Lung-on-a-Chip: Models & Applications

What Is a Lung-on-a-Chip?

A lung-on-a-chip is a microfluidic cell culture device that reconstitutes selected structural and functional features of human lung tissue.

In a common alveolar-capillary design, epithelial and endothelial cells are cultured on opposite sides of a porous membrane, with an air-liquid interface on the epithelial side and perfused medium on the vascular side; mechanically active models can add cyclic strain to mimic breathing [1].

In the canonical alveolar design, two microchannels are separated by a thin porous membrane carrying the epithelial-endothelial co-culture. The apical compartment can be exposed to air or aerosol, while the vascular channel is perfused with medium to provide nutrients and controlled shear. A general two-channel lung-on-a-chip architecture is illustrated in Figure 1.

Figure 1. General architecture of a two-channel lung-on-a-chip. (A) Cross-section: an epithelial layer at the air-liquid interface sits on a porous membrane, an endothelial layer and perfused medium occupy the channel below, and lateral vacuum chambers apply cyclic strain to reproduce breathing-related deformation. (B) Detail of the barrier, showing solute transport through the membrane pores. The two-channel architecture with lateral actuation chambers was introduced by Huh et al. [1].

The dimensions the device is trying to reproduce are demanding. Morphometry of the normal human lung gives an alveolar surface area of roughly 143 m² and an air-blood barrier with an arithmetic mean thickness of about 2.2 µm [3].

For diffusion calculations, the harmonic mean thickness is the relevant morphometric quantity and is near 0.6 µm in the same human dataset [3]; the stereological basis for distinguishing arithmetic and harmonic means was established in earlier morphometric work [4]. Membrane thickness and pore geometry are therefore important design parameters rather than incidental fabrication details.

This organ-specific configuration sits inside the broader field of microphysiological systems. For the underlying technology, cell sourcing strategies and readout methods common to every organ model, see Fluigent’s review of organ-on-chip technology.

Reconstituting Organ-Level Lung Function on a Chip

The phrase describes the founding result of the field. In 2010, Huh and colleagues published a microdevice that reproduced the alveolar-capillary interface and, critically, showed that it responded at the level of an organ rather than a monolayer; recruiting immune cells, mounting inflammatory responses, and transporting particulates across the barrier [1].

A later implementation of this epithelial-endothelial barrier using a collagen-elastin membrane is shown in Figure 2 [5].

Figure 2. Air-blood barrier reproduction in a stretchable alveolar array: confocal co-culture of primary human alveolar epithelial cells (E-cadherin, green) with primary endothelial cells (red), plus a TEM cross-section of the epithelial/endothelial layers across the collagen-elastin membrane. From Zamprogno et al. [5], Communications Biology (2021), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

The device reproduced integrated organ-level responses to bacteria and inflammatory cytokines introduced into the alveolar space. Circulating neutrophils adhered to activated endothelium, transmigrated across the capillary-alveolar interface, migrated toward bacteria and phagocytosed them under real-time microscopy [1]. This sequence illustrates the integrated immune-cell recruitment and pathogen-response behavior that the chip was designed to capture.

In nanotoxicology experiments, cyclic mechanical strain increased the toxic and inflammatory response to silica nanoparticles, enhanced epithelial and endothelial uptake, and increased transport into the microvascular channel; comparable effects of breathing on nanoparticle absorption were observed in an ex vivo whole-mouse-lung model [1]. These findings show that mechanical stimulation can materially change the measured biological response.

The same platform was later used to reproduce drug toxicity-induced pulmonary edema, where breathing motions again proved central to the development of vascular leakage [6].

The Air-Liquid Interface in Lung-on-a-Chip Models

An air-liquid interface (ALI) is created by removing medium from the apical compartment once the epithelium is confluent, leaving cells exposed to air while nutrients arrive from the basal side through the porous membrane.

ALI culture drives differentiation toward the in vivo phenotype: primary human airway epithelial cells form a polarized, pseudostratified epithelium of ciliated and mucus-secreting cells that submerged culture does not produce [7]. A representative primary alveolar epithelial ALI culture expressing adherens and tight-junction markers is shown in Figure 3 [5].

