Lung-on-a-chip: From design principles to disease applications

Abstract

To address the growing need for accurate lung models, particularly in light of respiratory diseases, lung cancer, and the COVID-19 pandemic, lung-on-a-chip technology is emerging as a powerful alternative. Lung-on-a-chip devices utilize microfluidics to create three-dimensional models that closely mimic key physiological features of the human lung, such as the air–liquid interface, mechanical forces associated with respiration, and fluid dynamics. This review provides a comprehensive overview of the fundamental components of lung-on-a-chip systems, the diverse fabrication methods used to construct these complex models, and a summary of their wide range of applications in disease modeling and aerosol deposition studies. Despite existing challenges, lung-on-a-chip models hold immense potential for advancing personalized medicine, drug development, and disease prevention, offering a transformative approach to respiratory health research.

NOMENCLATURE

AECs

Alveolar cells

ALI

Air–liquid interface

Ang-1

Angiopoietin-1

AT2

Alveolar type 2 cells

BEAS-2B

Bromodomain-containing protein 2B

CCSP

Club cell secretory protein

CF

Cystic fibrosis

CFTR

Cystic fibrosis transmembrane conductance regulator

COPD

Chronic obstructive pulmonary disease

CSE

Cigarette smoke extract

DEX

Dexamethasone

ECM

Extracellular matrix

EGFR

Epidermal growth factor receptor

FDA

U.S. Food and Drug Administration

GERD

Gastroesophageal reflux disease

HBSMCs

Human bronchial smooth muscle cells

HRV

Human rhinovirus

HUVECs

Human umbilical vein endothelial cells

IFN

Increased Interferon

IGF-1

Insulin-like growth factor 1

IPF

Idiopathic pulmonary fibrosis

LADC

Lung adenocarcinoma

LPS

Lipopolysaccharide

MAPK

Mitogen-activated protein kinase

MMPs

Matrix metalloproteinases

MSCs

Mesenchymal stem cells

NSCLC

Non-small cell lung cancer

PC

Polycarbonate

PCL

Poly(ε-caprolactone)

PDMS

Polydimethylsiloxane

PET

Polyester

PLECM

Lung extracellular matrix

PLGA

Poly(lactic-co-glycolic acid)

PLLA

Poly(l-lactic acid)

PMMA

Polymethyl methacrylate

PMN

Polymorphonuclear leukocyte

rhHGF

Recombinant human hepatic growth factor

SFTPC

Surfactant protein c

TGF-β

Transforming growth factor-β

TRPV4

Transient receptor potential vanilloid 4

I. INTRODUCTION

Respiratory diseases, such as chronic obstructive pulmonary disease (COPD) and asthma, are among the leading causes of mortality worldwide.^1,2 In addition, lung cancer remains a major contributor to cancer-related deaths.^3,4 The recent SARS-CoV-2 pandemic has further underscored the vulnerability of the respiratory system and highlighted the critical need for effective models to study lung infections and develop targeted therapies.^5–8 This pandemic resulted in millions of deaths worldwide and left many survivors with debilitating long-term sequelae, including persistent ground-glass opacities, pulmonary fibrosis, and impaired lung function.^9,10 These pressing challenges have catalyzed a growing demand for advanced in vitro lung models to elucidate disease mechanisms and accelerate the development of novel therapeutic strategies.

Traditional preclinical models have limitations. While animal models can recapitulate complex physiological and immune responses, their inability to fully recapitulate human-specific biology and genetic diversity limits the translatability of findings to clinical applications. In addition, ethical concerns and high costs further limit their use.^11–13 Similarly, 2D cell culture, while inexpensive and straightforward, cannot replicate the intricate 3D architecture and physiological microenvironment of native tissues. Cells in 2D systems often exhibit altered behavior due to the artificial flat substrate, potentially leading to inaccuracies in drug response assessments.^14,15 In contrast, 3D cell culture systems provide a more physiologically relevant environment by supporting cell–cell and cell–extracellular matrix (ECM) interactions, morphology, and gene expression patterns that more closely resemble those found in vivo.^16,17 These techniques, including spheroids, organoids, tissue-engineered constructs, and organ-on-a-chip devices, represent a significant step forward, but come with increased complexity and demand for specialized equipment and expertise.^18–20

Lung-on-a-chip devices, a breakthrough application of microfluidics, address many of these limitations by creating microscale 3D models that mimic the structure and function of the human lung. A typical lung-on-a-chip model cross section view is shown in Fig. 1(a). Lung-on-a-chip devices closely mimic critical physiological features such as the air–liquid interface (ALI), mechanical forces from respiration (e.g., cyclic stretch), and fluid flow (e.g., shear stress).^21,22 A typical lung-on-a-chip consists of two microchannels lined with epithelial and endothelial cells, allowing studies under dynamic, physiologically relevant conditions.^23–25 Lung-on-a-chips provide a powerful platform to study key lung functions, including gas exchange, immune responses and tissue inflammation, with greater fidelity than traditional in vitro models.

FIG. 1.

Structure of lung-on-a-chips and the human bronchial tree respiratory system. (a) Schematic representation of the structure and operation of a typical lung-on-a-chip, which has two parallel microfluidic channels separated by a porous membrane, representing airway and vascular compartments lined with epithelial and endothelial cells, respectively. The membrane allows gas exchange, while cyclic stretching is applied to simulate breathing-induced mechanical forces (F)—mimicking breathing—and an air–liquid interface supports cell function, mimicking key lung physiology. (b) Schematic of the branching structure of the lungs and the gas-exchange region. Within the lungs, air travels through branching bronchioles until reaching the alveoli, tiny air sacs where gas exchange occurs.

Current lung-on-a-chip research is focused on improving both their physiological relevance and practical utility through several key strategies. Efforts to improve biomimicry include developing advanced membrane materials that more closely mimic the basement membrane of the lung and incorporating different cell types, such as immune cells and fibroblasts, to create more accurate and complex tissue models.^26–28 To facilitate broader applications in drug screening and clinical research, researchers are prioritizing the standardization of lung-on-a-chip manufacturing processes and the development of high-throughput capabilities to improve experimental reproducibility and efficiency.^29,30 In addition, recognizing the interconnected nature of human physiology, there is growing interest in integrating lung-on-a-chips with other organ-on-a-chip systems, such as liver-on-a-chip and immune-on-a-chip, to study inter-organ interactions and systemic responses.^31,32 Beyond research applications, the clinical potential of lung-on-a-chips is beginning to emerge. Early studies demonstrate their promise in predicting patient-specific drug responses, paving the way for personalized medicine.^33,34 These advances position lung-on-a-chip technology as a transformative tool in biomedical research and healthcare. Despite their promise, current lung-on-a-chip technologies face challenges,^28,35,36 particularly in reproducing the structural and functional properties of the alveolocapillary barrier.

This review explores the development and application of lung-on-a-chip technology as a powerful platform for advancing lung disease research. It examines the evolution of in vitro lung models, emphasizing the principles and applications of lung-on-a-chip systems in the study of a wide range of lung diseases. Finally, it discusses future directions and the transformative potential of this rapidly evolving field.

