Advances in human respiratory organoid models for studying the pathogenesis and intervention strategies of COVID-19

Abstract

Coronavirus Disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), affects multiple organ systems, with the respiratory system being the primary target. Respiratory organoids, which closely mimic the structure and function of the human respiratory tract, have emerged as essential tools for studying SARS-CoV-2 infection. This review summarizes current methods for generating various respiratory organoids, including nasal, tonsil, airway, bronchial, and alveolar organoids, and highlights their application in investigating the mechanism of SARS-CoV-2 infection and evaluating potential therapeutic agents. Meanwhile, this review also introduces respiratory organoid-on-a-chip technology, which can precisely regulate culture conditions and incorporate vascularization and immune cells to enhance physiological complexity, thereby providing crucial support for investigating SARS-CoV-2-induced lung injury, immune responses, and conducting high-throughput drug screening. The aim of this review is to provide valuable insights for further research into the pathogenesis and intervention strategies of COVID-19.

Highlights

• •Respiratory organoids have emerged as essential tools for studying SARS-CoV-2 infection.
• •This review summarizes current methods for generating various respiratory organoids and highlights their applications.
• •It also introduces respiratory organoid-on-a-chip technology, aiming to offer insights for further SARS-CoV-2 research.

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), first emerging in December 2019, is a highly transmissible and pathogenic coronavirus responsible for Coronavirus Disease 2019 (COVID-19) (Hu et al., 2021; Zhou et al., 2020). It has spread globally, posing a severe threat to human health and public safety, leading the World Health Organization (WHO) to declare it a pandemic in March 2020 due to its widespread transmission and high fatality rate (Rahman et al., 2021; Ochani et al., 2021). As of 16 November 2025, over 0.77 billion cumulative confirmed cases have been reported worldwide, with more than 7.1 million deaths (World Health Organization, 2025). The virus constantly mutates, giving rise to numerous variants with increased transmissibility and susceptibility to re-infection (Chavda et al., 2022). To study SARS-CoV-2 and its variants, explore their infection and transmission mechanisms, and mitigate their harm to humans, establishing effective in vitro experimental models is essential.

Organoids, in vitro three-dimensional (3D) self-organizing structures derived from pluripotent stem cells (PSCs) or somatic stem cells, can replicate many aspects of the structural organization and function of their in vivo organ counterparts (Tang et al., 2022). They are widely used in disease modeling (Rossi et al., 2018), host-pathogen interactions (Sridhar et al., 2020), drug screening (Zhou et al., 2025) and patient-derived organoid biobanks (Zhou et al., 2021), showing great potential in biomedical research and translational applications (Kim et al., 2022b; Nie and Hashino, 2017). Respiratory organoids, with histological and genotypic characteristics highly similar to human respiratory systems, can reproduce various functions such as secretion and filtration, simulating the spatial structure and physiological functions of corresponding tissues and organs. Additionally, they have a short culture time and avoid ethical and moral issues (Jeziorski et al., 2023; Hostiuc et al., 2019), making them ideal tools for studying SARS-CoV-2 infection of respiratory systems. A key optimization of respiratory organoids is the regulation of cellular polarity, leading to two main configurations: apical-in and apical-out. Apical-in organoids mimic the natural in vivo structure, where the apical surface of epithelial cells (expressing viral receptors like ACE2) faces the internal lumen of the 3D organoid. In contrast, apical-out organoids are engineered to expose the apical surface outward to the culture medium, directly facilitating virus-cell contact without the need for mechanical dissociation or lumen access (Chiu et al., 2023b; Salahudeen et al., 2020). This polarity adjustment addresses a critical limitation of traditional apical-in organoids—difficulty in virus penetration into the internal lumen—which improves infection efficiency and experimental operability. For example, apical-out airway organoids have been shown to sustain productive replication of SARS-CoV-2 and its Omicron variants, with higher infectivity compared to apical-in counterparts (Chiu et al., 2023b).

Similarly, air-liquid interface (ALI)-differentiated human nasal organoids have been validated as stable models for SARS-CoV-2 infection, recapitulating viral shedding, cilia damage, and mucus hypersecretion (Tran et al., 2022; Rajan et al., 2021). ALI culture is a key in vitro complement to organoids, mimicking the respiratory tract’s physiological microenvironment by exposing epithelial cells' apical surface to air and basolateral surface to medium. This setup drives epithelial differentiation into polarized, stratified layers with in vivo-like phenotypes (cilia formation, mucus secretion, barrier function), which is essential for studying respiratory virus infection and host responses. ALI culture and organoids are interconnected rather than mutually exclusive, each with unique strengths. ALI can be established from primary respiratory epithelial cells or dissociated 3D organoid cells/clusters seeded on porous supports (e.g., transwell inserts) to form polarized epithelial layers. Organoids preserve the 3D structural complexity and cellular heterogeneity of native tissues, while ALI optimizes the simulation of epithelial barrier function and gas-exchange-related physiology, making it ideal for investigating viral entry, epithelial spread and aerosol exposure responses. Together, the two systems complement each other to address diverse research questions in respiratory disease modeling. To date, ALI cultures have been successfully established from a panel of respiratory tract organoids, including nasal organoids, tonsil organoids, lung organoids, airway organoids, bronchial organoids and alveolar organoids.

Respiratory organoids and COVID-19 research

Given that pulmonary disease is the primary lethal factor in this pandemic, there is an urgent need to investigate SARS-CoV-2-associated respiratory diseases. Infection with SARS-CoV-2 can lead to a series of changes from pneumonia to acute respiratory distress syndrome and ultimately respiratory failure (Chen et al., 2020; Tsai et al., 2021), posing a significant threat to public health. Therefore, various respiratory organoids, including nasal, tonsil, airway, bronchial, and lung organoids, are widely used in studying the pathological mechanisms of SARS-CoV-2 and testing the therapeutic effects of related drugs (Fig. 1).

Fig. 1.

