Here, we developed a human aorta smooth muscle cell organ-on-a-chip model to replicate the in vivo biomechanical strain of smooth muscle cells in the human aortic wall.
Conventional two-dimensional cell culture techniques and animal models have been used in the study of human thoracic aortic aneurysm and dissection (TAAD). However, human TAAD sometimes cannot be characterized by animal models. There is an apparent species gap between clinical human studies and animal experiments that may hinder the discovery of therapeutic drugs. In contrast, the conventional cell culture model is unable to simulate in vivo biomechanical stimuli. To this end, microfabrication and microfluidic techniques have developed greatly in recent years, providing novel techniques for establishing organoids-on-a-chip models that replicate the biomechanical microenvironment. In this study, a human aorta smooth muscle cell organ-on-a-chip (HASMC-OOC) model was developed to simulate the pathophysiological parameters of aortic biomechanics, including the amplitude and frequency of cyclic strain experienced by human aortic smooth muscle cells (HASMCs) that play a vital role in TAAD. In this model, the morphology of HASMCs became elongated in shape, aligned perpendicularly to the strain direction, and presented a more contractile phenotype under strain conditions than under static conventional conditions. This was consistent with the cell orientation and phenotype in native human aortic walls. Additionally, using bicuspid aortic valve-related TAAD (BAV-TAAD) and tricuspid aortic valve-related TAAD (TAV-TAAD) patient-derived primary HASMCs, we established BAV-TAAD and TAV-TAAD disease models, which replicate HASMC characteristics in TAAD. The HASMC-OOC model provides a novel in vitro platform that is complementary to animal models for further exploring the pathogenesis of TAAD and discovering therapeutic targets.
Thoracic aortic aneurysm and dissection (TAAD) is a localized dilatation or delamination of the aortic wall that is associated with high morbidity and mortality1. Human aortic smooth muscle cells (HASMCs) play a vital role in the pathogenesis of TAAD. HASMCs are not terminally differentiated cells, and HASMCs retain high plasticity, allowing them to switch phenotypes in response to different stimuli2. HASMCs are mainly subjected to rhythmic tensile strain in vivo, and this is one of the key factors regulating smooth muscle morphological changes, differentiation and physiological functions3,4. Therefore, the role of cyclic strain in the study of HASMCs cannot be ignored. However, conventional 2D cell cultures cannot replicate the biomechanical stimulation of cyclic strain experienced by HASMCs in vivo. Additionally, the construction of an animal TAAD model is not suitable for some types of TAAD, such as bicuspid aortic valve (BAV)-related TAAD. Moreover, the species gap between clinical human studies and animal experiments cannot be ignored. It hinders pharmaceutical translation in clinical practice. Thus, there is an urgent need for more complex and physiological systems to simulate the in vivo biomechanical environment in the research of aortic diseases.
Animal experiments used in biomedical research and drug development are costly, time-consuming, and ethically questionable. In addition, the results from animal studies frequently fail to predict the results obtained in human clinical trials5,6. The lack of human preclinical models and high failure rate in clinical trials have resulted in few effective drugs for the clinic, which has increased health care costs7. Thus, it is urgent to find other experimental models to complement animal models. Microfabrication and microfluidic techniques have developed greatly in recent years, providing novel techniques for establishing organoids-on-a-chip models that remedy the drawbacks of traditional 2D cell culture techniques and establish a more realistic, low-cost, and efficient in vitro model for physiological studies and drug development. Using microfluidic devices, organs-on-chips are established to culture living cells in micrometer-sized chambers with different stimuli to replicate the key functions of a tissue or organ. The system consists of single or multiple microfluidic microchannels, with either one kind of cell cultured in a perfused chamber replicating functions of one tissue type or different cell types cultured on porous membranes to recreate interfaces between different tissues. Microfluidic-based organoids combined with patient-derived cells have the unique advantage of bridging the large species difference between mouse and human disease models and overcoming the disadvantages of traditional 2D cell culture for disease mechanism research and drug discovery. With the rapid development of microfluidics in the past few years, researchers have realized the usefulness of in vitro organ-on-a-chip (OOC) models replicating complex in vivo biological parameters8. These microfluidic organoids simulate in vitro biomechanical environments, such as cyclic strain, shear stress, and liquid pressure, providing a three-dimensional (3D) cell culture environment. To date, several OOC models have been established to simulate biomechanical stimuli in organs such as the lung9, kidney10, liver11, intestine12, and heart13, but these have not been widely applied to the study of human aortic disease.