II. STRUCTURE AND FUNCTION OF HUMAN LUNGS

The human lungs are essential for the exchange of gases—oxygen in, carbon dioxide out—between the body and the outside world. This respiratory process relies on a complex interplay of anatomical structures and physiological mechanisms.^37–39

Air enters and leaves the lungs through the airway system: the trachea, bronchi, and progressively smaller bronchioles that branch throughout the lung tissue [Fig. 1(b)]. The trachea divides into the left and right main bronchi, each of which leads to a lung. These further divide into secondary and tertiary bronchi, then into bronchioles, and finally into alveolar ducts and sacs. This pattern of branching, some 23 generations in all, creates a vast network, with each branch narrowing as it approaches the alveoli. The airways also condition the inhaled air, warming and humidifying it before it reaches the gas exchange sites in the alveoli.^40,41 The lining of the airways, or epithelium, changes along the way. In larger bronchi, it is primarily a pseudostratified ciliated columnar epithelium, which changes to a simple columnar epithelium in smaller bronchioles. Mucus-secreting goblet cells and serous glands are present in the upper respiratory tract, decreasing in number as the airway descends. The cilia on these epithelial cells are essential for the clearance of inhaled particles, pathogens, and toxins and act as a key defense mechanism. Smooth muscle within the airway walls controls airway diameter by contracting and relaxing in response to various signals. This muscle regulates airflow, but in conditions such as asthma or COPD, excessive contraction or inflammation can narrow the airways, causing wheezing, coughing, and shortness of breath. Controlling smooth muscle tone and airway diameter is critical for effective ventilation and gas exchange.^42–44

Alveoli are the gas exchange units of the lung. An adult lung contains, on average, approximately 480 million alveoli, which make up about 90% of its volume.⁴⁵ Their structure is perfectly suited for the efficient exchange of oxygen and carbon dioxide between air and blood.⁴⁶ Each alveolus is lined with two types of epithelial cells: type I and type II alveolar cells (AECs). Type I alveolar cells are thin and flat, covering approximately 95% of the alveolar surface and providing a minimal barrier to gas diffusion. They facilitate the movement of oxygen into the blood and carbon dioxide from the blood into the alveolar air to be expired. Type II alveolar cells, although fewer in number, are essential for lung function. They produce and secrete surfactant, a substance that reduces surface tension within the alveoli, preventing them from collapsing and facilitating lung expansion during inhalation. Surfactant also contributes to the lung's innate immune defenses by trapping and neutralizing pathogens.^47–51 Tiny pores, called pores of Kohn, connect adjacent alveoli, allowing airflow between them and helping to equalize pressure. This interconnectivity also contributes to the lung's stability, preventing individual alveoli from collapsing under pressure changes during breathing.⁵²

The ability of the lung to expand and contract depends on the compliance of the alveolar wall and the elasticity of the surrounding lung tissue. These mechanical properties are primarily determined by the structure and composition of the airway membrane and alveolar interstitium.^53–56 The airway membrane, or air–blood barrier, consists of alveolar epithelial cells, capillary endothelial cells, and a common basement membrane. This thin, multilayered structure allows for efficient gas exchange by diffusion. The alveolar interstitium, located between the alveolar epithelium and the capillary endothelium, contains extracellular matrix (ECM) components such as elastic fibers, collagen, and other structural proteins. These components provide both compliance and elasticity, with elastic fibers being particularly important for lung recoil after inspiration, allowing the lung to return to its resting volume during expiration. The interstitium also harbors various cells, including fibroblasts, macrophages, and myofibroblasts, which contribute to the remodeling of lung tissue, particularly in response to injury or inflammation. Together, these structural and cellular elements ensure the mechanical and functional integrity of the lung during normal breathing and pathological conditions.^57,58

III. DEVELOPMENTS IN IN VITRO MODELING OF THE HUMAN LUNG

Recent years have seen remarkable advances in the development of in vitro lung models. These models, ranging from simple 2D cell cultures to sophisticated 3D organoids and lung-on-a-chip systems, attempt to recapitulate the complex structure and function of the human lung (Fig. 2).⁵⁹

FIG. 2.

Various in vitro human lung models for cell cultures. (a) Standard 2D air–liquid interface configuration.⁶⁰ Reproduced with permission from Bennet et al., Cells 10(7), 1602 (2021); licensed under a Creative Commons Attribution (CC BY) license. (b) A scaffold-free hanging drop spheroid formation method.⁶¹ Reproduced with permission from Nath et al., Pharmacol. Ther. 163, 94 (2016). Copyright 2016 Elsevier. (c) Organoid culture methods.⁶² Reproduced with permission from Gunti et al., Cancers 13(4), 874 (2021); licensed under a Creative Commons Attribution (CC BY) license. (d) Different types of scaffold-based 3D cellular models.⁶³ Reproduced with permission from Unnikrishnan et al., Front. Oncol. 11, 733652 (2021); licensed under a Creative Commons Attribution (CC BY) license. (e) A typical lung-on-a-chip structure with alveolar epithelial cells and pulmonary microvascular endothelial cells are cultured at the top and bottom of the ECM-coated membrane, respectively.³⁵ Reproduced with permission from Francis et al., Drug Discov. Today 27(9), 2593 (2022). Copyright 2022 Elsevier.

A. 2D and ALI cultures

A key early study by Elbert et al. established a method for culturing human alveolar epithelial cells in 2D using collagen-I or fibronectin-coated inserts.⁶⁴ These cultures showed the formation of tight junctions and a transepithelial electrical resistance indicative of a functional epithelial barrier. A standard 2D air–liquid interface configuration is shown in Fig. 2(a).⁶⁰ However, 2D cultures have limitations. Cells grown this way often lose their natural shape and ability to differentiate after a few days. For instance, alveolar type II cells, crucial for surfactant production, change morphologically and lose function within 3–5 days in culture.^65,66 To address these limitations, researchers have increasingly used more advanced techniques like the ALI model, which better mimics the lung's in vivo environment. Bluhmki et al. showed that small airway epithelial cells grown at ALI for over four weeks formed a pseudostratified epithelium with basal, club, goblet, and ciliated cells.⁶⁷ These cells also responded to fibrotic cytokines like TGF-β1 and TNF α, demonstrating the ALI model's value for studying lung pathophysiology. Xu et al. further confirmed that ALI enhances epithelial cell differentiation and metabolic activity compared to traditional 2D cultures.⁶⁸ Consequently, the ALI model has become a vital tool for studying lung diseases, providing a more physiologically relevant setting.

B. Spheroid and organoid cultures

While 2D and ALI cultures are useful, they do not fully capture the complex 3D architecture of the human lung. Spheroid and organoid cultures address this issue by allowing cells to self-organize into more complex, tissue-like structures.^62,69 Spheroids, often derived from patient cells, can be formed from a variety of lung cell types, including cancerous and healthy cells. They are particularly useful for studying lung cancer because they retain the genetic and histological characteristics of the parental tumor. There are several methods for generating spheroids, including matrix-based (matrix-on-top, -embedded, -encapsulation), spinner flasks, micropatterned/ultra-low attachment plates, hanging drops, and magnetic (levitation, 3D printing) techniques.⁶¹ Cores et al. developed a method to generate lung spheroids from healthy human lung tissue that expressed progenitor cell markers such as surfactant protein C (SFTPC) and club cell secretory protein (CCSP).⁷⁰ These spheroids have been used in personalized medicine. Spheroids have also shown promise for drug screening, particularly for lung adenocarcinoma (LADC), as they retain key tumor biomarkers and phenotypes. Figure 2(b) shows the scaffold-free hanging drop spheroid formation method.⁶¹ However, spheroid cultures present challenges, including variability in size, hypoxia in larger spheroids, and difficulty maintaining cellular diversity in smaller spheroids.⁷¹ Organoids, as shown in Fig. 2(c), which are more complex than spheroids, have been used to model lung development, disease, and therapy.^62,72,73 These structures can self-organize into multiple cell types, including airway and alveolar cells, providing a more accurate representation of lung tissue. However, similar to spheroids, organoids face challenges related to size, oxygen diffusion, and maintaining cellular heterogeneity.⁷⁴