Current study of COVID-19 using respiratory organoids. Respiratory organoids primarily encompass nasal organoids, tonsil organoids, airway organoids, bronchial organoids, lung organoids, and alveolar organoids. By recapitulating the physiological architecture of human respiratory tissues, these models facilitate in vitro investigations into the infection mechanisms of SARS-CoV-2 and its associated cell tropism. Furthermore, these respiratory organoids have been widely employed across multiple translational research domains, including pathogenesis exploration, preclinical drug development, precision therapy optimization, and regenerative medicine research. (This figure was created with BioRender software. https://biorender.com/).

Nasal organoids

The nasal mucosa is a key initial entry site for respiratory viruses, including SARS-CoV-2 (Tran et al., 2022). Studies have shown that SARS-CoV-2 can infect nasal turbinate tissues, mainly targeting respiratory epithelial cells, and triggers robust antiviral and inflammatory innate immune responses in the nasal mucosa (Alfi et al., 2021). Nasal organoids can effectively reproduce human nasal epithelium (HNE) and are used to assess the impact of viral evolution and develop preventive and therapeutic measures against SARS-CoV-2 (Liu et al., 2020).

Nasal organoids are typically generated from non-invasive samples such as human nasal washes and middle turbinate swabs obtained from healthy volunteers, following a reproducible and stable protocol (Rajan et al., 2021). The generation process involves tissue dissociation, cell isolation, and initial monolayer cultivation, followed by expansion through embedding in matrigel and subsequent differentiation via ALI culture when needed. The culture medium for nasal organoids is usually based on advanced DMEM/F12, supplemented with components such as HEPES, GlutaMAX, penicillin-streptomycin, B-27, N-2, and growth factors including FGF-7, FGF-10, and Heregulin beta-1 to support self-organization and differentiation (Table 1). These organoids faithfully recapitulate the cellular composition of HNE, comprising basal cells, ciliated cells, goblet cells, and club cells—key cell types involved in nasal mucosal barrier function, mucus secretion, and ciliary movement. Leveraging this physiological relevance, nasal organoids and their derived ALI-HNE cultures have been extensively used to study respiratory viral infections.

Table 1.

Recent human respiratory organoids used to study SARS-CoV-2.