In this study, we present a human aorta smooth muscle cell organ-on-a-chip (HASMC-OOC) model that can control the biomimetic mechanical forces and rhythms applied to TAAD patient-derived primary HASMCs. The chip consists of three-layered thick plates of polydimethylsiloxane (PDMS) etched with channels and two commercialized highly flexible PDMS membranes. HASMCs are cultured on the PDMS membranes. The channel in the middle of the chip is filled with a culture medium for cell culture. The top and bottom channels of the chip are connected to a vacuum pressure supply system that can control the rhythm and frequency of mechanical tensile strain of the PDMS membranes. Rhythmic strain experienced by HASMCs can be simulated in HASMC-OOC, replicating the biomechanical microenvironment of tissue or organ not functionally achievable with conventional 2D culture systems. With the advantage of high-resolution, real-time imaging, and biomechanical microenvironment, the biochemical, genetic and metabolic activities of living cells can be studied for tissue development, organ physiology, disease etiology, molecular mechanisms and biomarker identification ,cardiovascular disease and aortic disease. Combined with tissue-specific and patient cells, this system can be used for drug screening, personalized medicine, and toxicity testing. This HASMC-OOC model provides a novel in vitro platform for studying the pathogenesis of the aortic disease.
Human aortic specimens were utilized for primary HASMC isolation under the approval of Zhongshan Hospital, Fudan University Ethics Committee (NO. B2020-158). Aortic specimens were collected from patients who underwent ascending aorta surgery at Zhongshan Hospital, Fudan University. Written informed consent was obtained from all patients before participation.
1. Primary human aortic smooth muscle cell isolation
2. Preparation of PDMS chip
3. PDMS chip surface treatment and sterilization
4. Cell seeding on the PDMS chip and cell stretching process
5. Preparation of the mechanical control system
The HASMC-OOC model consists of a vacuum control system, a circulating system, and PDMS chips, and the schematic design of the HASMC-OOC model (Figure 1). The vacuum control system consists of a vacuum pump, solenoid valves, and a PLC controller. To act as the circulating system, a peristaltic pump was used to refresh the cell culture medium and add drugs. The PDMS chip was composed of two vacuum chambers and a middle chamber filled with SMCM for cell growth. According to the design of the chip structure, the PMMA mold was prepared, and the three PDMS slabs were prepared by pouring into molds and cross-linking at 70 °C for 2 h. After treatment with plasma, the three PDMS slabs and two PDMS membranes were bonded together (Figure 2). When the PDMS chips were prepared, the HASMCs were cultured on the PDMS membrane in the chip, and the PDMS chips were connected to the vacuum control system and the circulating system. A schematic of the entire system is shown in Supplementary Figure 1. The PDMS chip parameters are shown in Figure 3A, and the mechanical properties of the PDMS membrane are shown in Figure 3B. To quantify the tensile strains of the PDMS membrane, we captured the real-time deformations of the PDMS membranes from a cross-sectional view of the microfluidic model, with vacuum pressures of 0 kPa, 10 kPa, 15 kPa, 20 kPa, 25 kPa, and 30 kPa. We also measured the changes in length to evaluate the strain magnitude of the PDMS membrane in different culturing channel widths of 2 mm, 4 mm, and 6 mm (Figure 3C–E). Microchannels 6 mm in width were used in the PDMS chip. The results showed that a vacuum pressure of 10 kPa induced 7.18 ± 0.44% strain and 15 kPa induced 17.28 ± 0.91% strain (Figure 3E).