C. Scaffold-based 3D cultures

To better mimic the mechanical and biological environment of the lung, scaffold-based 3D cultures have gained attention. Scaffolds provide a support structure for 3D cell growth while preserving cell shape and function.^63,75 Figure 2(d) illustrates different types of scaffold-based 3D cellular models.⁶³ Crabbé et al. showed that recellularization of decellularized mouse lungs with mesenchymal stem cells (MSCs) and alveolar type 2 cells (AT2) in a bioreactor improved cell viability and reduced apoptosis, creating a useful model for studying lung tissue repair.⁷⁶ However, natural ECM scaffolds can have problems such as weaker mechanical properties, rapid degradation, and batch-to-batch variability. To overcome this, researchers have combined natural and synthetic materials to create hybrid scaffolds that provide both biological cues and mechanical strength. For example, Rezaei et al. developed a chitosan-polycaprolactone scaffold that supported lung fibroblast culture and could be used in lung tissue engineering.⁷⁷ 3D printing has also been used to create more complex lung tissue models. Using hydrogels or cell-laden inks, researchers have printed lung cancer models that replicate the tumor microenvironment, improving the study of cancer cell migration and invasion. Young et al. fabricated electrospun scaffolds from poly(l-lactic acid) (PLLA) and decellularized porcine lung ECM (PLECM) that supported the growth of human bronchial smooth muscle cells (HBSMCs) and allowed them to develop a contractile shape.⁷⁸ This highlights the potential of 3D scaffolds to study chronic lung diseases such as asthma and COPD.

D. Lung-on-a-chip platforms

The most advanced in vitro lung models are lung-on-a-chip systems that combine microfluidics with cell culture.^26,79,80 These platforms create dynamic, physiologically relevant environments that simulate the mechanical forces and fluid flows of the lung. Figure 2(e) shows a typical lung-on-a-chip structure with alveolar epithelial cells and pulmonary microvascular endothelial cells are cultured at the top and bottom of the membrane, respectively.³⁵ Lung-on-a-chip systems are particularly useful for mimicking the alveolar–capillary barrier, studying interactions between different lung cell types, and investigating disease mechanisms. These platforms often incorporate multiple cell types, including alveolar epithelial cells, endothelial cells, and fibroblasts, to mimic complex in vivo cell interactions. For example, Zhang et al. used a lung-on-a-chip to model the alveolar–capillary barrier and assess the pulmonary toxicity of nanoparticles. Human alveolar epithelial cells and human umbilical vein endothelial cells were cultured on opposite sides of the chip to study drug delivery, toxicity, and inflammation.⁸⁰ Campillo et al. discussed the use of lung-on-a-chip systems for regenerative medicine, particularly with stem cells.⁸¹ Embryonic and induced pluripotent stem cells can differentiate into lung-specific cell types and, when cultured in a microfluidic environment, could be used to model lung disease or develop therapies. However, maintaining the differentiation of induced pluripotent stem cells into mature lung lineages in vitro remains a challenge, requiring complex and carefully controlled conditions.

IV. KEY ELEMENTS OF LUNG-ON-A-CHIP SYSTEMS

Lung-on-a-chip technology aims to recreate the complex microenvironment of the human lung in vitro, providing a powerful tool for the study of lung physiology, disease modeling, and drug development. A key aspect of this technology is the precise design of the chip itself, which involves mimicking several critical structural and functional features of the lung.

A. The air–liquid interface

The ALI, where air and blood meet within the alveoli, is essential for gas exchange and also serves as the primary portal of entry for airborne pathogens and contaminants. In lung-on-a-chip devices, the ALI is typically created by separating an air-filled or gas-perfused channel from a fluid-filled channel using the porous membrane. Development of the airway membrane has progressed through several stages, each addressing the limitations of previous iterations. Early models often utilized Transwell membranes, which provided a basic platform for culturing epithelial and endothelial cells on opposite sides of a porous membrane.^82,83 However, Transwell systems lack compatibility with microfluidic systems and offer limited biomimicry of the dynamic lung environment. The advent of microfluidics has driven the adoption of polymeric membranes, particularly polydimethylsiloxane (PDMS), as the predominant material for lung-on-a-chip fabrication.^84,85 PDMS offers a unique combination of properties highly relevant to lung physiology: optical transparency for real-time imaging, gas permeability essential for nutrient and oxygen exchange, and elasticity that allows precise mimicry of the dynamic mechanical deformation induced by respiration. This elasticity is critical to mimic the physiological stress experienced by alveolar cells during respiration. Other polymers, including polycarbonate (PC), polyester (PET), and polymethyl methacrylate (PMMA), have also been explored for their ease of fabrication, flexibility, and cost-effectiveness, particularly in high-throughput screening applications where large numbers of chips are required.^86–88

A key design parameter of the respiratory membrane is the pore size (typically ranging from 1 to 10 μm), which ensures adequate exchange of nutrients and signaling molecules between adjacent epithelial and endothelial cells, while effectively preventing cell migration or leakage across the barrier.⁸⁹ To enhance biocompatibility and promote cell adhesion and spreading, membrane surfaces are often modified with ECM proteins such as collagen, gelatin, fibronectin, or fibrin. These ECM coatings provide a more natural substrate for cell attachment and influence cell behavior.^90,91 Precise control of flow rates in both the air and fluid channels is critical for maintaining ALI and delivering nutrients and oxygen to cells. The shear stress exerted by flowing blood (or culture medium) on endothelial cells must be carefully considered in the design of the fluid channel. This shear stress, typically in the range of 1–10 dyn/cm², plays a critical role in regulating endothelial cell behavior, including proliferation, differentiation, barrier function, and inflammatory responses.^92,93 The fluid channel can also be used to incorporate immune cells and model the complex interplay between the vasculature and the immune system in lung diseases. Recent advances have focused on integrating both airflow and blood flow within the same chip to more accurately reproduce the complex transport phenomena that occur in the lung.⁹⁴ The development of 3D multivessel networks within lung-on-a-chip devices enhances the physiological relevance of these models by better mimicking the complex vascular architecture of the lung. In addition, the air channel also allows the introduction of various airborne substances to model inhalation and study the pathogenesis of various lung diseases.^95,96

B. Cellular models

The selection of appropriate cell sources is critical to the success of any in vitro lung model. Primary cells derived directly from lung tissue offer high physiological relevance and retain many of the characteristics of their in vivo counterparts. However, primary cells have limited proliferative capacity and can exhibit donor-to-donor variability. Immortalized cell lines, while offering ease of use and consistent cell supply, may not fully represent the complex physiology of native lung cells.^97,98 Stem cells, including embryonic stem cells and induced pluripotent stem cells, offer the potential to generate multiple lung cell types through directed differentiation. While significant progress has been made in the development of differentiation protocols, the generation of fully mature and functional lung lineages from stem cells remains a challenge.⁹⁹

The alveolar–capillary interface is a complex structure composed of several key cell types: type I and II alveolar cells, microvascular endothelial cells, interstitial fibroblasts, and resident immune cells (e.g., alveolar macrophages).¹⁰⁰ The most common co-culture configuration includes endothelial, stromal (fibroblasts), and epithelial cells, often with the addition of immune cells to create more complex models for studying lung infection, inflammation, or immune responses.¹⁰¹ Co-culture systems incorporating multiple cell types such as epithelial cells, endothelial cells, fibroblasts and immune cells are increasingly being used to better mimic the complex cellular interactions that occur in the lung. The specific application of the in vitro model should ultimately guide the choice of cell source and culture conditions. For example, a model designed to study pulmonary fibrosis might prioritize the use of fibroblasts and epithelial cells, while a model designed to study lung cancer might focus on tumor cells and their interactions with the surrounding microenvironment. While significant progress has been made in recapitulating cellular organization, challenges remain in fully mimicking the complex interactions and dynamic behavior of different cell types within the alveolus. For example, the dynamic movement of type I and type II alveolar cells between alveolar spaces and their differential contributions to surfactant secretion and alveolar fluid clearance are aspects that are still difficult to fully recapitulate in current in vitro models.¹⁰²