Organoid Type	Stem Cell Type	Original Cell or Tissue	Initial Cultivation	Expansion and Differentiation	Culture Medium	Cellular Composition	References
Nasal organoids	SCs	Adult stem cells	Tissue dissociation and cell isolation	Embedded in matrigel for expansion ALI differentiation derived from organoids	Airway organoid medium	Basal cells, ciliated cells, club cells and goblet cells	Rajan et al. (2021)
				ALI differentiation derived from primary cells	PneumoCult™ airway organoid seeding Medium (STEMCELL Technologies, Vancouver, Canada); PneumoCult™ airway organoid differentiation medium (STEMCELL Technologies, Vancouver, Canada)	Basal cells, ciliated cells and epithelial cells	Tran et al. (2022)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, 1% HEPES, 1% GlutaMAX, and 1% Penicillin–Streptomycin; PneumaCult™-ALI (STEMCELL Technologies, Vancouver, Canada)	Basal cells, ciliated cells, club cells and goblet cells	Chiu et al. (2022)
				Embedded in medium for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, 1% HEPES, 1% GlutaMAX, and 1% Penicillin-Streptomycin	Basal cells, ciliated cells, club cells and goblet cells	Li C. et al. (2023)
				Embedded in medium for expansion ALI differentiation derived from organoids	BiomOrgan medium	Basal cells, ciliated cells, club cells and goblet cells	Wan et al. (2025)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, L-WRN Conditioned Medium, GlutaMAX, 1 M HEPES, 10000 U/mL Penicillin-Streptomycin, B-27, N-2, Primocin®, N-acetylcysteine, Nicotinamide, Y-27632 2HCl, A-83-01, SB202190, hFGF-7, hFGF-10, and Heregulin beta-1	Ciliated cells, club cells and goblet cells	Zhang et al. (2024)
Tonsil organoids	/	Tonsil tissues	Tissue dissociation and cell isolation	Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, Antibiotic-Antimycotic, GlutaMAX, 10% R-spondin 1, B-27, 50 ng/mL HGF, 100 ng/mL noggin, 50 ng/mL FGF10, 20 ng/mL bFGF, 20 nM A83-01, 10 μM PGE2, and 10 mM Nicotinamide	B cell subpopulations, T cell subpopulations, and APCs	Kim H.K. et al. (2022)
				Cultured in medium for expansion and differentiation	RPMI 1640 medium with GlutaMAX, 10% FBS, nonessential amino acids, sodium pyruvate, Penicillin-Streptomycin, Normocin, and insulin/selenium/transferrin cocktai	B cell subpopulations, T cell subpopulations, and APCs	Wagar et al. (2021)
Lung organoids	PSCs AT cells	iPSCs and airway cells	Tissue dissociation and cell isolation PSCs induced to endoderm, induced to anterior foregut, induced to spheroids	Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, Wnt3a, R-spondin, Noggin, B-27, GlutaMAX, Penicillin-Streptomycin, A-83-01, SB202190, Y-27632, FGF7, FGF10, N-Acetylcysteine, and Nicotinamide	AT1 cells, AT2 cells, basal cells, ciliated cells, goblet cells, club cells and generic lung lineage cells	Tindle et al. (2021)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, 500 ng/mL R-Spondin 1, 25 ng/mL FGF7, 100 ng/mL FGF10, 100 ng/mL Noggin, 500 nM A83-01, 5 μM Y-27632, 500 nM SB202190, B-27, 1.25 mM N-Acetylcysteine, 5 mM Nicotinamide, GlutaMAX, 10 mM HEPES, and Antibiotic-Antimycotic	AT1 cells, AT2 cells, basal cells, goblet cells, club cellsand generic lung lineage cells	Chiok et al. (2025)
				Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, B-27, N-2, 50 μg/m Lascorbic acid, GlutaMAX, 1% monothioglycerol, 0.05% BSA, and Penicillin-Streptomycin	Bronchial epithelial cells and alveolar epithelial cells	Kakizaki et al. (2025)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	/	AT1 cells, AT2 cells, basal cells, club cells, goblet cells, pulmonary neuroendocrine cells and smooth muscle cells	Leibel et al. (2024)
Airway organoids	AT cells PSCs ASCs MSCs	Airway cells ESCs Lung tissues Human airway stem cells	Tissue dissociation and cell isolation PSCs induced to endoderm, induced to anterior foregut, induced to spheroids	Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, 10 mM Nicotinamide, N-Acetylcysteine, B-27, 100 ng/mL Noggin, 50 ng/mL EGF, and 100 nM A83-01	AT2 cells, basal cells, ciliated cells and club cells	Salahudeen et al. (2020)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	RPMI1640 and complete serum free differentiation medium	Basal cells, ciliated cells and goblet cells	Duan et al. (2021)
				Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, B-27, N-2, GlutaMAX, Penicillin-Streptomycin, 10 ng/mL EGF, 10 ng/ML FGF10, 500 ng/mL R-spondin 1, 100 ng/mL Noggin, and 10 μM Y-27632	Basal cells, ciliated cells and goblet cells	Hysenaj et al. (2023)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	/	Basal cells, ciliated cells, club cells and goblet cells	Chiu et al. (2023b)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, B-27, N-2, GlutaMAX, Penicillin-Streptomycin, EGF, FGF10, R-spondin 1	Basal cells, ciliated cells, club cells and goblet cells	Lamers et al. (2021)
				Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, B-27, 1 mM N-Acetylcysteine, 10 nM [Leu15]-Gastrin I, 10 ng/mL EGF, 100 ng/mL Noggin, and R-spondin-1	Basal cells and club cells	Ebisudani et al. (2021)
				Embedded in medium for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, GlutaMAX, HEPES, Penicillin-Streptomycin, 10% R-spondin1, 1% B-27, 25 ng/mL Noggin, 1.25 mM N-Acetylcysteine, 10 mM Nicotinamide, 5 nM Heregulin beta-1, 100 μg/mL primocin, 5 μM Y-27632, 500 nM A83-01, 500 nM SB202190, 25 ng/mL FGF-7, and 100 ng/mL FGF-10	Basal cells, ciliated cells, club cells and goblet cells	Simoneau et al. (2024)
Bronchial organoids	PSCs	Human bronchial epithelial cells (HBECs) obtained via bronchoscopic sampling ESCs	Tissue dissociation and cell isolation Co-cultivated with fibroblasts PSCs induced to endoderm, induced to anterior foregut, induced to spheroids	Embedded in medium for expansion ALI differentiation derived from primary cells	/	Ciliated cells and goblet cells	He et al. (2021)
				Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, 5 ng/ml FGF2, 20 ng/ml FGF7, 100 ng/ml FGF10, 100 ng/ml Noggin, 300 ng/ml R-spondin 1, 10 μM Y-27632, 500 nM SB202190, 1 μM A83-01, B-27, 1.25 mM N-Acetylcysteine, 5 mM Nicotinamide, GlutaMAX, 10 mM HEPES, 100 U/ml Penicillin-Streptomycin, 50 μg Primocin	Basal cells, ciliated cells, club cells and goblet cells	Sano et al. (2022)
				Embedded in medium for expansion ALI differentiation derived from primary cells	Advanced DMEM/F12, B-27, N-2, 10 mM HEPES, Glutagro, 50 μg/mL Ascorbic Acid, and 0.4 mM Monothioglycerol	Bronchial epithelial cells and ciliated cells	Samuel et al. (2020)
				Embedded in medium for expansion ALI differentiation derived from primary cells	Bronchial epithelial cell medium (ScienCell, California, America)	Basal cells, ciliated cells, club cells and goblet cells	Park et al. (2024)
Alveolar organoids	PSCs AT cells	iPSCs Lung tissues	iPSCs induced to iAT2 Tissue dissociation and cell isolation PSCs induced to endoderm, induced to anterior foregut, induced to spheroids	Embedded in matrigel for expansion ALI differentiation derived from organoids	Advanced DMEM/F12, GlutaMAX, B-27, N-2, 10 μM SB431542, 2 μM DM, 10 μM Y-27632 dihydrochloride, 3 μM CHIR99021, 10 ng/mL rhBMP4, 100 nM All-trans RA, 10 ng/mL KGF, 50 nM Dexamethasone, 0.1 mM 8-Br-cAMP, and 0.1 mM IBMX	AT2 cells	Huang et al. (2020)
				Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, 10 mM HEPES, Penicillin-Streptomycin, 1 mM N-Acetylcysteine, 10 mM Nicotinamide, B-27, 10% R-Spondin-1, 50 ng/mL rhEGF, 100 ng/mL FGF7, 100 ng/mL FGF10, 100 ng/mL Noggin, 10 μM SB431542, and 3 μM CHIR99021	AT1 cells and AT2 cells	Youk et al. (2020)
				Embedded into matrigel for expansion and differentiation	Advanced DMEM/F12, 1% HEPES, 1% GlutaMAX, and 1% Penicillin-Streptomycin	AT1 cells, AT2 cells, basal cells, ciliated cells, club cells and goblet cells	Li C. et al. (2023)
				Embedded into matrigel for expansion and differentiation	/	AT2 cells	Chi et al. (2025)
				Embedded into matrigel for expansion and differentiation	cSFDM with 2 μM Dorsomorphin,10 μM SB431542,10 ng/ml rhBMP4, 100 nM RA, 50 nM dexamethasone, 0.1 mM 8-Br-cAMP, 0.1 mM IBMX, 10 ng/ml rhKGF, and 3 μM CHIR99021; STEMdiff™ Definitive Endoderm Kit (STEMCELL Technologies, Vancouver, Canada)	AT2 cells	Gandikota et al. (2024)

Abbreviations: ALI, air-liquid interface; AT1 cells, alveolar type 1 cell; AT2 cells, alveolar type 2 cell; ASC, alveolar stem cell; APC, antigen-presenting cell; DM, dorsomorphin; ESC, embryonic stem cell; EGF, epidermal growth factor; iPSC, induced pluripotent stem cell; iAT2, iPSC-derived alveolar type 2 cells; MSC, mesenchymal stem cell; PSC, pluripotent stem cell; PGE2, prostaglandin E2; RA, retinoic acid; SC, stem cell.

