Advanced models for respiratory disease and drug studies

Abstract The global burden of respiratory diseases is enormous, with many millions of people suffering and dying prematurely every year. The global COVID‐19 pandemic witnessed recently, along with increased air pollution and wildfire events, increases the urgency of identifying the most effective therapeutic measures to combat these diseases even further. Despite increasing expenditure and extensive collaborative efforts to identify and develop the most effective and safe treatments, the failure rates of drugs evaluated in human clinical trials are high. To reverse these trends and minimize the cost of drug development, ineffective drug candidates must be eliminated as early as possible by employing new, efficient, and accurate preclinical screening approaches. Animal models have been the mainstay of pulmonary research as they recapitulate the complex physiological processes, Multiorgan interplay, disease phenotypes of disease, and the pharmacokinetic behavior of drugs. Recently, the use of advanced culture technologies such as organoids and lung‐on‐a‐chip models has gained increasing attention because of their potential to reproduce human diseased states and physiology, with clinically relevant responses to drugs and toxins. This review provides an overview of different animal models for studying respiratory diseases and evaluating drugs. We also highlight recent progress in cell culture technologies to advance integrated models and discuss current challenges and present future perspectives.


| INTRODUCTION
The morbidity and mortality related to respiratory diseases impose substantial socioeconomic burdens on individuals and societies worldwide. 1 Five respiratory conditions predominantly contribute to the global health burden; chronic obstructive pulmonary disease (COPD), asthma, tuberculosis, lung cancer, and acute respiratory infections. 2,3 The common risk factors for chronic respiratory diseases include exposure to indoor and outdoor pollutants, tobacco use, allergens, occupational exposure, obesity, physical inactivity, and an unhealthy diet. 4 Increased exposure to these risk factors and the aging of populations have caused substantial increases in the prevalence and disease burden of respiratory diseases worldwide. The ongoing global coronavirus disease 2019  pandemic, caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), is still a major global crisis with over 317 million confirmed cases and 5.51 million deaths. 5 Disastrous bushfires and wildfires worldwide have also increased exposure to hazardous levels of air pollution with deleterious respiratory impacts. 6 With a dire need to address the growing global burden of respiratory diseases, rapid and efficient preclinical evaluation models are urgently required to fast-track the development of safe and efficacious new therapeutics and vaccines. 7,8 Some of these therapeutics target specific components of pulmonary disease such as inflammasomes, 9 interleukins (ILs), 10 matrix metalloproteinases (MMPs), 11 and mitochondrial dysfunction. 12 Drug development is an extensive, intricate, and expensive process, and has high attrition rates in clinical trials. 13 The average cost of research and development for taking a new product to the clinic is estimated to be $2.5 billion and takes up to 15 years. 14 Despite immense collaborative efforts and increased expenses to identify and develop the most effective and safe drugs, few are approved for clinical trials. 7 Compared to cardiovascular, neurological, and other diseases, the drugs approved for use in pulmonary medicine are few, and there are fewer drug candidates to address their growing healthcare burden. 15 To streamline drug development and limit costs, it is essential to eliminate ineffective drug candidates as early as possible, which can be achieved by utilizing new preclinical screening approaches with high efficiency and accuracy. 14 SHRESTHA ET AL.

| 1471
Laboratory-based preclinical and ex vivo, and in vitro drug models that use different cell types and cellbased assays are commonly employed by the pharmaceutical industry and research laboratories to assess drug candidates. 16,17 Although these models are simple to use, can be well controlled, and offer high throughput for basic screening and testing of drugs, they fail to mimic the in vivo tissue architecture, physiology, and cellular interactions in human cells and tissues and are unable to predict complex drug metabolism processes and adverse reactions. 18 Thus, advanced human-relevant models that mimic the human lung microenvironment along with tissue-tissue interfaces, chemical gradients, and mechanical factors are urgently required to better model respiratory diseases and expedite pulmonary preclinical research.
Here, we overview the existing ex vivo and in vitro platforms that are used to study human lung pathophysiology and host-respiratory pathogen interactions ( Figure 1). We summarize recent advances in technologies and cell culture platforms that replicate the in vivo environment to provide physiologically relevant clinical data. Additionally, we discuss the applications of these models in different pathologies and their use for screening drugs to alleviate the global burden of respiratory diseases.
F I G U R E 1 Cell culture models of human lung: (A) Commonly used cell culture models to study respiratory diseases. (B) Comparison of features of existing cell culture models. 2D, and 3D cell culture models, and animal models are extensively used to preclinically assess drugs in the drug development process. Recently, organoids and microfluidic organ-on-a-chip models have gained significant attention owing to their 3D physiologically relevant environment. All these models have their advantages and limitations. Models that mimic complex human physiology with clinically relevant and reproducible responses are required to identify the most effective and safe drugs. (Note: Animal models can incorporate human cells, e.g., human-derived xenografts.). [Color figure can be viewed at wileyonlinelibrary.com] T A B L E 1 Experimental models of respiratory disease.