C. Dynamic respiratory motions

Dynamic respiratory motion, a critical physiological factor influencing lung function and cell behavior, is increasingly being incorporated into lung-on-a-chip designs.¹⁰³ To achieve dynamic respiratory motion, various mechanical actuation methods have been employed. Pneumatic systems are among the most widely used due to their ability to generate precise and controllable cyclic strain. In these systems, air pressure is applied to deformable membranes or side chambers, creating rhythmic expansion and contraction that mimics breathing. For example, some designs utilize microfluidic channels integrated with thin, elastic membranes that respond to applied pressure gradients, enabling the simulation of alveolar stretching.^104–107 Early models used planar stretching of the membrane by applying vacuum pressure to side chambers adjacent to the cell culture area. While a significant improvement over static cultures, this approach only partially captures the complex, three-dimensional, arcuate expansion and contraction of the alveoli during breathing. More recent and advanced designs have focused on developing more biomimetic respiratory motions to bring lung-on-a-chip models closer to physiological pulmonary ventilation and alveolar expansion. Electromagnetic actuators and integrated microfluidic pumps have also been explored as alternative methods to generate respiratory motion, offering advantages in terms of miniaturization and control precision. These systems can be fine-tuned to replicate physiological parameters such as cyclic strain frequency (typically ∼0.2 Hz, corresponding to 12 breaths per minute) and tensile strain (ranging from 5% to 15%, depending on the lung region and breathing pattern).¹⁰⁸

In healthy adult lungs, transpulmonary pressure—the difference between airway pressure and pleural pressure—is typically low, often less than a few cm H₂O. This low pressure is largely due to the presence of pulmonary surfactants, which reduce surface tension within the alveoli, preventing collapse and facilitating effortless breathing.¹⁰⁹ Replicating these low pressures in lung-on-a-chip devices presents significant challenges. The mechanical actuation systems used to simulate breathing must be finely tuned to apply physiological levels of stress without exceeding the delicate pressure balance found in vivo. Recently, some advanced lab-on-a-chip platforms have integrated pressure sensors and feedback mechanisms to monitor and adjust the applied pressures to ensure they remain within the physiological range.^110,111 In addition, the incorporation of surfactant-like substances into the device can help mimic the surface tension dynamics of the alveoli, contributing to a more accurate simulation of lung mechanics.¹¹²

V. FABRICATING LUNG-ON-A-CHIP MODELS

Successful lung-on-a-chip design depends on mimicking key aspects of natural lung function, including maintaining tissue integrity, mimicking appropriate mechanical properties, and functionally simulating lung-specific processes such as gas exchange, respiratory motion, and fluid transport.⁷⁹ A functional lung-on-a-chip requires scaffolds that support both cellular growth and adequate airflow to create a viable air–blood interface. This interface must be thin, flexible, and able to withstand mechanical stress during simulated breathing while supporting cellular adhesion and growth.^113,114 Related fabrication methods are presented in detail (Fig. 3).

FIG. 3.

Fabrication methods for lung-on-a-chip models. (a) PDMS-based fabrication of lung-on-a-chips, the typical soft lithography technique.¹¹⁵ Reproduced with permission from Man et al., ACS Appl. Mater. Interfaces 15(30), 36888 (2023). Copyright 2023 ACS. (b) A PMMA-based thermoplastic methods lung airway-on-a-chip.¹¹⁶ Reproduced with permission from Humayun et al., Lab Chip 18(9), 1298 (2018). Copyright 2018 RSC. (c) Bioprinting techniques in different ways.¹¹⁷ Reproduced with permission from Mandrycky et al., Biotechnol. Adv. 34(4), 422 (2016). Copyright 2016 Elsevier.

A. Soft lithography and microfabrication

Soft lithography, a well-established microfluidic technique, is widely used to fabricate lung-on-a-chip models.^{104,105,118,119} In this method, molds are created using photolithography, and then elastomeric materials such as polydimethylsiloxane (PDMS) are cast to form microfluidic channels, as shown in Fig. 3(a).¹¹⁵ These PDMS-based channels can mimic air and blood flow, with cell cultures seeded in the upper and lower channels to replicate alveolar and vascular regions. Soft lithography excels at producing high-quality, transparent, and biocompatible microstructures that are ideal for lung-on-a-chip models that require real-time imaging. For example, Huh et al. have used soft lithography to develop a PDMS-based lung-on-a-chip model that mimics the alveolar-capillary interface, consisting of microchannels separated by a thin, flexible, microporous membrane. Cyclical application of vacuum to side chambers allowed stretching of the membrane, thereby simulating respiratory motion.⁷⁹ While effective, soft lithography has limitations in scalability and design complexity, making it difficult to create complex 3D structures or incorporate multiple tissue types. In addition, PDMS, while biocompatible and transparent, can exhibit non-specific protein binding, potentially affecting cell behavior.

B. Thermoplastic methods

Thermoplastic-based fabrication methods, such as micro-milling and solvent bonding, are gaining traction due to their ability to produce robust, reproducible, and scalable devices.^120,121 The schematic principle of this method is shown in Fig. 3(b).¹¹⁶ Researchers have used micro-milling to develop thermoplastic lung airway-on-a-chip models that replicate the airway microenvironment and the interactions between smooth muscle cells, epithelial cells, and extracellular matrix components.¹¹⁶ Thermoplastic devices tend to be more durable and less susceptible to PDMS-related limitations such as protein absorption. These devices can also be disassembled for post-processing, providing greater flexibility. However, these methods require precise control of temperature and material properties, which can limit design complexity. Recent advances have involved combining multiple techniques to overcome individual limitations. For example, the combination of soft lithography for channeling, bioprinting for cell deposition, and thermoplastic methods for structural support has been used to create more complex, functional lung models. Such hybrid approaches offer increased flexibility, precision, and scalability in the manufacturing process.

C. Bioprinting

Bioprinting, especially when combined with microfluidics, provides a powerful and versatile approach for fabricating three-dimensional tissue structures. This technique involves the layer-by-layer deposition of bioinks—containing cells, scaffolds, and biomolecules—to construct functional tissues.^122,123 In lung-on-a-chip applications, bioprinting is used to create cell-laden scaffolds that mimic the microstructure of the lung, including alveolar and airway regions. Currently, bioprinting methods include inkjet, laser-assisted, microextrusion, etc.¹²⁴ The schematic principles of these methods are shown in Fig. 3(c).¹¹⁷

1. Inkjet bioprinting

Inkjet bioprinting is a widely used method for creating highly detailed micro-scale structures. It uses piezoelectric or thermal mechanisms to deposit small droplets of bioink onto a substrate.^125,126 The high resolution enables the creation of intricate tissue architectures, such as the alveolar–capillary interface. A key advantage is its compatibility with low-viscosity bioinks, which are ideal for creating thin, precise tissue layers. For example, Park et al. have demonstrated the use of inkjet bioprinting to create a functional airway model with a blood capillary network using decellularized extracellular matrix bioinks derived from porcine trachea, providing a more biomimetic lung environment.¹²⁷ However, inkjet bioprinting is primarily suited for low-viscosity bioinks, making the construction of models with higher-viscosity tissues challenging. Vertical accuracy and resolution can also be limited for complex structures.

2. Laser-assisted bioprinting

Laser-assisted bioprinting uses laser pulses to vaporize and propel bioink onto a substrate for solidification. Offering high precision, with some systems capable of printing single cells per droplet, this technique is particularly well suited for creating highly detailed structures, such as those needed to simulate the alveolar–capillary interface.^128,129 While laser-assisted bioprinting offers exceptional precision, it can potentially damage cells due to the heat generated during printing. The process can also be time-consuming due to the need for precise control of the laser pulse and the production of cell-laden filaments.