Evaluations of two main types of human respiratory viruses, respiratory syncytial virus (RSV) and SARS-CoV-2, revealed that infected human nose organoid-ALI samples can provide insights into various aspects of RSV and SARS-CoV-2 infections, including viral shedding, cilia damage, innate immune response, and mucus hypersecretion (Rajan et al., 2021). An ALI differentiated HNE culture system has been tested as a realistic model of SARS-CoV-2 infection (Tran et al., 2022). It was demonstrated that SARS-CoV-2 infection with ALI-HNE established from different donors was stable and reproducible (Tran et al., 2022). A differentiation protocol was established to generate 3D differentiated nasal organoids, which outperformed existing models, especially two-dimensional (2D) ones (Chiu et al., 2022). Wan et al. established an organoid neutralization assay to determine the potency of the monoclonal antibody VIR-7831 (a neutralizing monoclonal antibody targeting SARS-CoV-2) (Wan et al., 2025). Cell line validation lacks in vivo relevance and is prone to misjudging therapeutic efficacy, whereas the nasal organoid neutralization assay can simulate human respiratory epithelium, thereby providing a reliable assessment of therapeutic efficacy (Wan et al., 2025). To evaluate the replicative fitness of Omicron and earlier variants in human respiratory epithelial cells, some scholars used nasal organoids (Li C. et al., 2023; Zhang et al., 2024). Results showed that Omicron BA.5 exhibited higher entry efficiency and significantly increased replicative capacity and infectivity compared to B.1.1.529 and the SARS-CoV-2 wild strain in human nasal and airway organoids (Li C. et al., 2023). Moreover, Zhang et al. found that the replication capacity of Omicron EG.5.1 and XBB.1.9.1 in nasal organoids derived from young individuals was superior to that from elderly counterparts, whereas their enhanced immune escape properties enabled EG.5.1 to emerge as the dominant strain within early sublineages (Zhang et al., 2024).

Nasal organoids are a robust choice for SARS-CoV-2 research, as they highly mimic human respiratory epithelium, enable non-invasive sample collection, and faithfully recapitulate early infection events like viral entry and cilia damage (Table 2). While their main limitations include the lack of immune/endothelial components and incomplete immune response modeling, these can be mitigated by integrating vascularization and immune cells via organoid-on-a-chip technology.

Table 2.

Infection characteristics of SARS-CoV-2 in human respiratory organoids.

Organoid Type	Tropism	Drug Screening	Application	Advantages	Disadvantages and Possible solutions		References
					Disadvantages	Possible solutions
Nasal organoids	Ciliated cells	Camostat and E64D	Exposure to SARS-CoV-2 and treatment; drug screening	Highly simulate human respiratory infection; non-invasive sample collection	Lack non-epithelial components (e.g., immune/endothelial cells); cannot model complete immune responses	Construct organoid-on-a-chip and introduce vascularization and immune cells	(Rajan et al., 2021; Tran et al., 2022; Chiu et al., 2022; Li et al., 2023a; Wan et al., 2025; Zhang et al., 2024)
Tonsil organoids	Basal cells	Remdesivir	Exposure to SARS-CoV-2 and drug screening	High viral infection efficiency; easily accessible samples	Require specialized medium; high operational complexity	Standardize operation procedures and reduce operational variables	(Kim H.K. et al., 2022; Wagar et al., 2021)
Lung organoids	AT2 cells, Ciliated cells, club cells, and goblet cells	EIDD-2801, Nafamostat, EST, Apilimod, Remdesivir, and SP-B	Exposure to SARS-CoV-2 and treatment; drug screening	Recapitulate in vivo pathology and infection processes	Not suitable for high-throughput screening; limited cell types	Introduce automated equipment to enhance screening efficiency; construct organoid-on-a-chip and introduce vascularization and immune cells	(Tindle et al., 2021; Chiok et al., 2025; Kakizaki et al., 2025; Leibel et al., 2024)
Airway organoids	AT2 cells, Ciliated cells, club cells, and goblet cells	GW6471 and IFN-λ1	Exposure to SARS-CoV-2; drug screening	Contain multiple epithelial and stem cell types; high physiological relevance	Complex culture procedure; relatively high cost	Batch standardized preparation and simplification of operational procedures	(Salahudeen et al., 2020; Duan et al., 2021; Hysenaj et al., 2023; Chiu et al., 2023b; Lamers et al., 2021; Ebisudani et al., 2021; Simoneau et al., 2024)
Bronchial organoids	AT2 cells, Ciliated cells, and goblet cells	Camostat and Remdesivir	Exposure to SARS-CoV-2 and treatment; drug screening	High viral infection efficiency	Require specialized medium; operationally complex	Introduce batch standardized preparation and automated equipment	(He et al., 2021; Sano et al., 2022; Samuel et al., 2020)
Alveolar organoids	AT2 cells	Remdesivir, Camostat and E64D	Exposure to SARS-CoV-2 and drug screening	High physiological relevance; reusable for long-term studies, reducing cost	Generally low infection rate; lack immune and stromal cells	Utilize the ALI to expose apical receptors of cells (e.g., ACE2), or perform mechanical dissociation of 3D organoids to enhance viral contact; construct organoid-on-a-chip and introduce vascularization and immune cells	(Huang et al., 2020; Youk et al., 2020; Li C. et al., 2023; Chi et al., 2025; Gandikota et al., 2024)

Abbreviations: ALI, air-liquid interface; EST, nafamostat, (2S, 3S)-trans-Epoxysuccinyl-L-leucylamindo-3-methylbutane ethyl ester; SP-B, pulmonary surfactant protein.