Disease Inducer/mouse model Features References
Asthma OVA-induced in BALB/c mice Eosinophilia, increased eosinophil peroxidase, IL-4, IL-13, nitric oxide, mucus hypersecretion, AHR, airway remodeling with chronic exposure 21,29 HDM induced in BALB/c mice Severe and persistent lung inflammation with the accumulation of CD4 + lymphocytes and overproduction of helper T cell Type 2related cytokines by splenocytes, airway remodeling with chronic exposure 25,30 Cockroach allergen-induced Abr-/mice with FVBJ inbred background Infiltration of leukocytes, increased eosinophil in BALF, increased serum IgE, mucus hypersecretion, impaired lung function with decreased pulmonary compliance, and increased airway resistance 31 Fungus (Alternaria alternata) induced in BALB/c mice This includes predominant eosinophilic Th2 responses in wild-type mice and neutrophilic in Il4, Il13, and Stat6 -/mice. The neutrophilic responses in Stat6 -/mice characterized by increased lung levels of CXCL1, TNF-α, CXCL2, and CXCL5 32 Ambient air PM induced in BALB/ c mice Upregulation of mRNA expression of TNF-α in lung tissue 33 Diesel exhaust particle-induced in BALB/c mice Increased eosinophil, IL-5, IL-17F, and CCL20, decreased levels of IFN-γ in BALF and AHR 33 COPD CS-induced in BALC/c and C57BL/ 6 mice Chronic airway and lung inflammation with increased macrophages and neutrophils in BALF, expression of TNF-α, CxCl2, KC, and IL-1B in the lung, airway remodeling, and fibrosis, alveolar enlargement/emphysema, impaired lung function, and gas exchange 22,34 Porcine pancreatic elastase-induced in BALB/c and C57BL/6 mice Increase neutrophils and macrophage and TNFα, IL-6, KC, GCSF, MCP-1 in BALF 35,36 Porcine pancreatic elastase-induced in BALB/c mice  These include increased airway and neutrophil and macrophage counts and   cytokines (TNF-α, IL-6, KC, granulocyte-macrophage colony-stimulating factor (GCSF), monocyte chemoattractant protein-1 (MCP-1), were significantly increased on Day 3. In addition, there were increased mRNA levels of TNF-α, IL-6, GCSF, macrophage inflammatory protein-2 (MIP-2), KC, and protein expression of nuclear factor kappa-lightchain-enhancer of activated B cells (NF-κB) in lung tissue and the destruction of airway spaces (a feature of emphysema). 35 Another study observed that elastase (1U/mouse, intratracheal instillation, followed by 1-3 weeks rest) increased the mRNA expression of MMP-2 and MMP-9 key enzymes involved in the cleavage of extracellular matrix (ECM). 36 Lipopolysaccharide (LPS) is an endotoxin of Gram-negative bacteria. 73 LPS can be detected in the BALF of patients with COPD, which has led to speculation of roles for it in the progression of the disease. 74,75 It can be administered alone or with other challenges such as cigarette smoke to induce COPD features in mice. 76 LPS administration induces hallmark features of airway inflammation such as excessive mucus production, narrowing of the bronchial lumen, and chemotaxis of neutrophils and macrophages to airway spaces. 77,78 LPS is administered intratracheally into mice at a dose of 5 μg/installation/mouse 77 or 0.1 ml of a 100 mg/ml LPS solution in normal saline for rats. 78 This dose is equivalent to LPS entering the human lung after smoking approximately 25 cigarettes. 79 In guinea pigs and mice, the chronic damage in the lungs induced by LPS is associated with the overexpression of TNF-α and IL-18. 37 Human exposure to biomass smoke for a prolonged period is another cause of chronic respiratory disease. 80 Compared to unexposed healthy people, those who are exposed have an odds ratio of 2.44 for developing COPD. 81 Thus, another animal model of COPD is in vivo exposure to varying doses and periods of biomass smoke. Rabbits exposed to biomass smoke by burning 80 g of biomass-dried dung 1 h/day for a month had increased proliferation of respiratory epithelial cells and significant alveolar destruction compared to air-exposed controls. 39 Other animals such as C57BL/6 mice and Sprague-Dawley rats are exposed to biomass smoke to induce COPD. 82 This suggests that people cooking indoors with dung in low-and middle-income countries may develop COPD features after repeated exposure. 83 Although animal models are widely used to model COPD, in contrast to the variable pathology and different stages of COPD severity in humans, the existing animal models are confined to mimic only a few (but not all) characteristic features of human COPD which must be taken into account when investigating the clinical application of these models. Therefore, the assessment of animal models and the interpretations of results need to be based on the alignment of the characteristic features of human COPD to advance the understanding of disease pathogenesis. 84 85 A/J and Swiss mice are genetically susceptible to lung cancer when exposed to whole-body cigarette smoke (reference cigarette 2R4F) for 5 months with a 4-month rest period.