3. Micro-extrusion-based bioprinting

Micro-extrusion-based bioprinting involves the continuous extrusion of bioink through a nozzle to create layered structures. It is ideal for printing higher viscosity materials, including hydrogels commonly used in lung-on-a-chip models, and allows the creation of multi-material structures that mimic different lung tissue types, such as epithelial, endothelial, and fibroblast layers.^130,131 A key advantage is its ability to handle high-viscosity bioinks, which are essential for creating scaffolds that closely mimic natural lung tissue. However, cell viability can be compromised by extrusion pressure, and resolution can be lower compared to inkjet or laser-assisted bioprinting. Nozzle clogging and inconsistent flow rates can also be a challenge.

4. Stereolithography bioprinting

Several studies on the stereolithography bioprinting present an advanced bioprinter capable of producing complex, tissue-like structures.^132–134 This method works by irradiating a photosensitive material with a light source of a specific wavelength (typically ultraviolet or visible light), which triggers a polymerization reaction in selected regions to form a solid structure. Compared to traditional bioprinting methods, stereolithography offers several key advantages: it achieves high resolution with micron-level printing accuracy that can accurately replicate complex biological structures, making it ideal for creating tissue engineering scaffolds and organ models. In addition, it enables rapid printing, accommodates multiple materials, and ensures biocompatibility, making it highly versatile and suitable for a wide range of biological applications.

D. Emerging materials

Recent advancements in materials science have introduced innovative materials that hold promise for enhancing the physiological relevance of lung-on-a-chip devices. For example, researchers have developed a biodegradable, stretchable membrane composed of collagen and elastin, designed to emulate the mechanical properties of lung tissue.¹³⁵ This membrane facilitates a more accurate replication of the lung's dynamic environment within lung-on-a-chip systems. Additionally, the integration of ultrathin hybrid membranes made from poly(ε-caprolactone) (PCL) and gelatin has demonstrated mechanical properties and biocompatibility akin to native lung tissue.¹³⁶ These membranes support the cultivation of lung cells under air–liquid interface conditions, thereby improving the physiological relevance of lung-on-a-chip devices. The application of these novel materials is anticipated to enhance the ability of lung-on-a-chip devices to mimic the human lung microenvironment, thereby advancing research in respiratory diseases and drug development.

VI. APPLICATIONS OF LUNG-ON-A-CHIP TECHNOLOGY

A. Lung disease studies based on lung-on-a-chip

Microfluidic chip is a promising new technology that better mimics the human lung environment, providing a more effective tool for studying complex lung diseases.^137,138 Compared to traditional cell cultures and animal models, lung-on-a-chip can more accurately simulate the process of disease development, helping researchers to better understand disease mechanisms and providing opportunities for personalized medicine. Related investigations using lung-on-a-chip are described in detail (Table I).

TABLE I.

Studies on various diseases using lung-on-a-chip and related models and findings.

Investigations	Models and findings	References
Lung cancer	Cyclic mechanical strain simulating respiration suppressed tumor growth, suggesting a link between respiration and drug resistance.	Hassell et al.139
	Fibroblast-secreted IGF-1 inhibited EGFR pathways, contributing to drug resistance and tumor invasion.	Yang et al.26
	Co-culture of cancer and stromal cells under flow showed fibroblasts promoting tumor formation via TGF-β, MMPs, and biophysical forces.	Lee et al.140
	A 3D-printed lung-on-a-chip model with simulated respiration and stiffness validated cigarette smoke extract effects on metastasis and drug screening with paclitaxel.	Das et al.141
	Extracellular acidification was used to monitor lung cancer cell viability under drug treatment in a microfluidic device.	Khalid et al.142
	A multi-organ micro-physiological system revealed fibroblast and macrophage transformation in lung metastasis to other organs.	Xu et al.32
Asthma	A lung small airway chip with circulating immune cells allowed the study of inflammatory diseases such as asthma and COPD, as well as therapeutic responses	Benam et al.143
	Human airway musculature-on-a-chip reproduced asthmatic muscle responses to IL-13 and demonstrated reversal with standard asthma treatments.	Nesmith et al.144
	Cyclic mechanical stress on the pulmonary alveolar epithelium increased IL-8 secretion, demonstrating the contribution of physical stress to inflammation.	Stucki et al.145
	A micro-engineered model of rhinovirus-induced asthma exacerbations showed that IL-13 pre-treatment impaired antiviral immune responses and recapitulated key features of asthma cytopathology upon HRV infection.	Nawroth et al.146
Cystic fibrosis	A CF Airway Chip mimicked mucus accumulation, cilia activity, inflammation, and P. aeruginosa growth.	Plebani et al.147
Idiopathic Pulmonary Fibrosis	Human lung microtissues were used to replicate key biomechanical events of fibrogenesis, including tissue stiffening and contraction, and to evaluate antifibrotic drug effects	Asmani et al.148
	Exposure of alveolar cells on a lung chip to gastric contents disrupted the epithelial barrier, highlighting the role of epithelial injury in pulmonary fibrosis.	Felder et al.149
	A breathing lung-on-a-chip model incorporating cyclic stretch showed that while stretch impaired wound healing, rhHGF partially mitigated this effect, suggesting potential IPF therapies.	Felder et al.150
Pulmonary edema	A microdevice mimicking pulmonary edema identified mechanical forces in vascular leakage and potential therapeutics like Ang-1 and TRPV4 inhibitors.	Huh et al.151
Pulmonary thrombosis	A lung-on-a-chip model with perfused whole blood demonstrated that LPS-induced thrombosis is mediated indirectly by alveolar epithelial activation, and validated the model for drug testing using a PAR-1 antagonist.	Jain et al.152
Smoking lung injury	A chip using CSE demonstrated permeability assays and drug delivery studies, showing CSE-induced changes in IL-6/IL-8 release and E-cadherin expression, partially reversed by budesonide.	Shrestha et al.153
	A biomimetic smoking robot coupled with a lung-on-a-chip delivered fresh cigarette smoke in a cyclic pattern, revealing similar gene expression changes in oxidation–reduction pathways as seen in human smokers.	Benam et al.154
	Transcriptome analysis of a bionic lung chip identified MAPK pathway involvement in smoke-induced inflammation, validated by biomarkers and DEX anti-inflammatory effects.	Li et al.155

1. Lung cancer

Lung-on-a-chip technology has emerged as a valuable tool for the study of lung cancer, providing a more physiologically relevant in vitro environment compared to traditional methods.¹⁵⁶ Several studies have demonstrated the utility of lung-on-a-chips in studying various aspects of lung cancer biology, including tumor growth, invasion, drug response, and metastasis. Studies have highlighted the importance of the microenvironment and mechanical forces in lung cancer progression. A study by Hassell et al. used a lung cancer chip model of non-small cell lung cancer (NSCLC) to demonstrate that tumor cells proliferated faster in alveolar chips compared to airway chips, mirroring in vivo observations in lung adenocarcinoma patients.¹³⁹ Specifically, cyclic mechanical strain simulating respiratory motion suppressed tumor cell growth and invasion into vascular channels, suggesting a link between respiration and the formation of drug-resistant cancer cells. This highlights the impact of physiological exercise on tumor development. Another study by Yang et al. developed a lung-on-a-chip model using poly(lactic-co-glycolic acid) (PLGA) and co-cultured NSCLC cells with lung fibroblasts.²⁶ Their research showed that fibroblasts secrete insulin-like growth factor 1 (IGF-1), which inhibits the epidermal growth factor receptor (EGFR) pathway targeted by the drug gefitinib, potentially contributing to reduced drug sensitivity. In addition, when endothelial cells were added to the co-culture, the NSCLC cells induced endothelial cell apoptosis, suggesting a mechanism for tumor cell invasion.