Tonsil organoids

Data analysis indicates that COVID-19 patients often present with enlarged tonsils (El-Anwar et al., 2020). SARS-CoV-2 is suggested to induce mucosal immune responses through the nasal passages and Waldeyer’s ring of tonsils and adenoids, potentially confining the virus to the upper respiratory tract (Russell and Mestecky, 2022). However, small laboratory animals like mice and hamsters lack the systems to replicate the basic features of human adaptive immunity (Cai et al., 2022). The reassociation of human tonsillar cells in ex vivo culture leading to organoid formation provides a new tool for assessing human humoral immune responses to infection (Kenter and Richner, 2021). Therefore, establishing a tonsil organoid system based on the wide availability of human tonsils offers a new direction for SARS-CoV-2 research.

Kim et al. inoculated tonsil epithelial cells after enzymolysis of tonsil tissue into the human prostatic organoid culture medium (HPM) and found that the organoid formation efficiency was highest when cultured in HPM without epidermal growth factor (Kim H.K. et al., 2022). Subsequent studies revealed that SARS-CoV-2 infection in tonsil epithelial organoids activated tumor suppressor M signals, remodeled lipid metabolism, but inhibited innate immune signals (Kim H.K. et al., 2022). Additionally, remdesivir was found to reduce viral copy number and intracellular viral protein levels in organoid culture supernatant in a dose-dependent manner (Kim H.K. et al., 2022). Tonsil epithelial organoids are considered a platform for preclinical and translational research to study the infectivity and transmissibility of SARS-CoV-2 and evaluate antiviral candidates. Furthermore, as the tonsil is an accessible lymphatic organ, the developed tonsil organoids can be used to evaluate immune responses to vaccine candidates and the effects of different adjuvants. Wagar et al. used replication-deficient type 5 adenovirus (Ad5) encoding either the full-length viral spike protein, spike and nucleocapsid proteins, or the S1 spike subunit with nucleocapsid in the Ad5 E1 region as vaccine candidates against SARS-CoV-2. At 14 days post stimulation, they observed plasmablast differentiation and significant CD8⁺ T cell activation in a subset of donors compared to unstimulated controls (Wagar et al., 2021). Moreover, specific IgA and IgG antibodies for spike and nucleocapsid proteins were detected by protein microarray in a few donors (Wagar et al., 2021). Although epithelially derived tonsil organoids are valuable for modeling certain aspects of adaptive immunity to SARS-CoV-2, they primarily recapitulate epithelial components and lack a full spectrum of immune cell subtypes. Moreover, they cannot fully mimic the complex in vivo microenvironment (Te Marvelde et al., 2025), which limits their ability to replicate integrated immune regulatory networks present in human tissues.

Lung organoids

Lung organoids (LOs) which can be obtained from primary cell culture and stepwise induction of PSCs could be used for virus pathogenesis, drug development, regenerative medicine and precision treatment (Peng et al., 2022). The cultivation of LOs consists of four definitive periods: (1) definitive endoderm (DE), (2) anterior foregut endoderm (AFE), (3) lung progenitor cells (LPCs), and (4) various types of LOs (Peng et al., 2022; Chen et al., 2017). Monolayers derived from adult LOs, primary airway cells, or hiPSC-derived alveolar epithelial type II (AT2) pneumocytes can be infected with SARS-CoV-2 to create in vitro lung models of COVID-19 (Tindle et al., 2021). Studies had demonstrated that the host’s response to SARS-CoV-2 was largely dependent on pulmonary surfactant protein-B (SP-B), which played a crucial role in signal transduction, viral resistance, inhibiting the production of systemic inflammatory cytokines, and reducing cell apoptosis (Leibel et al., 2024). Infection with SARS-CoV-2 relied on the protease-mediated activation of the viral spike (S) protein, while Omicron variants could still replicate efficiently in TMPRSS2-knockout respiratory organoids (Kakizaki et al., 2025). In contrast, infection with the Delta variant could be blocked by protease inhibitors, indicating that TMPRSS2 played a crucial role in SARS-CoV-2 infection (Kakizaki et al., 2025). By establishing lung organoids, organoid-derived monolayers, and organoid-derived ALI cultures, Chiok compared the infection kinetics and host antiviral responses between SARS-CoV-2 and MERS-CoV (Chiok et al., 2025). Among these, ALI cultures optimally mimicked the respiratory tract environment, supporting infection by both viruses and eliciting diverse immune responses, while monolayer cultures exhibited restricted infection and incomplete immune responses (Chiok et al., 2025).

Airway organoids

The airways, particularly the distal airways, suffer severe infection consequences during the COVID-19 epidemic (Salahudeen et al., 2020). This is mainly because ciliated cells from the upper airway can easily confer downstream infection potential, and the virus reaching the larynx (lower airway) continues to infect the alveoli. It indicates that once SARS-CoV-2 infects the lower airway, it can replicate and continue to infect respiratory cells (de Oliveira et al., 2021; Cespedes and Souza, 2020; Qi et al., 2020). Therefore, establishing a suitable airway model is crucial for further understanding the mechanisms of SARS-CoV-2 infection. The US FDA has approved several therapeutics, including Paxlovid (Harris, 2023), remdesivir (Cassidy et al., 2020), nirmatrelvir-ritonavir, and molnupiravir (Toussi et al., 2023), through emergency use authorization to combat COVID-19. However, the efficacy of these treatments has been challenged by viral escape mutations (Rehman et al., 2021). Thus, it is essential to establish stable screening models to identify potential drugs to protect against SARS-CoV-2 infection. Pulmonary airway-like organoids can be obtained by inducing PSCs, multipotent stem cells, and alveolar cells, containing ciliated cells, basal cells (BCs), secretory cells, and CC10-secreting club cells (de Oliveira et al., 2021) (Table 1). Regarding cultivation methods, these organoids can be maintained in a 3D configuration. The 3D culture environment is typically established using a scaffold (e.g., Matrigel, one of the most commonly used scaffolds in current research), a non-adherent surface, or other specialized substrates that support the self-organization and structural integrity of the organoids (de Oliveira et al., 2021). To keep these organoids as a 2D monolayer for easier pathogen exposure, they can be dissociated and seeded onto transwell inserts (de Oliveira et al., 2021).

