Coronavirus disease 2019 (COVID-19), which is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has rapidly reshaped global scientific priorities by exposing fundamental limitations in our ability to model human respiratory infection [1]. SARS-CoV-2 is an enveloped, positive-sense RNA virus belonging to the Coronaviridae family and shares phylogenetic similarity with severe acute respiratory syndrome coronavirus and Middle East respiratory syndrome coronavirus [2,3]. Despite these genetic and structural similarities, SARS-CoV-2 is distinguished by its exceptionally high transmissibility and its capacity to cause widespread lower respiratory tract disease, with clinical manifestations ranging from mild upper-airway infection to viral pneumonia, acute respiratory distress syndrome, and systemic organ dysfunction [4]. This broad clinical spectrum reflects not only intrinsic viral pathogenic determinants but also marked heterogeneity in host responses shaped by age, comorbid conditions, genetic background, and immune status [5]. The SARS-CoV-2 life cycle is initiated when the viral spike (S) glycoprotein binds angiotensin-converting enzyme 2 (ACE2), which is expressed on airway and alveolar epithelial cells (Fig. 1) [6,7]. Following receptor engagement, the S protein requires proteolytic activation to trigger membrane fusion and viral entry [8,9]. At the plasma membrane, transmembrane serine protease 2 (TMPRSS2) cleaves the S protein and facilitates rapid fusion at the cell surface. In contrast, in cells with low TMPRSS2 expression, viral particles are internalized into endosomal compartments, where cathepsins activate the S protein under acidic conditions to enable endosomal fusion [10,11]. Once released into the cytoplasm, the approximately 30-kb positive-sense viral RNA genome, which bears a 5′ cap and a 3′ poly(A) tail, undergoes cap-dependent translation to generate the ORF1a and ORF1ab polyproteins. These polyproteins are subsequently cleaved into nonstructural proteins that assemble into the replication–transcription complex (RTC) [12–14]. The RTC remodels the endoplasmic reticulum (ER) into double-membrane vesicles and interconnected membranous webs that provide a protected microenvironment for viral RNA synthesis [15]. Within these replication compartments, the RTC produces both full-length genomic RNA and a nested set of subgenomic RNAs through a process of discontinuous transcription [16,17]. The subgenomic RNAs serve as templates for the translation of structural proteins, including spike (S), envelope (E), membrane (M), and nucleocapsid (N), as well as accessory proteins required for virion assembly within the ER–Golgi intermediate compartment [17]. Viral replication and the accompanying host responses induce epithelial injury, metabolic reprogramming, and sustained activation of pro-inflammatory signaling pathways. Together, these processes drive the pulmonary pathology and systemic inflammatory manifestations characteristic of COVID-19 [18].

To elucidate SARS-CoV-2 pathogenesis and accelerate therapeutic development, researchers have traditionally relied on in vitro infection systems and animal models to define viral entry mechanisms, replication kinetics, and host responses [19,20]. Two-dimensional (2D) cell cultures, including Vero E6, Calu-3, and A549 cells engineered to overexpress human ACE2, have served as essential early platforms for quantifying viral propagation, dissecting host factor requirements, and enabling high-throughput screening of antiviral candidates [20]. However, these monolayer systems lack key attributes of the human respiratory epithelium, including multicellular organization, apical–basal polarity, mucus secretion, and specialized epithelial subsets such as ciliated cells and alveolar type II (AT2) cells [21]. Consequently, 2D cultures fail to adequately reproduce cell type–specific infection patterns, interferon signaling dynamics, and the tissue-level injury characteristic of COVID-19 [22]. Given these limitations, animal models have traditionally been employed to examine SARS-CoV-2 infection, transmission, and systemic immune responses in vivo [20,23]. Transgenic mice expressing human ACE2, Syrian hamsters, ferrets, and non-human primates have each contributed important mechanistic insights into viral pathogenicity, upper and lower airway tropism, and vaccine-induced protection [23]. Nonetheless, species-dependent differences in ACE2 receptor compatibility, TMPRSS2 expression, innate and adaptive immune responses, alveolar architecture, and mucociliary physiology substantially alter susceptibility and disease manifestations relative to humans [20,23]. Moreover, viral adaptation strategies required to establish infection in certain species, such as mouse-adapted viral strains or ectopic ACE2 overexpression, may bypass natural entry routes and confound interpretation of pathogenesis studies [24]. Practical, ethical, and cost-related constraints further limit the scalability and rapid deployment of animal models during emerging outbreaks [20]. Collectively, these limitations highlight the need for human-specific, physiologically relevant systems capable of modeling SARS-CoV-2 infection within a more faithful respiratory tissue context.