Further research has focused on incorporating more complex features. Lee et al. created a microfluidic model to study the role of stromal cells in tumorigenesis.¹⁴⁰ Co-culturing cancer cells with endothelial cells and fibroblasts in a collagen matrix under continuous flow revealed that transforming growth factor-β (TGF-β), matrix metalloproteinases (MMPs), and biophysical forces generated by fibroblasts promoted tumor formation. The presence of fibroblasts also upregulated metastasis-associated genes and angiogenesis while downregulating apoptosis-related genes. This model has also been used to test chemotherapeutic agents. Another study by Das et al. developed a fully 3D-printed lung-on-a-chip model that mimicked in vivo conditions by co-culturing lung cancer cells with lung fibroblasts in a hydrogel with a surface stiffness similar to lung tissue.¹⁴¹ The device included an air channel to simulate respiratory cycles. This model was used to study disease progression (metastasis) and drug efficacy. Exposure to cigarette smoke extract exacerbated metastatic characteristics, while dose-dependent drug testing with paclitaxel validated the model for both metastasis studies and drug screening. Real-time monitoring capabilities have also been incorporated into lung-on-a-chip systems. Khalid et al. developed a microfluidic chip with integrated optical pH and impedance sensors and fluorescence microscopy to monitor the viability of lung cancer cells under drug treatment.¹⁴² The system successfully detected extracellular acidification due to cell death and demonstrated the differential effects of docetaxel and doxorubicin on cell viability. To address the issue of metastasis, Xu et al. proposed a multi-organ micro-physiological system to mimic lung metastasis to brain, bone, and liver.³² By co-culturing lung cells with cells from these target organs, they observed the transformation of fibroblasts and macrophages into cancer-associated phenotypes and detected damage-associated markers in cells from the distant organs. This platform provides a valuable tool to study the complex process of lung cancer metastasis.

2. Asthma and COPD

Asthma is multifactorial, involving genetic, host, and environmental factors.¹⁵⁷ Approximately 300 million people worldwide are affected by the disease. The SARS-CoV-2 pandemic has also highlighted the potential of asthma to exacerbate viral effects on the respiratory system.^158,159 Several microfluidic models have been developed to investigate the complex mechanisms of asthma and COPD, and to evaluate therapeutic interventions. Benam et al. developed a lung small airway chip that supported a differentiated, mucociliary bronchiolar epithelium underlain by lung microvascular endothelium.¹⁴³ Immune cells circulated through the underlying fluid flow, allowing the study of complex inflammatory diseases such as asthma and COPD, as well as therapeutic responses. Nesmith et al. constructed a human airway musculature-on-a-chip containing bronchial smooth muscle cells on elastomeric films.¹⁴⁴ This model reproduced the in vitro asthmatic muscle response to IL-13 and demonstrated reversal by treatment with a muscarinic antagonist and a β-agonist. Stucki et al. demonstrated that cyclic mechanical stress on the pulmonary alveolar epithelium increased IL-8 secretion, illustrating how physical stress contributes to inflammation.¹⁴⁵ Nawroth et al. developed a micro-engineered model to study rhinovirus-induced asthma exacerbations, incorporating airway epithelial cells and human umbilical vein endothelial cells (HUVECs) cultured on opposite sides of a porous PET membrane within a microfluidic device.¹⁴⁶ Induction of an asthmatic phenotype with IL-13 resulted in goblet cell hyperplasia, reduced cilia beating, and endothelial activation, mimicking key asthma pathologies. Subsequent human rhinovirus (HRV) infection in this system increased interferon (IFN) secretion, IFN-induced chemokines, and IL-8 secretion, while IL-6 levels decreased. These results suggest that IL-13 impairs normal immune responses to viral infection and confirm the ability of the model to recapitulate key biochemical features of asthma cytopathology.

3. Fibrosis

Both cystic fibrosis (CF) and idiopathic pulmonary fibrosis (IPF) are serious lung diseases associated with fibrosis; however, despite sharing “fibrosis” as a common feature, CF and IPF arise from distinct biological mechanisms and progress through separate pathological pathways.^160,161 Lung-on-a-chip platforms have emerged as powerful tools for dissecting the complex cellular and molecular mechanisms driving fibrosis in a controlled in vitro environment. These models recapitulate key aspects of the lung microenvironment, including interactions between epithelial, endothelial, and fibroblast cells. While CF is primarily characterized by defective mucociliary clearance due to mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, leading to mucus accumulation and susceptibility to bacterial infections, fibrosis contributes significantly to disease progression.¹⁶² Plebani et al. developed a novel microfluidic CF Airway Chip to recapitulate key features of human cystic fibrosis (CF) airways.¹⁴⁷ By co-culturing primary human CF bronchial epithelial cells at an air–liquid interface with pulmonary microvascular endothelial cells exposed to fluid flow, the chip successfully mimicked enhanced mucus accumulation, increased cilia density and beating frequency, and elevated IL-8 secretion characteristic of CF. Furthermore, the CF Airway Chip demonstrated increased polymorphonuclear leukocyte (PMN) adhesion and transmigration, and provided a more conducive environment for Pseudomonas aeruginosa growth, leading to enhanced inflammatory responses and PMN recruitment. IPF is characterized by progressive scarring of alveolar tissue, resulting in reduced lung compliance and impaired gas exchange.¹⁶³ Asmani et al. developed a novel in vitro model of pulmonary fibrosis using membranous human lung microtissues to recapitulate key biomechanical events of fibrogenesis, including progressive tissue stiffening and contraction, reduced tissue compliance, and traction-induced bronchial dilation. This model was used to evaluate the therapeutic efficacy of two U.S. Food and Drug Administration (FDA)-approved antifibrotic drugs, pirfenidone and nintedanib.¹⁴⁸ Epidemiological studies have suggested an association between IPF and gastroesophageal reflux disease (GERD). Felder et al. developed a lung chip with an epithelial wound scaffold to evaluate the effect of gastric contents (hydrochloric acid and pepsin) on alveolar cells.¹⁴⁹ Exposure to these substances resulted in significant disruption of the alveolar epithelial barrier, highlighting the role of epithelial injury in driving pulmonary fibrosis. The group also developed a breathing lung-on-a-chip model incorporating cyclic mechanical stretch to study wound healing processes.¹⁵⁰ This model was used to evaluate the effects of recombinant human hepatic growth factor (rhHGF), a key regulator of tissue repair, and cyclic mechanical stretch. While cyclic stretch impaired healing, rhHGF partially mitigated this effect, providing insight into potential therapeutic strategies for IPF.

4. Pulmonary edema

Pulmonary edema refers to the condition characterized by fluid accumulation in the lung's alveolar spaces, which disrupts normal respiratory function. This medical condition can result from various cardiac issues that increase left ventricular pressure, leading to elevated hydrostatic pressure in the pulmonary capillaries.¹⁶⁴ Huh et al. developed a microdevice designed to mimic pulmonary edema, specifically to study the toxicity of IL-2 and its associated pulmonary edema observed in cancer patients undergoing IL-2 treatment.¹⁵¹ This on-chip model revealed that mechanical forces from breathing play a crucial role in vascular leakage leading to edema, independent of circulating immune cells. Furthermore, the model identified potential therapeutic candidates, angiopoietin-1 (Ang-1) and a transient receptor potential vanilloid 4 (TRPV4) ion channel inhibitor (GSK2193874), for preventing this IL-2-induced toxicity.