A mature airway organoid was generated by deriving adult stem cells from the lung tissue and inducing proximal differentiation following a defined protocol (Li et al., 2022; Chiu et al., 2023a). A “2D airway organoid” screening system has been established for SARS-CoV-2 infection and drug screening (He et al., 2021). Anti-receptor-binding domain (anti-RBD) polyclonal antibodies (pAbs), generated from rabbit serum, showed effective viral neutralization and protected airway organoids from SARS-CoV-2 infection and tissue damage (He et al., 2021). High-throughput screening using human pluripotent stem cell-derived airway organoids (hPSC-AOs) revealed that GW6471 blocks SARS-CoV-2 wild-type and B.1.351 variant by inhibiting HIF1α (Duan et al., 2021). Additionally, the hPSC-AO platform demonstrated that three compounds, including xanthohumol, 5-(tetradecyloxy)-2-furoic acid, and ND-646, that inhibited fatty acid biosynthesis are also effective in blocking SARS-CoV-2 infection (Duan et al., 2021). Phenotypically stable airway organoids from 20 different subjects characterized epithelial responses to SARS-CoV-2 infection and identified tetraspanin-8 (TSPAN8) as a facilitator of SARS-CoV-2 infection (Hysenaj et al., 2023). In patients infected with non-SARS-CoV-2 respiratory viruses, airway brush samples exhibited reduced TSPAN8 expression, whereas this indicator remained stable in COVID-19 patients. Additionally, cells co-expressing both ACE2 and TSPAN8 were present in the lungs of two COVID-19 patients (Hysenaj et al., 2023). TSPAN8-blocking antibodies reduced SARS-CoV-2 infection (Hysenaj et al., 2023), highlighting the utility of airway organoids for mechanistic studies and drug screening. By utilizing an adult stem cell-derived human airway organoid model overexpressing the ACE2 receptor (ACE2-OE), Simoneau et al. demonstrated that the NF-κB inhibitor alpha gene (encoding IκBα) can reduce NF-κB nuclear translocation and enhance viral infection, revealing the regulatory role of the NF-κB feedback loop in viral replication (Simoneau et al., 2024). A recently established apical-out airway organoids (AOs), which morphologically and functionally recapitulated the human airway epithelium comparably to apical-in AOs, sustained productive and multicycle replication of SARS-CoV-2. It exhibited higher infectivity and replicative fitness of SARS-CoV-2 and the Omicron variants BA.5 and B.1.1.529 (Chiu et al., 2023b). In summary, airway organoids are a robust choice for SARS-CoV-2 research, characterized by their inclusion of multiple epithelial and stem cell types and high physiological relevance, which facilitate faithful simulation of viral tropism and reliable drug screening. While they suffer from complex culture processes and relatively high costs, these drawbacks can be alleviated via batch standardized preparation and simplified operating procedures, underscoring their significant value in investigating SARS-CoV-2 pathogenesis and therapeutic interventions (Table 2).

Bronchial organoids

Existing studies indicate that SARS-CoV-2 often causes severe acute bronchopneumonia through bacterial superinfection (Menter et al., 2020), with imaging features of bronchial thickening observed in children (Castagnoli et al., 2020). Exploring the mechanism of SARS-CoV-2 on the bronchus can provide new insights for COVID-19 research.

Significant differences exist between mouse and human bronchi, such as the absence of BCs and low numbers of goblet cells in mice (Rock et al., 2010). Meanwhile, cynomolgus and rhesus macaques, which are closest physiologically to humans, have a slower reproduction rate (Takayama, 2020). Bronchial organoids (BOs) can simulate human organ physiological conditions and are suitable for large-scale drug screening (Ranga et al., 2014). Fang et al. analyzed high-throughput sequencing data of uninfected and SARS-CoV-2 infected human BOs in the Gene Expression Omnibus (GEO) database (Fang et al., 2020). They found that granulocyte colony-stimulating factor (CSF3) was significantly up-regulated after SARS-CoV-2 infection (Fang et al., 2020), suggesting CSF3 as a potential drug target. BOs, lung biopsies, and lung-derived Calu-3 and A549 cells were used to screen microRNAs that could inactivate the key untranslated regions at the 5′ ends of all encoded SARS-CoV-2 transcripts. Meta-analysis showed that miR-5004-3p expression in trachea, lung, and BOs decreased significantly after SARS-CoV-2 infection (P < 0.01), indicating that hsa-miR-5004-3p, which is a unique human microRNA, could target the precursor sequence of SARS-CoV-2 (Mohammadi-Dehcheshmeh et al., 2021). Sano et al. amplified and differentiated BOs from normal bronchial epithelial cells and analyzed cellular responses to SARS-CoV-2 infection (Sano et al., 2022). They found that differentiated BOs were almost not infected with SARS-CoV-2 due to limited access to ciliated cells located in the lumen of BO (Sano et al., 2022). Park et al. utilized human bronchial epithelial (HBE) organoids to investigate the infection dynamics of four β-coronaviruses-namely HCoV-OC43, SARS-CoV, MERS-CoV, and SARS-CoV-2. They found that all of these viruses are capable of replicating in the organoids, with SARS-CoV-2 exhibiting the strongest replicative capacity and a propensity for infecting ciliated cells (Park et al., 2024). Meanwhile, host viral receptor genes and protease genes were upregulated, and these findings provided insights into elucidating the pathogenic mechanisms of β-coronaviruses (Park et al., 2024). Using a BO-derived ALI model for infection, they discovered that BCs played a vital role in repairing the upper bronchial cortex after SARS-CoV-2 infection (Sano et al., 2022). However, as BOs are still in the early stages of SARS-CoV-2 research, challenges remain, such as low infection rates in spherical BOs (Sano et al., 2022). New organogenic technologies, including apical-out polarity optimization, organoid-on-a-chip integration with vascular/immune components, and gene modification, are being developed to address these issues and enhance infection rates, with BOs holding broad prospects for future research.