In response to these challenges, organoid technology has emerged as a powerful and physiologically relevant alternative for modeling respiratory viral infections [25]. Organoids are 3-dimensional (3D) self-organizing structures derived from pluripotent or adult stem cells that recapitulate key features of native tissues, including cellular heterogeneity, epithelial polarity, and context-dependent signaling networks [26]. Human lung organoids generated from induced pluripotent stem cells (iPSCs) or primary airway and alveolar progenitors comprise the major epithelial populations naturally targeted by SARS-CoV-2, including ciliated cells, goblet cells, club cells, basal stem cells, and AT2 cells [23]. Relative to 2D monolayer cultures, lung organoids more accurately reproduce epithelial differentiation states, spatial organization, and multicellular interactions that shape viral entry routes, replication dynamics, and innate immune responses [22,23,27]. Unlike animal models, they also circumvent species-specific differences in ACE2 compatibility, immune regulation, and airway physiology, thereby enabling direct interrogation of human-specific infection processes [27,28]. As a result, lung organoid systems have yielded critical insights into viral entry mechanisms, replication competence across epithelial subsets, innate immune activation, and variant-dependent shifts in epithelial tropism [27,28]. Furthermore, their genetic tractability, scalability, and compatibility with patient-derived tissues position lung organoids as versatile platforms for antiviral drug discovery and personalized disease modeling [29,30].

In this review, we synthesize current lung organoid platforms used to study SARS-CoV-2 infection, critically evaluate their advantages over traditional experimental models, and examine the conceptual and technical limitations that must be addressed to further enhance physiological relevance. We also highlight emerging strategies, including immune cell co-culture, vascular and mesenchymal integration, mechanical stimulation, and microphysiological organ-on-chip approaches, that hold promise for expanding the translational utility of lung organoids in the study of COVID-19 and future emerging respiratory pathogens.

Ethics statement: This study was a literature review of previously published studies and was therefore exempt from institutional review board approval.

Building on the limitations of conventional in vitro and in vivo models, diverse lung organoid systems have been developed to model SARS-CoV-2 infection with substantially greater human-specific fidelity [31]. Unlike 2D monolayer cultures or animal models, lung organoids comprise multiple epithelial lineages, exhibit defined apical–basal polarity, and preserve tissue-specific transcriptional programs that collectively shape viral tropism, innate immune activation, and cell type–specific susceptibility (Fig. 2) [28,32]. These properties enable lung organoids to model viral entry, replication, and host responses within an architectural and molecular context that more closely approximates the human respiratory epithelium. Human lung organoids can be generated from 2 principal stem cell sources: iPSCs and adult stem or progenitor cells isolated from airway or alveolar tissues (Table 1) [31]. Although both approaches rely on self-organization within a 3D extracellular matrix, they differ fundamentally in developmental trajectory, epithelial maturity, scalability, and suitability for modeling inter-individual variation. iPSC-derived lung organoids recapitulate early developmental programs and give rise to airway- or alveolar-like epithelial structures, whereas adult stem cell–derived organoids retain donor-specific genetic and epigenetic features and more faithfully reflect mature epithelial phenotypes [33–35]. Together, these complementary platforms enable investigation of SARS-CoV-2 infection biology across developmental states, with iPSC-derived systems providing insight into developmental susceptibility and adult stem cell–derived models capturing features of mature lung infection.

Despite their transformative impact on experimental modeling of SARS-CoV-2 infection, lung organoids remain constrained by several biological and technical limitations that prevent full recapitulation of human lung physiology [26,31,76]. A central limitation is their inability to reconstruct the complex multicellular microenvironment of the respiratory system. Although current lung organoids faithfully reproduce many epithelial features, they lack immune, vascular, and biomechanical components that are critical determinants of lung homeostasis and disease [31,77,78]. The absence of resident immune populations, including macrophages, dendritic cells, and lymphocytes, restricts investigation of epithelial–immune crosstalk, interferon-driven antiviral programs, and cytokine-mediated tissue injury that are central to COVID-19 pathogenesis. Likewise, the lack of an endothelial interface precludes accurate modeling of oxygen and nutrient exchange, inflammatory cell trafficking, endothelial dysfunction, and microvascular thrombosis, all of which are hallmark features of severe SARS-CoV-2 infection. In addition to missing cellular components, lung organoids are typically maintained under static culture conditions and do not experience the cyclic mechanical stretch generated by breathing [26,31]. Mechanical forces play a critical role in regulating AT2 maturation, surfactant homeostasis, ion transport, and epithelial repair responses, and their absence contributes to persistent immaturity, particularly in iPSC-derived systems. In response to these limitations, immune-integrated organoid platforms have begun to emerge. Recent macrophage–AT2 assembloid models and innate immune co-culture approaches partially restore epithelial–immune interactions that shape SARS-CoV-2 entry, interferon signaling, cytokine induction, and epithelial injury [55]. These systems more closely approximate early antiviral responses and provide an improved framework for investigating immune-driven disease heterogeneity.