Figure 3. Primary human alveolar epithelial cells cultured on a collagen-elastin membrane after 4 days and 2 days at air-liquid interface, stained for E-cadherin (red), ZO-1 (green), and nuclei (Hoechst, blue). From Zamprogno et al. [5], Communications Biology (2021), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

ALI differentiation requires sustained culture. Transcriptional profiling of human bronchial epithelial cells across a 28-day ALI time course identified more than 2,000 genes changing at least two-fold, with cilia formation, adhesion, immunity and transport among the dominant programs [7].

In practice, ALI differentiation protocols commonly extend over several weeks and vary with cell source and medium, so long-term sterility and stable culture conditions are important.

Be-Transflow (OOC-TRNS-07, manufactured by Beonchip) is designed for ALI and barrier co-culture: a culture well is connected to a perfusable microfluidic channel through a 1 µm porous membrane. Epithelial cells can be cultured in the upper well and endothelial cells in the lower channel.

The channel is 375 µm high, 1.5 mm wide and 45 mm long, with a listed 44 µL volume; the wells are positioned on standard 96-well plate locations for automated microscopy [8] (Figure 4).

Figure 4. Be-Transflow configuration for air-liquid-interface and barrier models. (A) Be-Transflow device. (B) Schematic showing an epithelial culture in the upper compartment and an endothelial culture in the perfused channel below, separated by a porous membrane. Source: Fluigent/Beonchip, Be-Transflow [8].

Material choice is important in drug-exposure experiments. Be-Transflow is made from lipophobic thermoplastic rather than PDMS, reducing the nonspecific absorption concern associated with PDMS-based devices [8].

PDMS has been shown to absorb small hydrophobic molecules and deplete them from solution [9]; in organ-on-chip drug-response assays, the extent of absorption is time-dependent and is not predicted by hydrophobicity alone [10].

Where a model requires perfusion on both sides of the barrier – for example for circulating immune or tumor cells, transport studies, or controlled gas conditions – Be-Doubleflow (OOC-DBLE-08, manufactured by Beonchip) provides two perfusable channels separated by a porous membrane. Each channel is 375 µm high, 1.5 mm wide and 45 mm long, with a listed 31.2 µL volume.

The standard specification lists a 1 µm membrane, while the product description also notes customizable pore-size options [11].

Common barrier readouts include trans- or transepithelial electrical resistance (TEER), tracer-permeability assays and tight-junction immunostaining; the exact combination depends on the cell model and device. These approaches are demonstrated in mechanically active lung-chip studies using molecular tracers, junctional staining and TEER measurements [12,13].

Breathing Motion and Mechanical Stimulation

Breathing-related mechanics are reproduced by cyclically deforming the membrane or tissue-supporting layer using controlled pneumatic actuation. The original Huh design and later lung-chip arrays commonly used approximately 10% linear strain at 0.2 Hz, corresponding to 12 cycles per minute [1,12].

Published lung-chip studies place physiological alveolar strain in roughly the 5-15% range [14], and a 10% linear stretch therefore sits within that range [12].

Strain magnitude is biologically relevant: in Stucki et al., physiological cyclic strain increased transport of a small hydrophilic tracer by about 46% without a significant increase in 70 kDa dextran permeability, indicating altered small-molecule transport without loss of gross barrier integrity [12].

Two actuation geometries are common. In the original configuration, vacuum applied to side chambers stretches the membrane in-plane [1]. In diaphragm-based systems, pneumatic actuation of a microdiaphragm is transmitted through the fluidic compartment to deflect the culture membrane out of plane, generating three-dimensional cyclic strain [12].

These geometries require a pressure or vacuum source with sufficient dynamic response to reproduce the desired waveform; some custom systems also benefit from bidirectional pressure control. Fluigent’s Mechanical Stimulation Pack combines three LineUp Push-Pull controllers, a LINK module, a Flow Unit M and a Flow EZ, with OxyGEN for protocol automation and data recording [15]. The complete configuration is illustrated in Figure 5.