| Lung cancer
There were significant increases in the incidence of lung tumors and tumor multiplicity in mice exposed to cigarette smoke. 40 The most potent carcinogens in cigarette smoke are polycyclic aromatic hydrocarbons, tobacco-specific nitrosamine, and Benzo [a] pyrene. 86 These carcinogens induce different types of lung cancer. Nicotine-derived nitrosamine ketone (NNK) induces adenocarcinoma, 87 whereas, N-nitroso-tris-chloroethylurea promotes the development of squamous cell carcinoma. 88 A single high dose (15 μg/g body weight; intraperitoneally) of SHRESTHA ET AL. | 1477 diethylnitrosamine followed by 1-year rest is enough to induce lung cancer in 70% of FVB/N mice. 45 The primary tumors express the cytokeratin-7 characteristic of epithelial neoplasms and thyroid transcription factor-1 specific to the thyroid and lung adenocarcinoma. No mutations were observed in KRAS and epidermal growth factor receptor (EGFR) genes while mTOR was highly activated. 45 Several transgenic mouse models develop features that resemble those of human lung cancers. Among them are mice deficient in tumor suppressor genes such as tumor suppressor protein 53 (Trp53), mutated in colorectal carcinoma protein (Mcc), retinoblastoma protein (Rb), phosphatase and tensin homolog (PTEN), and RNA-binding protein 5 (Rbm5) or mice with mutation of proto-oncogenes such as Kirsten rat sarcoma virus (Kras), tyrosine kinase receptors (Trk), extracellular signal-regulated kinase (Erk), Wnt and Myc. 85,89 Ferrets are also used to model lung cancer. Intraperitoneal injection of NNK in ferrets (50 mg/kg body weight) once a month for 4 months, followed by postexposure rest periods of 24, 26, and 32 weeks led to the development of both preneoplastic and neoplastic lesions in the lungs. 41 The preneoplastic lesions were characterized by squamous metaplasia and dysplasia, and atypical adenomatous hyperplasia. The major discovery in this study was the high expression of nicotinic acetylcholine receptors on broncoepithelial cell membranes. 41 Urethane is another potent tobacco carcinogen used in mice models to induce lung adenocarcinoma. A recent study of C57BL/6 mice injected with 1 g/kg urethane intraperitoneally for 10 consecutive weeks followed by 14 weeks of rest showed the development of lung adenocarcinoma. This study revealed that airway epithelial cells were more susceptible to urethane-induced KrasQ61R mutations than alveolar type II cells, suggesting that tobacco-induced lung adenocarcinoma is initiated in airway epithelial cells. 90 Collectively, these studies show that mouse models can be used to investigate genetic mechanisms and discover drugs for lung cancer. 91

| Lung fibrosis
Lung fibrosis is the primary feature of idiopathic pulmonary fibrosis (IPF) and an important feature of asthma and COPD. 92 Bleomycin is commonly used to induce pulmonary fibrosis in mouse models. Other inducers of lung fibrosis are HDM extract, silica, cigarette smoke, and asbestos. Although mice are preferred, other animals such as rats, hamsters, rabbits, guinea pigs, dogs, and primates are also used in lung fibrosis models. 93,94 Administration of a single dose of bleomycin (intranasal with 0.05U/mouse) with or without 4 weeks of rest in one mouse model and treatment with HDM extract (intranasal with 25 μg in 30 μl of sterile saline) for 5 days/week for 5 weeks in another resulted in significant airway inflammation and collagen deposition around the airways. 46 The airway inflammation and remodeling in both bleomycin and HDM-treated mice were dependent on the ECM protein fibulin-1c activation of TGF-β. 25,46 Knockout of the Fbln1c gene and targeted inhibition of Fbln1c could decrease collagen deposition in the airway and protect against AHR. 25 Administration of asbestosis to animals can replicate the pathologic changes that occur in human pulmonary fibrosis in asbestosis. For animal models, asbestos should only be used after baking to avoid contamination with LPS and to avoid LPS-related inflammatory changes. 95 A single dose of asbestos fibers delivered via intratracheal administration can induce pulmonary fibrosis, although it may take more than a month to develop the characteristic features, such as collagen deposition, oxidative stress, fibroblastic foci, and alveolar epithelial cell injury. 96 However, such models have limitations; fibrosis is mostly central but not subpleural, and collagen deposition is unequally distributed around the lungs. Pulmonary fibrosis due to the deposition of asbestos fibers is due to apoptosis of alveolar epithelial cells followed by the polarization of macrophages to M2 phenotype and increased production of pro-fibrotic cytokines by activated T lymphocytes. 97,98 The chronic instillation of silica into rat lungs results in the formation of fibrotic nodules, which are similar to those in the lungs of exposed humans. 99 In mice, the nature of the experimental silicosis that develops depends on the dose of silica, route, length of exposure, and mouse strain. 100 Silica can be administered in mice via inhalational/ aerosolization, 100 oropharyngeal aspiration, 47 or intratracheal delivery. 101 Comparatively, C3H/HeN, MRL/MpJ, and NZB mice are more prone to developing aerosolized silica-induced fibrosis than BALB/c mice which have little fibrotic response. 100 In intratracheal silica administration, the C57BL/6 strain is more susceptible than CBA/J mice. 102 After administration, silica accumulates in the lung and induces persistent inflammatory responses followed by the development of fibrotic nodules around silica deposits. Intratracheal models take 2-4 weeks to develop, 47 whereas inhalation induces disease that closely resembles human silicosis but takes 1-6 months. 101