A second major constraint relates to epithelial maturation and developmental fidelity. iPSC-derived lung organoids frequently retain fetal-like transcriptional and functional profiles that can alter epithelial responses to SARS-CoV-2, whereas adult stem cell–derived organoids mitigate some of these developmental limitations [33,37,39,79,80]. However, adult stem cell–derived systems introduce distinct challenges, including donor-to-donor variability and limited long-term expansion capacity [81,82]. Differences in age, smoking history, chronic lung disease, and genetic background influence epithelial differentiation states and viral susceptibility, complicating cross-study comparisons and standardization efforts [41,63,83,84]. Across both platforms, batch-to-batch variability in organoid size, cellular composition, and maturation state further contributes to experimental heterogeneity. Technical and scalability challenges also remain substantial [85,86]. Organoid cultures are labor-intensive, rely on specialized extracellular matrices, and are less amenable to high-throughput screening than traditional 2D monolayer systems. Moreover, variability in differentiation protocols, culture media formulations, and matrix composition can generate pronounced differences across laboratories, underscoring the need for standardized workflows, benchmarking criteria, and well-characterized organoid biobanks [85–88].

Future innovations are expected to progressively address these limitations by incorporating missing biological components and enhancing structural and functional maturation [37,78]. Strategies aimed at vascularization, including endothelial co-culture, microfluidic perfusion, and organ-on-chip platforms, hold promise for restoring nutrient delivery, mechanotransduction, and endothelial–epithelial interactions [89–91]. Mechanical conditioning through bioreactors or engineered scaffolds may further promote epithelial maturation and more closely replicate alveolar biomechanics [92–94]. In parallel, efforts to standardize differentiation protocols, improve reproducibility, and develop miniaturized or automated organoid systems will expand the scalability and experimental robustness of organoid-based infection studies [26,87]. As these approaches converge, lung organoids are expected to increasingly approximate adult lung physiology and to serve as powerful platforms for studying SARS-CoV-2 and future respiratory pathogens [29,73,95]. With continued technological refinement, lung organoids are poised to assume an increasingly central role in respiratory virus research.

Human lung organoids have emerged as essential platforms for modeling SARS-CoV-2 infection, bridging the gap between reductionist in vitro systems and the complex physiology of the human respiratory tract (Fig. 2) [29]. Both iPSC-derived and adult stem cell–derived lung organoids capture critical aspects of airway and alveolar epithelial biology, enabling detailed interrogation of viral entry mechanisms, replication dynamics, innate immune responses, and variant-dependent shifts in epithelial tropism [30]. By retaining developmentally immature or adult-like transcriptional states, reproducing epithelial polarity, and expressing endogenous viral entry factors, these systems permit mechanistic analyses that more closely approximate in vivo human infection.

Nonetheless, current lung organoid platforms remain limited by incomplete incorporation of immune, vascular, and mechanical components, as well as by challenges related to epithelial maturation, donor variability, and scalability [29,31,37,41]. Ongoing efforts to integrate immune cells, endothelial interfaces, and biomechanical cues, together with initiatives to standardize culture protocols, promise to substantially enhance the physiological fidelity and reproducibility of organoid-based models [76,78]. As these innovations intersect with advances in single-cell profiling, CRISPR-based genome engineering, and microphysiological technologies, lung organoids are positioned to play an increasingly central role not only in elucidating SARS-CoV-2 pathogenesis but also in preparing for future emerging respiratory viruses [29,37,53]. By enabling mechanistic, translational, and personalized approaches to infection research, lung organoid systems will continue to expand our understanding of human respiratory biology and accelerate the development of effective antiviral strategies.