5. Pulmonary thrombosis

Pulmonary thrombosis, a significant contributor to mortality in various lung diseases, initiates with platelet activation, often triggered by inflammatory stimulation of the lung vascular endothelium. Jain et al. developed a microfluidic lung-on-a-chip model with a vascular channel lined with vascular endothelial cells that allowed perfusion with human whole blood.¹⁵² This model successfully recapitulated in vivo platelet-endothelial dynamics and demonstrated that lipopolysaccharide (LPS)-induced thrombosis is mediated indirectly by alveolar epithelial activation rather than direct endothelial activation. The utility of the chip for drug testing was demonstrated by analyzing the inhibitory effect of a PAR-1 antagonist on endothelial activation and thrombosis.

6. Smoking lung injury

Cigarette smoking is a leading cause of lung diseases and a major contributor to cancer-related mortality worldwide.^165–167 To investigate the pathogenic mechanisms of cigarette smoke-induced lung injury and explore potential therapeutic interventions, various lung-on-a-chip platforms have been developed. Shrestha et al. developed a chip using cigarette smoke extract (CSE) to simulate the effects of smoking in vitro.¹⁵³ The suitability of the chip for permeability assays, toxicological testing, and pulmon,ary drug delivery studies was further confirmed by examining the effects of CSE on IL-6 and IL-8 release and E-cadherin expression in Calu-3 cells. CSE treatment increased IL-6 and IL-8 secretion and decreased E-cadherin expression, effects that were partially reversed by budesonide. Benam et al. developed a biomimetic smoking robot coupled with a lung-on-a-chip platform.¹⁵⁴ The robot delivers fresh cigarette smoke to an epithelium-lined microchannel in a cyclic inhalation–exhalation pattern while applying physiological shear stress and cyclic strain. Genome-wide microarray analysis revealed quantitative similarities in oxidation–reduction pathway gene expression between acute exposure results from the device and pathology samples from human smokers. A novel methodology was developed by Li et al. that combines a custom-designed bionic lung chip with transcriptome analysis to evaluate the effects of on-line cigarette smoke exposure on bromodomain-containing protein 2B (BEAS-2B) cells.¹⁵⁵ By performing in situ cigarette smoke exposure within the microfluidic chip, followed by in situ cell lysis and RNA-Seq, the study identified the significant involvement of the mitogen-activated protein kinase (MAPK) pathway in CS-induced inflammatory responses. This finding was corroborated by in situ fluorescence validation of inflammatory biomarkers and further confirmed by experiments demonstrating the anti-inflammatory effects of dexamethasone (DEX) within the chip.

7. Clinical applications

Lung-on-a-chip devices have shown significant potential in bridging the gap between preclinical research and clinical applications, particularly in the context of drug development and personalized medicine.^168,169 These microfluidic platforms provide a more accurate and human-relevant model compared to traditional 2D cell culture and animal models, which has led to their increasing use in clinical trials and patient care pathways. For example, they have been used to evaluate the efficacy and toxicity of chemotherapeutic agents for lung cancer, which have been validated in clinical trials.¹³⁹ Lung-on-a-chip models have also been used in respiratory clinical trials, such as COVID-19, which mimics the air–liquid interface and mechanical forces of the lung to study viral pathogenesis and screen antiviral drugs. In personalized medicine, lung-on-a-chip devices have been integrated into patient care pathways for diseases such as IPF, using patient-derived cells to predict individual responses to antifibrotic therapies.¹⁴⁹ In addition, lung-on-a-chip devices are being explored in regenerative medicine, with clinical trials underway to develop bioengineered lung tissue for transplantation to address the shortage of donor lungs.¹⁶⁰

B. Aerosol transport and deposition

The study of respiratory particles is of great importance, which can help disease prevention, drug development, and drug delivery optimization in medicine, and provide a basis for pollution impact assessment and particle behavior studies in environmental science, which has far-reaching implications for maintaining human health and environmental safety. Many researchers have already conducted such studies.^170–172 Historically, aerosol transport studies have used scaled-up models, limiting the exploration of realistic inhaled aerosol dynamics. This is due to the difficulty of simultaneously fitting dimensionless flow and particle numbers (e.g., Reynolds and Womersley numbers).¹⁷³ Lung-on-a-chip models provide a more accurate representation of lung structure and function, enabling detailed studies of aerosol deposition. These models provide valuable insights into the pathophysiology of smoking-related diseases and potential therapeutic strategies.^174,175

Several studies have used pumps to simulate the human breathing process and to study the deposition and transport of aerosols. Fishler et al. developed a full-scale experimental model of the pulmonary alveolus (deep lung airways) to study particle transport and deposition during inhalation, as shown in Fig. 4(a).¹⁰⁷ Using bifurcating alveolar ducts with respiratory-like wall motion, the researchers experimentally tracked inhaled polydispersed smoke particles (0.2–1 μm) and monodispersed microspheres (0.1–2 μm). The experiments demonstrated the importance of particle motion (gravity and diffusion) in aerosol dispersion and deposition, particularly through a streamline crossing mechanism during flow reversal and within alveolar cavities. Lin et al. developed a spontaneous breathing lung model using multilayer fabrication technology to mimic the human lower respiratory tract, including branching morphology and deformable alveolar features, as shown in Fig. 4(b).¹⁰⁶ This model mimics cyclic and spontaneous breathing and allows inhalation and exhalation of nebulized aerosols under simulated disease conditions (obstructive and restrictive). Experiments with 4.2 μm aerosols showed deeper lung penetration (up to generation 19) in the obstructive model compared to the restrictive model (up to generation 17).

FIG. 4.

Lung-on-a-chip models for aerosol transport and deposition. (a) A full-scale experimental model of the pulmonary alveolus to study particle transport and deposition.¹⁰⁷ Reproduced with permission from Fishler et al., Sci. Rep. 5, 11 (2015); licensed under a Creative Commons Attribution (CC BY) license. (b) A breathing lung model using multilayer fabrication technology to mimic the respiratory tract.¹⁰⁶ Reproduced with permission from Lin et al., Biomicrofluidics 13(4), 044109 (2019). Copyright 2019 AIP Publishing LLC. (c) A real-scale single alveolar chip with movable walls to experimentally study micro/nanoparticle transport.¹⁰⁴ Reproduced with permission from Dong et al., Micromachines 12(2), 184 (2021); licensed under a Creative Commons Attribution (CC BY) license. (d) A multilayer microfluidic lung chip designed to accurately mimic human respiratory bronchi for particle deposition studies.¹⁷⁶ Reproduced with permission from Qiu et al., Lab Chip 23(19), 4302 (2023). Copyright 2023 RSC.

Some researchers studied particle deposition in a single alveolar model. Dong et al. developed a novel microfluidic, real-scale alveolar chip with movable walls to experimentally study micro/nanoparticle transport in human lung alveoli under rhythmic respiration, as shown in Fig. 4(c).¹⁰⁴ A new method of introducing particles in aqueous solution (rather than air) was used for improved visualization. The chip allows tracking of particle trajectories under different force conditions over multiple breathing cycles, providing higher resolution and clearer images of particle velocities. The study investigated the effects of flow patterns, drag, and gravity (including different directions of gravity) on particle transport. The results showed that while drag force contributes to reversible particle motion, gravity is the dominant factor determining particle deposition within the alveoli. Zhang et al. builds on previous work by developing a microfluidic experimental platform to study particle behavior in a rhythmically expanding alveolar chip over multiple cycles.¹⁷⁷ Their study measured alveolar flow patterns at different acinar generations with gravity oriented both parallel and perpendicular to the alveolar duct. Observations revealed that particles within the alveoli either escaped through the duct inlet or remained trapped, demonstrating irreversible particle transport. Particle deposition was more prevalent on the distal alveolar walls in earlier acinar generations. The deposition rates of particles of different sizes were also compared.