Figure 5. Fluigent Mechanical Stimulation Pack schematic showing Flow EZ, Push-Pull controllers, FLOW UNIT, LINK/OxyGEN and a mechanically actuated microfluidic device. Source: Fluigent, Mechanical Stimulation Pack [15].

Lung-on-a-Chip Applications

Four application areas account for most published work. Each exploits a different property of the device; the air interface, the mechanical cycling, the vascular channel, or the human cell source.

Inhaled drug and aerosol testing

Inhaled compounds or aerosols can be delivered to an epithelial surface maintained at ALI and then assessed for epithelial response and transport toward the basal compartment. Mechanically active lung-chip studies also show that cyclic strain can modify epithelial transport [12]. For compound-testing workflows, material choice is important because PDMS can sequester small hydrophobic molecules [9,10].

Respiratory infection models

Respiratory-infection models introduce pathogens or pathogen surrogates into the epithelial compartment while immune or vascular components are maintained in the opposing channel. The 2010 lung chip directly visualized neutrophil recruitment, transmigration and bacterial phagocytosis [1], while later alveolar-chip models have also been used with viral surrogates such as SARS-CoV-2 pseudovirus [14].

Figure 6. Example dual-channel configuration for barrier studies requiring perfusion on both sides. (A) Be-Doubleflow device. (B) Schematic of a perfused barrier with circulating cells or particles. Source: Fluigent/Beonchip, Be-Doubleflow [11].

Toxicology and air-pollutant exposure

Toxicology models expose the epithelial side to airborne or particle-associated insults under controlled ALI and mechanical conditions. In the original lung chip, cyclic strain increased silica-nanoparticle uptake and inflammatory responses [1]; later alveolar-chip work has also incorporated cigarette-smoke exposure [14]. These studies show why static and mechanically active models should not be assumed to yield identical responses.

Lung cancer models

Lung cancer-on-chip models combine tumor cells with epithelial, endothelial, stromal or immune components under controlled microfluidic conditions to study tumor growth, invasion and therapeutic response [16]. Dual-channel barrier devices are particularly relevant when vascular interactions or circulating tumor cells are part of the experimental question [11,16].

Case Study: A Breathing Lung-on-a-Chip for Drug Testing

Researchers at the ARTORG Center for Biomedical Engineering, University of Bern, with AlveoliX AG, developed a microengineered alveolar array in which a suspended collagen-elastin membrane supported by a hexagonal scaffold is pneumatically deflected to reproduce breathing-like motion [13]. The primary article reports a stable membrane deflection of 43 µm and a maximum surface-area strain of 7.8%.

TGF-β1-induced extracellular-matrix gene upregulation reached statistical significance under cyclic stretch, and nintedanib showed clearer protective effects on barrier integrity under dynamic than static conditions, while also reducing fibrotic markers [13]. The experimental setup used Fluigent Flow EZ (-800 mbar) control for cyclic stretch, PushPull control for valve actuation, and OxyGEN software [13,17]. The chip architecture and pneumatic breathing mechanism are shown in Figure 7.

schematic of Lung on chip system and array of alveoli

Figure 7. Microengineered alveolar-array system used for breathing-like mechanical stimulation: chip architecture, diaphragm-driven actuation and valve principle. Adapted from Weber et al. [13], Biofabrication (2026), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/), ARTORG Center, University of Bern and AlveoliX AG; the same work is presented in the related case study [17].

Full details: breathing lung-on-chip case study.

Equipment for Lung-on-a-Chip Experiments

A mechanically active ALI lung-on-a-chip typically combines a barrier-capable chip, controlled perfusion of the vascular compartment, pneumatic actuation when breathing motion is required, and a sterile fluidic layout suitable for long-term cell culture. Simpler lung-chip models may omit mechanical actuation when it is not part of the biological question.

Perfusion and mechanical actuation impose different fluid-control requirements. Perfusion benefits from stable, low-pulsatility flow, whereas pneumatic actuation requires a reproducible pressure or vacuum waveform.