| COVID-19
SARS-CoV-2 is the causative agent of COVID-19. It causes severe viral infection resulting in excessive inflammatory responses driven by a pro-inflammatory cytokine storm syndrome, which results in extensive airway inflammation. 50,103 The ongoing COVID-19 pandemic has driven the rapid discovery of effective vaccines and drugs. To develop and test novel therapeutics, preclinical animal models are essential to facilitate the translational investigation of candidate drugs and vaccines and expedite human implementation. 49,104 Various small animal models with mice, ferrets, and golden hamsters are commonly used to study COVID-19 pathophysiology and identify the most effective therapeutics. 20 However, the most critical aspect to consider is selecting an appropriate animal model where the SARS-CoV-2 Spike (S) protein binds to the viral entry receptor, Angiotensin-converting enzyme 2 (ACE2).
Wild-type mice are not permissive to the ancestral SARS-CoV-2 infection due to incompatibility between the S protein and mouse ACE2 (mACE2). Consequently, several genetic strategies have been implemented to overcome this barrier to produce mouse models that are an excellent preclinical platform for COVID-19 research. The most common approach utilizes transgenic mice expressing human ACE2 (hACE2) under the control of the human cytokeratin promotor (K18), which is localized exclusively in epithelial cells of the lung, gastrointestinal tract, liver, kidney, heart, and brain. 105 This mouse line, K18-hACE2, was originally developed to study the pathophysiology of SARS-CoV, responsible for the first SARS epidemic in 2003. However, as both SARS-CoV and SARS-CoV-2 enter host cells via binding of the S protein to hACE2, K18-hACE2 mice are highly susceptible to infection with both with significant weight loss, extensive lung inflammation, high viral lung titers, and uniform lethality with relatively low viral inocula. 106 Interestingly, K18-hACE2 mice also develop extensive neurological inflammation with high viral titers during the late stages of SARS-CoV-2 infection. However, the mechanisms of how this occurs are unknown, which is a caveat with this severe mouse model. 107  Therapeutic administration of Remdesivir and Plitidepsin, both antiviral inhibitors of SARS-CoV-2 replication, was protective in reducing lung viral titers and lung inflammation in K18-hACE2 mice. 108 Additionally, monoclonal antibody therapy at both the early and late stages of SARS-CoV-2 infection rescued mice from lethal infection. 109,110 Alternative approaches involve the use of adenovirus (AdV) or adeno-associated virus (AAV) transduction systems to transiently induce expression of hACE2 in recipient cells or following intranasal administration to mice to induce SARS-CoV-2 infectivity in the absence of endogenous hACE2. 110,111 AdV and AAV systems are advantageous as they can utilize readily available genetically modified mouse lines without the need to backcross to K18-hACE2 mice. However, both AdV and AAV SARS-CoV-2 mouse models result in only mild weight loss, clinical presentation, and lung inflammation when challenged with high viral inocula in wild-type; however, they both support high viral replication during the early stages of infection when hACE2 is abundantly expressed. 110,111 SHRESTHA ET AL.
| 1479 While the use of AdV and AAV mouse models is useful in understanding the genetic determinants of COVID-19, they are relatively ineffective in testing novel therapeutic agents and vaccines.
Another approach is to mouse adapt SARS-CoV-2. This involves the continual passage of highly virulent and human-specific SARS-CoV-2 through mice, followed by the recovery of infectious virus to repeat the process and facilitate adaptation of SARS-CoV-2 to the mouse host. This has resulted in mouse-adapted SARS-CoV-2 variants that cause significant weight loss, clinical scores, and severe lung inflammation, which recapitulates wild-type SARS-CoV-2 in K18-hACE2 mice. 51,52 However, the continuous passage of SARS-CoV-2 through mice to induce evolutionary pressure on the virus to mutate and adapt to the host may cause unknown and potentially deleterious genetic changes, which may deviate from the original SARS-CoV-2 variant and question the relevance of the mouse-adapted SARS-CoV-2 model. Interestingly, mouse-adapted SARS-CoV-2 models have been used to test the efficacy of preclinical vaccine candidates and the endogenous administration of IFN-λ, both of which reduce viral replication and protect mice from severe disease. 51 Other animal models that are employed to investigate the pathophysiology of SARS-CoV-2 include ferrets and hamsters, although these are less commonly used compared to traditional mouse models. Both ferret and hamster ACE2 have a high degree of homology with hACE2, thus enabling SARS-CoV-2 infection in these animals. 112 Finally, nonhuman primates (NHP) are considered the "gold standard" as preclinical animal models for understanding the pathophysiology of COVID-19. However, they are much less commonly used than smaller animal Although animal models are widely used for respiratory disease-specific investigations, preclinical drug development, and testing, they also come with several limitations. In some cases, they do not reproduce the structural, mechanical, and functional properties of the human tissue and thus fail to mimic the inherently complex nature of tissues and organs. Moreover, promising discoveries and treatments in animal models do not necessarily prove to have favorable outcomes for humans. 115 Thus, advanced models like organs-on-chip are required to decrease the need for both traditional 2D cell culture methods and animal studies. These advanced models are described in detail in the following section.