Alveolar organoids

Alveolar lung-like organoids can be obtained from primary cell cultures containing mesenchymal cells and alveolar epithelial cells (Peng et al., 2022). Studies have shown that in the alveolar model, the main targets of SARS-CoV-2 are AT2-like cells, which exhibit strong chemokine induction upon sensing (Han et al., 2021). Huang et al. demonstrated that SARS-CoV-2 infected AT2-like cells develop an inflammatory phenotype, inducing upregulation of NF-κB signaling and loss of the mature alveolar program (Huang et al., 2020). This finding highlighted the critical role of AT2-like cells in SARS-CoV-2 infection and drug performance testing.

To establish alveolar lung-like organoids in vitro, Salahudeen et al. used single adult AT2 or KRT5⁺ BCs to generate distal lung-like organoids with apical outward polarity, presenting ACE2 on the exposed outer surface to facilitate SARS-CoV-2 infection of AT2 and basal cultures (Salahudeen et al., 2020). Huang et al. generated iPSC-derived AT2s cultured in a 2D ALI and 3D epithelial spheres expressing surfactant protein-C (Huang et al., 2020). Youk et al. developed a 3D culture technique for hAT2 from healthy donor lungs, establishing cell polarity and stabilizing AT2 and showing transcriptional changes, IFN responses, and expression of interferon-stimulated genes three days post SARS-CoV-2 infection (Youk et al., 2020). Comparatively, 2D cultures exhibited superior performance in viral titer detection compared to 3D organoids, while 3D organoids created a more confined environment with less air exposure, making them more suitable for studying viral infection mechanisms than 2D cultures (Lamers et al., 2021). Furthermore, Gandikota et al. used iPSC to construct three types of lung alveolar organoids: apical-in, sheared, and apical-out (Gandikota et al., 2024). All of them were able to effectively support the infection of the novel coronavirus and its variants (Gandikota et al., 2024). Among them, the WA1 strain had the strongest replication ability, while the Omicron strain was the weakest (Gandikota et al., 2024). Following infection, it primarily triggered inflammatory and interferon responses centered on the NF-κB signaling pathway, serving as a valid tool for investigating viral infection mechanisms and host responses (Gandikota et al., 2024). For drug screening, protease inhibitors like lopinavir and nelfinavir, used for HIV infection treatment, and the nucleotide prodrug remdesivir showed modest antiviral effects against SARS-CoV-2 in constructed alveolar-like spheres (Ebisudani et al., 2021). Low-dose interferon λ1 treatment also reduced viral replication and infection (Salahudeen et al., 2020; Lamers et al., 2021; van der Vaart et al., 2021). Studies have shown that coronaviruses encode nonstructural protein 15 (Nsp15), which, through its endonuclease activity, has been proven to inhibit type I and type III interferon responses and RNase L activation in alveolar organoids, thereby exerting an immunosuppressive effect (Chi et al., 2025). Endonuclease-deficient mutants exhibited attenuated virulence in K18-hACE2 mice, demonstrating that Nsp15 was a key factor promoting viral infection and thus making it an ideal antiviral target (Chi et al., 2025). Current alveolar organoids lack the full cellular complexity of native tissues and cannot form intact alveoli (Li et al., 2020). Nonetheless, these organoids retain notable utility in the pathological research of respiratory diseases (Li et al., 2020). They serve as a robust model for investigating SARS-CoV-2-induced cellular mechanisms, which elucidates COVID-19-associated lung infections (Li et al., 2020).

Across nasal, tonsil, lung, airway, bronchial, and alveolar organoids, a common reliance on 3D embedding in Matrigel® or similar scaffolds is evident for expansion, while ALI differentiation is frequently employed to enhance physiological relevance (Table 1). Notably, the cellular composition of these organoids reflects their respective anatomical origins: nasal and airway models are enriched with ciliated, goblet, and basal cells, whereas alveolar organoids predominantly feature AT2 cells. Despite these shared strategies, significant variations exist in culture media formulations and growth factor cocktails, underscoring the need for standardized protocols to improve reproducibility and cross-model comparisons. Nasal and airway organoids, which exhibit high infectivity in ciliated and secretory cells, are particularly suitable for studying early viral entry and transmission, as well as for neutralization assays and drug screening. In contrast, alveolar organoids, primarily targeting AT2 cells, better model distal lung pathology and cytokine-driven immune responses, though their generally lower infection rates can be improved through apical-out polarity or ALI adaptations. While each model offers unique advantages—such as the physiological relevance of airway organoids and the accessibility of nasal and tonsil organoids—common limitations include the lack of immune/stromal components and operational complexity (Table 2), which are increasingly addressed through organoid-on-a-chip integration and protocol optimization.

Respiratory organoid-on-a-chip and COVID-19 research

Compared with traditional organoids, which typically lack a controllable microenvironment and diverse cellular composition, the organoid-on-a-chip platform exhibits significant advantages (Li et al., 2025). Not only can it precisely regulate culture conditions, but it also enhances physiological complexity by incorporating vascularization and immune cells while preserving genetic heterogeneity (Li et al., 2025; Zhao et al., 2024). Thus, this technology can closely simulate the microenvironment and functions of human organs on a chip, making it highly suitable for preclinical human physiology research (Zhou et al., 2025).

