatment at 6 h and 48 h, including HADHA (trifunctional enzyme subunit alpha) and very long-chain acyl-CoA dehydrogenase isoform 2; additional decreases in cytochrome c oxidase subunit 2 and fumarate hydratase isoform were noted. Western blots confirmed HADHA downregulation. Seahorse assays showed inhibition of fatty acid β-oxidation by envelope and membrane at both time points.
• Pharmacologic FAO inhibition mirrors phenotype: Etomoxir treatment increased HPLF proliferation, apoptosis, and senescence, and elevated Collagen I production and 3D deposition, recapitulating SARS-CoV-2 envelope/membrane effects. Overall, envelope and membrane proteins are the main drivers of fibrotic degeneration-like changes in PDL fibroblasts via suppression of mitochondrial fatty acid β-oxidation.

The study addressed whether SARS-CoV-2 can directly affect periodontal fibroblasts and promote fibrotic pathogenesis. Demonstrating ACE2/TMPRSS2 expression in PDL tissues and HPLFs and spike binding supports direct susceptibility of fibroblasts to SARS-CoV-2. Functional experiments showed that envelope and membrane proteins, rather than spike alone, drive a fibrotic phenotype characterized by concurrent hyperproliferation, apoptosis, senescence, increased collagen deposition, and decreased MMP1, indicating impaired matrix turnover. Mechanistically, proteomic, immunoblot, and bioenergetic data converge on downregulation of mitochondrial fatty acid β-oxidation, including reduced HADHA, implicating FAO suppression as a central pathway. Replication of these phenotypes by etomoxir underscores causality between FAO inhibition and fibrosis-like changes. Because envelope and membrane proteins are conserved across variants, these mechanisms may be broadly relevant and could contribute to oral/periodontal sequelae of COVID-19, including in long COVID. The findings suggest the FAO pathway, including CPT1-HADHA axis, as potential therapeutic targets to mitigate COVID-19-associated fibrotic remodeling in periodontal and possibly other tissues.

This work reveals that SARS-CoV-2 envelope and membrane proteins directly induce fibrotic degeneration phenotypes in human periodontal ligament fibroblasts by downregulating mitochondrial fatty acid β-oxidation, notably reducing HADHA, resulting in altered cell proliferation, apoptosis, senescence, and excessive collagen deposition with reduced MMP1. Pharmacologic FAO inhibition reproduces these effects, highlighting FAO as a mechanistic driver and therapeutic target. Future research should delineate the precise biochemical pathways linking viral structural proteins to FAO suppression and evaluate in vivo relevance and potential interventions targeting mitochondrial fatty acid metabolism to prevent or treat COVID-19-related fibrosis in oral and other tissues.

The mechanistic link between SARS-CoV-2 structural proteins and suppression of fatty acid β-oxidation requires further biochemical elucidation, as acknowledged by the authors. The study relies on in vitro models (2D HPLFs, 3D equivalents/hydrogels) and surrogate exposure methods (lentiviral expression of individual viral structural proteins and recombinant spike protein) rather than whole-virus infection or in vivo validation, which may limit generalizability to clinical settings.