| LUNG BIOPSIES AND TISSUES IN RESPIRATORY RESEARCH
Animal models of different lung diseases are valuable in increasing our understanding of lung development, pathophysiology, and diseases; the inherent differences between animal and human lung physiology make it imperative to develop human lung model systems. These models complement and are complemented by in vivo animal models. Clinical biopsies of the airways or lungs are collected and cells or tissues are extracted for examination. 62 Bronchoscopies are performed to collect airway brushings for epithelial culture, and airways pinch biopsies can also be obtained. 65,116 Primary airway epithelial cells are cultured at the air-liquid interface (ALI) and differentiate into differentiated columnar epithelium with the basal, club, goblet, and ciliated cells. These cultures are valuable in pathogenesis, drug, and infection (e.g., influenza, respiratory syncytial virus, SARS-CoV-2) studies. In asthmatic individuals, airway biopsies are performed to study different phenotypes of severe asthma to direct treatments.
Lung biopsies are performed to ascertain an accurate diagnosis of the onset of interstitial lung disease or confirm the absence/presence of cancer. Various lung biopsy procedures include needle, thoracoscopic, transbronchial, and open biopsy. 117,118 Many studies have utilized airway and lung tissues obtained by a biopsy to study the pathogenesis and effects of treatments in the context of asthma, COPD, lung fibrosis, and cancer. 92,[119][120][121] Airway and lung biopsies contain all of the cell and tissue compartments affected by respiratory diseases, including the airways and the lung parenchyma. The samples obtained by biopsy can be snap frozen (to study the gene expression and proteins levels of different markers of asthma, and also epithelial cells can be resurrected from them for ALI culture), paraffin-embedded (to assess histological changes like goblet cell hyperplasia/metaplasia and immune cell infiltration in distal and central airways), and processed for high magnification imaging (scanning electron microscopy [SEM] and transmission electron microscopy [TEM]). 122 Biopsies are also helpful in diagnosing respiratory conditions. They can be used to differentiate severe asthma from allergic bronchopulmonary aspergillosis, autoimmune airway disease, and hypersensitivity pneumonitis. 119 However, since asthma is primarily an airway disease, BAL and sputum are more widely used than bronchial biopsies. 123 In COPD patients, inflammation occurs primarily in the small airways and lung parenchyma and all airway lung samples are important. 124 Indeed COPD-linked structural changes and expression of inflammatory proteins can be studied utilizing bronchial and lung biopsies and their use provides insights into COPD pathogenesis. 125 Stefan et al. reported increased NF-κB expression in bronchial biopsies from smokers and patients with COPD. NF-κB is a redox-sensitive transcription factor that regulates multiple pro-inflammatory pathways. 126 Additionally, decreased histonedeacetylase activity has been reported in bronchial biopsies that explain the limited effectiveness of steroids in individuals with COPD. 127 Many studies have used lung biopsies and tissues to study COPD, such as defining the roles of microRNAs in driving pathogenesis through a SATB1/S100A9/NF-κB pathway, and the role of necroptosis. 65,128 IPF is a chronic, progressive, and fibrotic lung disease where healthy lung tissue is permeated with excessive ECM, resulting in reduced lung function. 92 Transbronchial lung biopsy and lung cryobiopsy are used to diagnose IPF and to study disease pathogenesis and progression. 129 Although airway and lung biopsies offer an accurate and detailed method to study the pathogenesis and effects of treatments in different lung diseases, there are multiple challenges associated with this approach. Obtaining enough tissues to perform multiple analyses is challenging. Also, the patients that undergo lung biopsy may have complications of hemoptysis, pulmonary venous air embolism, pulmonary hemorrhage, and pneumothorax. 130,131 Thus, optimizing the use of airway and lung biopsies for studying different respiratory diseases is needed.
Resected lung tissues from severe COPD, IPF, and lung cancer patients can be used for the interrogation, manipulation, or extraction of cells for further analysis (Table 2). 65,128 They are valuable in defining pathogenic mechanisms and testing therapies compared to healthy control tissues from failed lung transplants. Tissues from the lungs as far as possible from lung tumors can also be used for healthy or COPD studies depending on the diagnosis of the donors. These tissues have some obvious confounding factors.