Pressure-based control can support both types of operation, although the controller configuration depends on the chip; Flow EZ provides pressure/vacuum control and can be paired with Flow Unit feedback for closed-loop flow-rate control [18].

For automated organ-on-chip workflows, Omi supports controlled perfusion, recirculation, sampling and injection, with pressure, flow-rate and shear-stress control in an incubator-compatible platform [19]. For a more conventional external-controller setup, the Organ on Chip Perfusion Pack combines pressure-based flow control, a bidirectional flow sensor and OxyGEN automation for stable chip perfusion [20]. An example of the Omi platform is shown in Figure 8.

omi recirculation sampling and injection under flow

Figure 8. Omi automated organ-on-chip platform for perfusion, recirculation, sampling and injection under controlled flow and pressure conditions. Source: Fluigent [19].

If you are specifying a first system, Fluigent’s 5 key tips for starting organ-on-chip models covers the practical decisions; cell sourcing, coating, bubble management and flow rate selection, that determine whether the first experiments succeed.

Frequently Asked Questions

What is a lung-on-a-chip used for?

Lung-on-a-chip models are used for inhaled-drug and toxicity studies, respiratory-infection research, particle toxicology, fibrosis and cancer modeling. Their value is that human cells can be studied under controlled air-liquid-interface, perfusion and, where relevant, breathing-like mechanical conditions; they can complement conventional culture and help reduce reliance on some animal experiments [1,6].

How is an air-liquid interface created on a chip?

Cells are seeded in the apical compartment and first expanded under submerged conditions. Once the epithelium is established, apical medium is removed so that the epithelial surface is exposed to air while nutrients are supplied from the basal side. Differentiation then proceeds over a model-dependent period; the cited bronchial epithelial dataset follows a 28-day ALI time course [7].

How is breathing motion reproduced in vitro?

By cyclically deforming the membrane or tissue-supporting layer using controlled pneumatic actuation. Common published lung-chip conditions include approximately 10% linear strain at 0.2 Hz, within the physiological range used in alveolar models [1,12,14]. The exact actuation geometry and strain definition depend on the device.

Which chip is best for lung-on-a-chip culture?

Be-Transflow (manufactured by Beonchip) suits ALI and barrier culture with a perfused channel beneath an upper culture well; its standard membrane pore size is 1 µm [8]. Be-Doubleflow (manufactured by Beonchip) provides two perfusable channels for models where flow on both sides is important; its current standard specification also lists a 1 µm membrane, while the product description notes customizable pore-size options [11]. Both are compatible with Fluigent pressure-based flow controllers [8,11].

Related Expertises

References

1. Huh D, Matthews BD, Mammoto A, Montoya-Zavala M, Hsin HY, Ingber DE. Reconstituting organ-level lung functions on a chip. Science. 2010;328(5986):1662–1668. doi:10.1126/science.1188302

2. Wang Y, Marucci L, Homer ME. In silico modelling of organ-on-a-chip devices: an overview. Frontiers in Bioengineering and Biotechnology. 2025;12:1520795. doi:10.3389/fbioe.2024.1520795 (cited for context; no longer a figure source.)https://doi.org/10.3389/fbioe.2024.1520795

3. Gehr P, Bachofen M, Weibel ER. The normal human lung: ultrastructure and morphometric estimation of diffusion capacity. Respiration Physiology. 1978;32(2):121–140. doi:10.1016/0034-5687(78)90104-4

4. Weibel ER, Knight BW. A morphometric study on the thickness of the pulmonary air-blood barrier. The Journal of Cell Biology. 1964;21(3):367–384. doi:10.1083/jcb.21.3.367

5. Zamprogno P, Wüthrich S, Achenbach S, et al. Second-generation lung-on-a-chip with an array of stretchable alveoli made with a biological membrane. Communications Biology. 2021;4:168. doi:10.1038/s42003-021-01695-0