Several researchers have constructed biomimetic human disease models based on lung-on-a-chip, which simulate the physiological microenvironment of the respiratory tract and recapitulate SARS-CoV-2-induced lung injury and immune responses in vitro at the organ level, thereby enabling personalized medicine and high-throughput drug screening (Cao et al., 2022; Fisher et al., 2023; Faley et al., 2025). Fujimoto et al. developed a microfluidic chip co-culturing bronchial organoids with a vascular bed. They demonstrated that the virus preferentially infects bronchial epithelial cells, causes vascular injury via the paracrine secretion of type I interferons, and that inhibition of interferon-related pathways or the JAK/STAT pathway can alleviate this injury-indicating that this microphysiological system (MPS) can reflect SARS-CoV-2-induced vascular damage (Fujimoto et al., 2024). To elucidate the mechanism underlying pulmonary vascular injury following SARS-CoV-2 infection, Thacker et al. utilized a human lung-on-a-chip model. They revealed that SARS-CoV-2, after infecting alveolar epithelial cells, spread to the underlying endothelial cells, leading to impaired barrier integrity, a procoagulant phenotype, and sustained inflammatory responses. Notably, this vascular injury was independent of cytokine storms, and the inhibitory effect of tocilizumab was limited (Thacker et al., 2021). Furthermore, studies have shown that SARS-CoV-2 can activate cGAS-STING signaling in endothelial cells through the release of mitochondrial DNA, thereby triggering cell death and type I interferon production. Persistently high levels of interferons in the later stage were associated with abnormal inflammation and poor prognosis (Domizio et al., 2022). In mice, inhibiting STING improved virus-induced pulmonary inflammation and enhanced prognosis (Domizio et al., 2022). Lung-on-a-chip platforms can be used not only for investigating the infection mechanisms of viruses such as SARS-CoV-2 but also for drug screening and efficacy evaluation. As a high-throughput organoid-on-a-chip system, PREDICT96-ALI can accurately mimic human tracheobronchial tissues, making it suitable for assessing the viral kinetics of SARS-CoV-2 variants and screening antiviral drugs. It thus provides an efficient and predictive platform for drug screening (Lopez Quezada et al., 2024).

Conclusions and perspective

Since its emergence in late December 2019, SARS-CoV-2 has spread globally, posing a major threat to people’s lives and health. Organoids, mouse models, and clinical samples are complementary in COVID-19 translational research. Mouse models can recapitulate systemic physiological responses but are constrained by inherent species differences, while clinical samples reflect authentic pathological conditions yet suffer from limitations in accessibility and experimental manipulability. Organoids effectively bridge these gaps, combining human tissue-specificity with experimental controllability to provide a robust platform for mechanistic studies and therapeutic development (Fig. 2). Organoid studies enable cell-to-cell communication within a single organ and reproduce pathogen-cell interactions, including detecting viral targets of action and releasing associated factors. Respiratory organoids can serve as powerful disease models to investigate SARS-CoV-2 pathogenesis and provide a valuable platform for drug screening to identify potential COVID-19 candidate therapies and vaccines. Proximal airway cells are critical for sustained SARS-CoV-2 infection, while distal alveolar differentiation (AT2→AT1) is key for triggering the overzealous host immune response in fatal disease. Adult lung organoid monolayers with well-mixed proximodistal airway components can recapitulate both processes (Tindle et al., 2021). Regarding COVID-19 vaccine responses, many variabilities exist, with some individuals unable to mount B cell and/or T cell responses (Wagar et al., 2021), further highlighting the need for respiratory organoid systems to assess broad immunogenicity before clinical studies.

Fig. 2.

Comparative evaluation of organoids, animal models, and clinical samples for translational research. Organoids offer a semi-physiological human-derived platform suitable for mechanistic studies and drug screening but lack systemic features such as vasculature and immune components. Animal models provide physiological and systemic context but are limited by species differences. Clinical samples represent the highest physiological and pathological fidelity yet are constrained by sample accessibility and individual variability. Together, this comparison highlights the complementary roles of each model in advancing translational research, with organoids bridging the gap between conventional cell cultures and in vivo systems. (This figure was created with BioRender software. https://biorender.com/).

Although respiratory organoid platforms have greatly contributed to exploring SARS-CoV-2 pathogenesis and drug development, they still have limitations. The human body features organ interconnections, and a key factor preventing organoids from replacing animal models is the lack of immune and vascular system interventions and organ communication (Li M. et al., 2023). Additionally, organoid culture presents technical challenges, such as varying media compositions, uneven sizes, and different differentiation degrees, which hinder the uniform and standardized application of disease modeling (Peng et al., 2022). In the future, the organoid platform will face more challenges. The combined use of different organoids or more complex organoids will help explore disease mechanisms and offer more possibilities for personalized treatment of future epidemic viruses.

This review summarizes the use of respiratory organoids in studying SARS-CoV-2 and other respiratory viral infections; however, these models have also been widely applied to investigate bacterial, parasitic, and fungal respiratory pathogens (Blutt and Estes, 2022; Heo et al., 2018; Kim et al., 2024; Tisdale-Macioce et al., 2021). Looking forward, organoid-on-a-chip systems integrating immune cells, vascular networks, and neural components will better simulate immune responses and neurovascular injury during pulmonary infection, providing a more physiologically relevant microenvironment (Verstegen et al., 2025; Surina et al., 2025). Furthermore, while current organoid phenotyping often relies on manual processing, the integration of artificial intelligence, high-content imaging, and machine learning is expected to enable ultra-high-throughput drug screening based on respiratory organoids (Surina et al., 2025; Ma et al., 2025). Despite their considerable potential, respiratory organoids still face several challenges. Even with incorporation of immune and vascular components, the models lack systemic factors such as hormonal signals, metabolic feedback, and organism-level regulation, which may limit their ability to fully recapitulate in vivo pathophysiology. In addition, the complexity of organoid and organoid-chip construction, combined with substantial batch-to-batch variation, poses challenges for reproducibility and large-scale standardization (Surina et al., 2025). Therefore, establishing standardized characterization protocols, consensus construction guidelines, and clear regulatory frameworks will be essential to promote the broader application of these platforms (Wang et al., 2025). With continued technological advances and deeper mechanistic insights, organoid-based in vitro models are poised to significantly accelerate the development of personalized therapies for respiratory infectious diseases.

Conflict of interest

The authors declare no conflict of interest.