| ORGANOIDS
Organoids, or miniaturized organs, are intricate multicellular clusters produced during in vitro culture of stem cells related to a specific organ. 145 Stem cells start to form organoids via innate developmental self-assembly and form complex tissue-like structures and simplified versions of organs in 3D microenvironments. 146 The recent progress in stem cell manipulation techniques has led to the development of in vitro culture systems mimicking the native microstructure of organs to produce functional and relevant cellular complexes. 147 Further improving the systems by generating organoids with vessel-like structures that are vital for the transportation of nutrients and waste is a demanding concern. The lack of nutrient transportation to the inner sides of these cellular complexes can limit the survival, proliferation, differentiation, and functionality of organoids. 148 Recently, researchers have generated different organoid models for regeneration, 149 disease modeling, 150 cancer, 151 and pharmacologic studies 152 of the respiratory system ( generated by coculturing different somatic, progenitor, and stem cells. To generate organoids for each part of the lung, from the airways to the alveoli, the complicated 3D architecture, ECM, fluid dynamics of the microenvironment, cell composition, and intracellular interactions of that specific parts need to be recapitulated. 153 Ideally, the vascularized and complicated regions of the alveoli would also be recapitulated. There have been substantial recent advances in lung organoid studies that we review and highlight.

| Stem cell-derived lung organoids
Lung organoids can be generated using mesenchymal (MSCs), embryonic (ESCs), and induced pluripotent (iPSCs) stem cells. The study of stem cell-derived organoids can increase our understanding of developmental processes and underlying pathophysiology to treat lung disease and injuries. ESCs were the first stem cells to be used and were isolated from embryos. They have been used in cell biology studies for over 30 years. 166 These stem cells have many attributes for studying cellular behavior and signaling during the developmental process in the ESC period.
The self-organization of ESCs is a critical point in organoid biology and is genetically encoded inside cells leading to the formation of biological structures. 167  into granulous pneumocytes (alveolar type II cells) to form alveolar organoids. 161 Alveolar type II cells produce pulmonary surfactants and play an important role in gas exchange.
The dysfunctionality of alveolar type II cells is fatal in lung diseases, especially lung cancer, which is a huge clinical issue. Yamamoto et al. generated alveolar organoids using hiPSCs and expanded them to model human alveolar development and study the in vitro toxicity of drugs. 162 In a recent study related to the 2020 COVID-19 pandemic, Bose et al. generated 3D human lung organoids and 3D bronchial transient epithelial/progenitor cells by differentiating ethnicity-based iPSCs. These organoids and cells modeled the lower respiratory tract, providing researchers a way to understand differences in severity and infectivity of SARS-CoV-2. 179 The high costs of generating, gene editing, purifying, and expanding iPSCs are significant limitations and need advanced technology facilities.

| Patient-derived lung cancer organoids
Organoids derived from the cancerous lung of patients can mimic the native architecture of tumors and are considered the optimal in vitro models for anticancer drug studies. 180

| Normal lung-derived organoids
Various organoids have been generated by culturing cells isolated from different regions of normal Lungs such as the trachea, airways, and nasal epithelium. The main elements of these organoids are basal progenitor cells, airway secretory cells such as goblet cells, alveolar type II cells, multiciliate cells, and Alveolar type 2 cells. 164 Hild et al.
developed lung organoids termed mature bronchospheres composed of functional multiciliate, airway basal, and mucin-producing goblet cells for high-throughput studies of human airway epithelium. 164 Lung organoids can also be used to identify crucial genes that control airway function, such as fluid/gas transport, selective permeability, innate immunity, and barrier formation using clustered regularly interspersed short palindromic repeats (CRISPR)/ CRISPR-associated protein 9 (Cas9) gene-editing technology. 182 The findings can be useful in predicting possible oncogene-activated mutations introduced by CRISPR/Cas9 and cancer in lung organoids. Recently, different groups have reported the use of primary lung alveolar organoids to better understand the underlying pathogenesis of SARS-CoV-2 infections and identify effective drugs against the virus. 183,184 Despite its widespread usage, controlling and reproducing the biochemical and biophysical environment required for organoid development is challenging. The lack of dynamic vascular supply and dependence on passive diffusion for growth make it difficult to grow large organoids. Additionally, the variations in size, structure, and gene expression among organoids restrict their application in drug testing and disease modeling. However, altering the inherent properties of cells provides a potent means of enhancing the stability of organoids and customizing them for specific uses. The field of genetic engineering, personalized organoids, and bioprinting organoids and organoidforming cells are still in its early stages, but they hold the potential to advance the use of organoids in tissue engineering and facilitate the creation of functional organs and the large-scale cultivation of synthetic tissues.
Recent advances in tissue engineering combined with additive manufacturing techniques have given rise to organson-a-chips, miniaturized dynamic in vitro cell culture models that recapitulate the vital structural and functional elements of human organs. 185,186 These systems recreate the tissue-tissue interface, vascular perfusion and biochemical gradients, and infusion of immune cells and other connective tissues. 187 Primary cells, iPSCs, or patientderived organoids can be used to line the device with a continuous flow of media in the endothelium-lined channels resulting in an active exchange of nutrients, drugs, and waste. 188 Creating an entire human lung with all the critical structural and functional elements is not technically feasible. 186 Lung-on-a-chips offer the ability to reproduce the well-defined functional units of the lungs. They can mimic the mucociliary barrier of the airways, alveolar-capillary interface, and inflammatory responses, which is critical to studying the disease process and evaluating new drugs and toxins.