6. Huh D, Leslie DC, Matthews BD, Fraser JP, Jurek S, Hamilton GA, Thorneloe KS, McAlexander MA, Ingber DE. A human disease model of drug toxicity-induced pulmonary edema in a lung-on-a-chip microdevice. Science Translational Medicine. 2012;4(159):159ra147. doi:10.1126/scitranslmed.3004249

7. Ross AJ, Dailey LA, Brighton LE, Devlin RB. Transcriptional profiling of mucociliary differentiation in human airway epithelial cells. American Journal of Respiratory Cell and Molecular Biology. 2007;37(2):169–185. doi:10.1165/rcmb.2006-0466OC

8. Fluigent. Be-Transflow: Air-Liquid Interface and Co-Culture Chip (OOC-TRNS-07). Product page and technical specifications. https://www.fluigent.com/research/instruments/microfluidic-chips/cell-culture-organ-on-a-chip-microscopy/be-transflow/ Accessed 24 August 2026.

9. Toepke MW, Beebe DJ. PDMS absorption of small molecules and consequences in microfluidic applications. Lab on a Chip. 2006;6(12):1484–1486. doi:10.1039/b612140c

10. van Meer BJ, de Vries H, Firth KSA, van Weerd J, Tertoolen LGJ, Karperien HBJ, et al. Small molecule absorption by PDMS in the context of drug response bioassays. Biochemical and Biophysical Research Communications. 2017;482(2):323–328. doi:10.1016/j.bbrc.2016.11.062

11. Fluigent. Be-Doubleflow: Dual-Channel Microfluidic Cell Culture Chip (OOC-DBLE-08). Product page and technical specifications. https://www.fluigent.com/research/instruments/microfluidic-chips/cell-culture-organ-on-a-chip-microscopy/be-doubleflow/ Accessed 24 August 2026.

12. Stucki AO, Stucki JD, Hall SRR, Felder M, Mermoud Y, Schmid RA, Geiser T, Guenat OT. A lung-on-a-chip array with an integrated bio-inspired respiration mechanism. Lab on a Chip. 2015;15(5):1302–1310. doi:10.1039/C4LC01252F

13. Weber TA, Zamprogno P, Schneider S, Hajari MA, Büchler P, Hobi N, et al. Microengineered alveolar array lung-on-chip with hydrogel membrane and simulated breathing mechanics for anti-fibrotic testing. Biofabrication. 2026;18(1):015026. doi:10.1088/1758-5090/ae2e44

14. Huang D, Liu T, Liao J, Maharjan S, Xie X, Pérez M, et al. Reversed-engineered human alveolar lung-on-a-chip model. Proceedings of the National Academy of Sciences. 2021;118(19):e2016146118. doi:10.1073/pnas.2016146118

15. Fluigent. Mechanical Stimulation Pack. Product page and package specifications. https://www.fluigent.com/research/instruments/packages/application-packages/mechanical_stimulation_package/ Accessed 24 August 2026.

16. Li L, Bo W, Wang G, Juan X, Xue H, Zhang H. Progress and application of lung-on-a-chip for lung cancer. Frontiers in Bioengineering and Biotechnology. 2024;12:1378299. doi:10.3389/fbioe.2024.1378299

17. Fluigent. Breathing Lung-on-Chip Platform for Dynamic Anti-Fibrotic Testing. Customer case study based on Weber et al. (2026). https://www.fluigent.com/resources-support/expertise/customer-case-studies/breathing-lung-on-chip-platform/ Accessed 24 August 2026.

18. Fluigent. Flow EZ™ microfluidic pressure and flow controller. Product page and specifications. https://www.fluigent.com/research/instruments/pressure-flow-controllers/flow-ez/ Accessed 24 August 2026.

19. Fluigent. Omi, an Automated Organ-On-A-Chip Platform. Product page and technical specifications. https://www.fluigent.com/research/instruments/automated-organ-on-chip-platform/ Accessed 24 August 2026.

20. Fluigent. Organ on Chip Perfusion Pack. Product page and package specifications. https://www.fluigent.com/research/instruments/packages/application-packages/cell_perfusion_package/ Accessed 24 August 2026.

Lung-on-a-Chip: Models, Air-Liquid Interface and Applications