Draw inspiration and experience from the contributions of previous study, we reported a multilayer microfluidic lung chip designed to accurately mimic human respiratory bronchi, enabling controlled particle deposition studies, as shown in Fig. 4(d).¹⁷⁶ The chip uses a deformable PDMS membrane to quantitatively control fluid velocity and mimic passive human breathing, simulating the respiratory cycle. Combining time-lapse photography of fluorescent particles in liquid and a chip-aerosol exposure device with microscopy, the researchers observed particle deposition. Both experiments and numerical simulations showed a decreasing particle concentration toward the distal lung generations and demonstrated that breathing patterns significantly influence deposition. Increased particle residence time (achieved by adjusting breath-hold time, exhalation time, cycle length, and tidal volume) facilitated deeper particle deposition. This microfluidic lung chip offers a promising, efficient, cost-effective, and ethical alternative to in vivo studies for respiratory health and inhaled drug delivery research.

The complex environment within the respiratory tract, in which the mucociliary clearance system plays a critical role in particle transport and deposition, has driven considerable research into the development of in vitro models using artificial cilia.^178–183 These models aim to mimic the function of natural cilia in clearing mucus and trapped particles from the airways. We recently reported on a novel platform designed to mimic the ciliated mucus clearance system, a key defense mechanism of the respiratory tract.¹⁸⁴ We focused on the effect of artificial cilia on particle transport in this complex environment. Our system uses PDMS and iron powder to fabricate micro-cilia arrays that dynamically respond to alternating magnetic fields. Under the action of a periodically varying magnetic field, these artificial cilia exhibit an asymmetric oscillation pattern that effectively drives the microspheres to move in a directional manner in the fluid medium, mimicking the motion of mucus. We investigated the key factors influencing the particle transport efficiency, including cilia oscillation frequency, microsphere size, cilia density, and fluid viscosity. This work enhances our ability to model and study human respiratory function in vitro and contributes to the development of innovative solutions for respiratory health and microchannel cleaning.

Previous views on aerosol deposition have primarily focused on the size of the particles, considering it the dominant factor in determining where aerosols deposit within the respiratory tract. However, recent experimental evidence has further highlighted the role of aerosol hydrophobicity in determining regional deposition within the lungs. Studies by Li et al. demonstrated that the hydrophobicity of inhaled particulate matter significantly influences its deposition patterns, with more hydrophobic particles tending to be trapped in the nasal cavity, while less hydrophobic particles penetrate deeper into the alveoli.¹⁸⁵ Similarly, Dong et al. revealed that airborne fine particles can facilitate the transport of viruses deep into the lower respiratory tract and even to distant organs, emphasizing the importance of physicochemical properties, including hydrophobicity, in aerosol behavior and its implications for respiratory health.¹⁸⁶ These findings suggest that both particle size and hydrophobicity should be considered in future studies of aerosol transport and deposition, as they play critical roles in understanding the biological impacts of inhaled particles. Looking ahead, lung-on-a-chips hold great promise for advancing our understanding of aerosol deposition and its health effects.^187,188 These advanced in vitro models can simulate the complex environment of the respiratory tract, including the effects of mucociliary clearance and the role of hydrophobicity in particle deposition.

VII. FUTURE CHALLENGES AND PERSPECTIVES

Lung-on-a-chip devices are positioned as powerful tools for studying diseases such as SARS-CoV-2 (COVID-19), pulmonary fibrosis, and lung cancer. These models provide a superior platform for biomedical researchers studying lung disease and inhalation toxicity testing. However, realizing their full potential depends on overcoming several critical challenges.

A major challenge is to improve biomimicry. Many current lung-on-a-chips rely on non-biomimetic materials and oversimplified designs, limiting their physiological accuracy. The cell models at the heart of lung-on-a-chips are often established on non-biomimetic scaffolds or membranes. Future systems will need to incorporate advanced biomaterials along with modern bio-fabrication techniques, vascularization, and scaffold-free techniques to better mimic native lung structures and interactions. The incorporation of advanced features such as vascularization can lead to the development of more realistic and biomimetic lung-on-a-chips. Importantly, scaffold-free approaches that rely on the inherent potential of stem cells or patient-derived primary cells cultured on an engineered scaffold could also play a critical role in the next generation of translational lung-on-a-chips. These innovations will enhance biomimicry and expand the scope of lung-on-a-chip applications. However, materials challenges remain a key area for improvement. While common materials such as PDMS offer advantageous properties such as gas permeability and optical transparency, they can interact with small molecules and drugs, potentially skewing experimental results. Research into alternative materials with better biomimicry and chemical stability is essential to ensure the accuracy and reliability of lung-on-a-chip experiments.

Imaging limitations are also a significant obstacle. Traditional 2D imaging techniques are insufficient to capture the complexity of 3D lung-on-a-chips. As systems become more complex and three-dimensional, the development of advanced imaging techniques that can track morphological and histological changes in 3D environments in real time is essential. These techniques will provide a more complete understanding of cellular behavior and disease progression within the lung-on-a-chip environment.

The integration of multiple organ systems into a single lab-on-a-chip platform also presents challenges beyond those of single organ models. Key issues include proper scaling of organ sizes and cell numbers to ensure physiologically relevant interactions, which is critical to accurately mimic human responses. The inclusion of immune components adds complexity, as immune responses significantly influence drug efficacy and toxicity, requiring careful consideration of immune cell interactions with other tissues. Standardization across organ models is also essential to ensure reproducibility and reliability, requiring the development of uniform protocols for cell culture, microfluidic design, and data analysis, which can be challenging given the diversity of tissues involved. Additionally, the utilization of primary cells, stem cells, and patient-derived cells in lung-on-a-chip devices introduces several ethical and regulatory challenges. The procurement of patient-derived cells requires informed consent to ensure that donors are fully aware of the scope and potential implications of the research. This process must address concerns about privacy, data security, and the potential disclosure of genetic information. The use of stem cells, particularly those derived from embryonic sources, raises ethical debates about the moral status of embryos and the conditions under which they are used or destroyed. To address these issues responsibly, researchers must adhere to established ethical guidelines and regulatory frameworks to ensure that all cell sources are obtained and used with respect for donor rights and societal ethical standards.

In conclusion, interdisciplinary collaboration is critical to the advancement of lung-on-a-chip technologies, which inherently combine principles from engineering, biology, and medicine. Integrating microengineering techniques with cell biology and clinical knowledge can greatly enhance the development and application of lung-on-a-chips. For example, engineers can design microfluidic systems that accurately replicate the mechanical and structural complexity of the human lung, while biologists can contribute their understanding of cellular behavior and interactions within these systems. Clinicians, on the other hand, provide critical insight into the physiological relevance and clinical applicability of these models, ensuring that the lung-on-a-chips address real-world challenges in pulmonary medicine. Collaborations with computational scientists are also essential, as they develop algorithms for modeling fluid dynamics and simulating biological processes that can predict how lung-on-a-chips will behave under various conditions without extensive physical experimentation. In addition, partnerships with industry can facilitate the scaling and standardization of these technologies, making them more accessible for widespread clinical and research use. Such interdisciplinary efforts are critical to overcoming current limitations and realizing the full potential of lung-on-a-chips as transformative tools in respiratory disease research and therapeutic development.

AUTHOR DECLARATIONS

Conflict of Interest

The authors declare no competing interests.

Author Contributions

Yan Qiu: Data curation (lead); Investigation (lead); Methodology (equal); Visualization (lead); Writing – original draft (lead). Guoqing Hu: Conceptualization (lead); Funding acquisition (lead); Project administration (lead); Resources (lead); Supervision (lead); Writing – review & editing (lead).

DATA AVAILABILITY

Data sharing is not applicable to this article as no new data were created or analyzed in this study.