| Early models
The lung-on-a-chip sector has gained substantial traction in terms of its number and design following the model of Huh et al., which aimed to mimic the human lung environment more closely. 189 Numerous comprehensive reviews have extensively discussed lung-on-chip models, including the study of different sections of the human respiratory system (alveolus, bronchi), fabrication processes, cell types (e.g., primary cells, co-/triple cultures), and use of different materials (polydimethylsiloxane [PDMS], poly(methyl methacrylate), bio-ink). 185,[189][190][191] Takayama and his group developed a small human airway-on-a-chip and cultured primary lung epithelial cells at the ALI for >3 weeks to form a structurally intact fully differentiated airway epithelium with barrier functions. 192

| Infections
The potential to infect lung epithelial cells with pathogens (bacteria, viruses, or fungi) in a lung-on-a-chip platform provides the opportunity to study host-pathogen interactions. 191 desethylamodiawuine, showed promising results both in vitro and in vivo, suggesting that they may be promising therapies for COVID-19. Findings from these studies suggest that the lung-on-a-chip may be a suitable model to replicate clinically relevant organ-level responses to viral infections. Combined with existing cell-based and animal assays, lung-on-a-chip models offer a powerful platform to study the underlying pathophysiology of respiratory diseases and expedite the discovery of effective and safer drugs to combat pandemic viral infections. 211,212 Different studies conducted using lungon-a-chip models have been summarized in Table 4.

| Organoids-on-a-chip
Organoids and organs-on-a-chip are two inherently different yet complementary approaches to replicate the specific functions and major features of human organs in vitro. 169 These two advanced approaches can be synergistically combined to recapitulate the complexities of human organs in an easily reproducible, controllable, and accessible manner. 225 Organoids being similar to actual organs are an improved model for identifying targets and validation in the drug discovery process. In contrast, organs-on-a-chip provide greater efficacy and safety testing through more controlled and reproducible results. 226 Organoids-on-a-chip offer control of the biochemical and biophysical environment that are key in developing fully mature organoids SHRESTHA ET AL.
| 1491 T A B L E 4 Different studies utilizing Lung-on-a-chip models.

Disease studied Model features References
Asthma IL-13 induced acute exacerbation of asthma and evaluated tofacitinib, a JAK inhibitor, as a potential new therapeutic 199 HRV-16-induced asthmatic exacerbation and assessed the efficacy of MK-7123, a CXCR2 antagonist 201 Human asthmatic musculature to mimic asthmatic bronchoconstriction and bronchodilation and evaluated HA1077, as a potential treatment for treating acute bronchoconstriction and preventing hypercontraction of bronchial smooth muscles. 202 COPD Investigated the role of STAT3 phosphorylation in cigarette smoke-associated COPD and malignant transformation of bronchial epithelium and evaluated HJC0152 as a potential treatment 206 The model used epithelial cells from COPD patients and replicated the viral and bacterial infection-led clinical exacerbation of COPD and evaluated BRD4-inhibitor, a new anti-inflammatory drug 199 Bronchial epithelium isolated from COPD patient lungs to study the in vitro responses to e-cigarettes 171 Lung cancer Multiorgan chip to study the in vivo lung cancer metastasis 214 Evaluated the effects of EGFR-targeted Antitumor drugs: Gefitinib, Osimertinib, and Afatinib 215 Effects of different patterns of hypoxia on major subtypes of lung cancer 216 Replicated the lung cancer growth and invasion patterns and effects of physiological breathing motions on the pattern 203 Lung fibrosis Structured color materials to explore IPF phenotypes with mechanical visuals and the role of cyclic stretching 217 iPF derived fibrotic disease model for cystic fibrosis, s and studied fibroblast-endothelium-epithelium interactions 218 Investigated CXCL12-CXCR4 axis mediated ECP induced migration of fibrocytes towards epithelium in airway remodeling 219 Cystic Fibrosis model and response to Pseudomonas aeruginosa and circulating polymorphonuclear leukocytes 220

COVID-19
Tested FDA-approved drugs on SARS-CoV-2 pseudoviral infection 211 Efficacy of multiple antiviral drugs on SARS-CoV-2 pseudoviral infection 197 Inflammatory responses and pathological changes following SARS-CoV-2 infection. Tested potential antiviral therapies 212 Inflammatory responses and entry process of SARS-CoV-2 pseudovirus to explore SARS-CoV-2 pathogenesis 221 Pulmonary Thrombosis Studied pathophysiology and tested new therapeutic 222 Toxins Interleukin-2 induced pulmonary edema and testing of potential new therapeutics Angiopoietin-1, and TRPV4 193 Nanotoxicity of TiO 2 and ZnO nanoparticles 223 with physiological relevance and functionality. Stable flow of chemical gradients in specific time courses, implementation of mechanical forces such as shear stress, and solid mechanical forces like physiological and biomechanical cues develop a fully functional and mature organoid. 227,228 The ability to mimic perfusable blood vessels or incorporate functional vasculature in organs-on-chips provides the growing organoids with sufficient nutrition and oxygen for maturation and longer survival. 229 The organs-on-chips platform also allows the co-culture of different cells and tissues with the capability of replicating multiorgan physiological interactions, providing multiorganoid systems as an enhanced platform for preclinical drug screening. 169 Organoids-on-chips can be used to create patient and population-specific disease models using patientderived organoids, which can be applied to personalized and possibly regenerative medicine. 230 Despite being reliable, cost-effective, reproducible, and physiologically relevant, lung-on-a-chip systems still face important challenges. PDMS, the commonly used material to fabricate organ-on-a-chip devices, can interfere with pharmacological studies as it can absorb lipophilic drugs. 188 Moreover, ECM-coated PDMS membranes used to mimic tissue interfaces and culture media as a surrogate of blood does not precisely represent the in vivo physical and transport properties. 231 To resolve this issue, researchers are exploring different fabrication materials and techniques such as 3D printing, micro-milling, and solvent bonding of glass, thermoplastics, and silicon to create models capable of incorporating multiple features. 232 Maintaining primary human cell heterogeneity is another major challenge. Significant donor-to-donor variation of primary human cells and organoids or even with cells from a single source impairs quality control. 233 Commonly used cell lines in the devices do not recapitulate the tissue-specific functions and lack adequate cell heterogeneity.
The small cell numbers used in devices can also lead to limited secretions, which may not be enough to be detected by widely used traditional assays. Complex designs with multiple features are difficult to fabricate and use, leading to limited throughput. The growth of the lung-on-a-chip sector has recently accelerated, as reflected by bibliometrics with the increase in the number of publications and patents over recent years.
With pharmaceutical companies beginning to adopt organ-on-a-chip technology, a report from Yole Développement, a marketing, technology, and strategy consulting company, expects the market to grow at a compound annual growth rate of 28.6% from $29.6 million in 2018 to $133.9 million by 2024. 234 Although it is a new technology with numerous challenges, further integration of automation and inbuilt sensors, and international funding strategies, organ-on-chips technology, and its applications will grow substantially in the near future. 235 T A B L E 4 (Continued)

Disease studied Model features References
Assessment of exposure to different concentrations of air pollutant fine particulate matter (PM 2.5 ) 224 Physiological responses to smoke generated by electronic cigarettes 171 Assessed the effects of cigarette smoke and the anti-inflammatory drug, Budesonide The risk and incidence of chronic respiratory diseases continue to increase globally. The ongoing COVID-19 pandemic, uncontrolled forest/bushfires, and exposure to outdoor air pollution have contributed to an increased incidence and exacerbation of respiratory diseases. Human-relevant preclinical models of respiratory diseases are valuable in expediting pulmonary research and in developing pulmonary therapies at sustainable costs. 15 The efficacy of preclinical models in predicting human drug responses is vital to avoid failures of expensive and lengthy clinical trials. Animal models are the mainstay of therapeutic research and provide valuable data on in vivo pharmacokinetics, responses, and efficacy, and are mandatory for preclinical evaluation and approval of any novel or repurposed drugs. The investigation of complex physiological processes, their interactions, inflammatory and immune responses, and disease pathogenesis at the tissue and organ level and multiorgan interconnections is not possible with conventional cell culture models and must be performed in animal models. 188 Unfortunately, the cross-species variation of animal models in lung physiology, underlying molecular and cellular mechanisms, and gene expressions with humans creates issues with the predictive power for human drug responses. 188 Moreover, they are expensive, time-consuming, and may have ethical issues. 231 Organoids and organs-on-a-chip are novel technologies with promising applications in organ development, disease modeling, and the physiology of human organs. 191 Organoids replicate critical structural and functional in vivo properties, whereas the organ-on-a-chip model provides reproducible results with precise control of the microenvironment and inclusion of biophysical and chemical cues. The combination of these two systems can create a state-of-the-art personalized disease model using organoids derived from a patient's tissue specimen. 169 Multiple organs-on-chips, each representing an organ of the human body, can be integrated to develop multiorgan systems or body-on-chips that provide an additional prediction of drug responses in the human body. 231 However, it is worth noting that lung organoids, lung-on-a-chip, or lung organoids-on-a-chip do not mimic the function of the whole lung; instead, they represent a major functional subunit of the lungs. They should be used as complementary methods with in vivo animal models to provide greater predictive power particularly of studies of pathogenesis and therapies in different organs. Recent advances in technology and refinement of existing models have led to significant progress in preclinical testing models. A definitive preclinical model that addresses all the issues of translational pulmonary is elusive. All models fundamentally complement each other, each assisting in drug development processes. Thus, the selection of screening models at each stage of the drug development process is vital to successfully define the most effective and safest drug in the market at a reduced cost.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
Data sharing not applicable to this article as no data sets were generated or analyzed during the current study.