To verify the cell viability of the HASMC cell line (CRL1999) in the PDMS chip, a LIVE/DEAD assay was performed on days 3 and 5 after the cells were cultured. The results showed that the cell viability was higher than 90% at day 3 and day 5 (Figure 4A). Cytoskeleton (F-actin) staining of CRL1999 in the PDMS chip showed normal cell morphology and cells aligned perpendicularly to the strain after stretching for 24 h (Figure 4B–C). Immunofluorescence staining of the contractile markers SM22 and CNN1 was performed using HASMC-OOC after 24 h of rhythmic strain (Figure 4D–E). To start, 2 mL of PBS was slowly poured into the channel and discarded with a 2 mL syringe. Subsequently, 2 mL of 4% (v/v) paraformaldehyde was poured into the channel and incubated for 30 min at room temperature. After this, 2 mL of 0.2% (v/v) Triton X-100 was poured into the channel by a 2 mL syringe and incubated for 15 min at room temperature. This was followed by 2 mL of PBS being slowly poured into the channel and discarded by a 2 mL syringe after washing the cells. Next, 2 mL of 5% (w/v) bovine serum albumin was slowly poured into the channel and incubated for 30 min at room temperature. Afterward, 1 mL of SM22 or CNN1 antibody was slowly poured into the channel and incubated overnight at 4 °C. After incubation, 1 mL of goat anti-rabbit secondary antibody was slowly poured into the channel and incubated for 1 h at room temperature. Finally, 1 mL of DAPI solution was slowly poured into the channel and incubated for 10 min at room temperature. The results showed that the fluorescence intensity of SM22 and CNN1 under strain conditions was higher than that under static conditions (Figure 4F). Compared with static conditions, the SM22 and CNN1 genes in HASMCs were upregulated in strain conditions (Figure 4G). These data suggested that CRL1999 aligned perpendicularly to the strain direction and that a contractile phenotype was more pronounced under strain conditions than under conventional cell culturing under static conditions.
Using bicuspid aortic valve-related TAAD (BAV-TAAD) and tricuspid aortic valve-related TAAD (TAV-TAAD) patient-derived primary HASMCs, we established BAV-TAAD and TAV-TAAD disease models. The results of F-actin staining of BAV-TAAD and TAV-TAAD patient-derived primary HASMCs showed normal cell morphology and cells aligned perpendicularly to the strain direction after stretching for 24 h (Figure 5A-B). Immunofluorescence staining showed that SM22 and CNN1 expression in BAV-TAAD and TAV-TAAD patient-derived primary HASMCs was higher under strain conditions than under static conditions (Figure 5C–G,I). The gene expression levels of SM22 and CNN1 in BAV-TAAD and TAV-TAAD patient-derived primary HASMCs were upregulated under strain conditions in contrast to static conditions (Figure 5H,J). These results in BAV-TAAD and TAV-TAAD patient-derived primary HASMCs in the PDMS chip were consistent with the results from HASMC cell line CRL1999, indicating that the combination of patient-derived primary HASMCs and the HASMC-OOC model replicate HASMC characteristics in TAAD, which provides a BAV-TAAD and TAV-TAAD in vitro disease model for further investigation of disease's molecular mechanisms and drug screening.
Figure 1: Schematic design of the HASMC-OOC system. The system consists of a vacuum control system, a circulating system, and PDMS chips. Please click here to view a larger version of this figure.
Figure 2: PDMS chip preparation procedure. The PDMS gel was poured into a PMMA mold and cross-linked at 70 °C for 1-2 h. Then, the three PDMS slabs and two PDMS membranes were treated with plasma for 5 min and bonded together. Finally, the prepared PDMS chips were sterilized, and a concentration of 2 x 105 cells/mL HASMC suspension was slowly poured into the medium microchannel of the PDMS chip. Please click here to view a larger version of this figure.
Figure 3: The detailed parameters and mechanical properties of the PDMS chip. (A) The detailed parameters of the PDMS chip. The PDMS slab size was 100 mm × 40 mm × 6 mm, and the microchannel size was 70 mm × 6 mm × 4 mm. (B) The parameters of the commercial PDMS membrane. The calculated stretching amplitude of the PDMS membrane at different vacuum pressures for microchannel widths of 2 mm (C), 4 mm (D), and 6 mm (E). The data shows mean ± standard deviation (SD). Please click here to view a larger version of this figure.
Figure 4: Cell morphology, orientation, and phenotype alteration of HASMCs under cyclic strain conditions. (A) Representative image of the LIVE/DEAD assay at day 3 and day 5 after HASMC cell line (CRL1999) culturing on a PDMS chip. (B) F-actin staining of CRL1999 under static and strain conditions. (C) The cell orientation of CRL1999 under static and strain conditions. Representative image of the immunofluorescence staining of SM22 (D) and CNN1 (E) in CRL1999 under static and strain conditions. (F) The relative intensity of immunofluorescence staining of SM22 and CNN1. (G) SM22 and CNN1 mRNA expression in CRL1999 under static and strain conditions. n = 3, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, two-tailed Student's t-tests were used to compare the two groups. The data shows mean ± SD. Please click here to view a larger version of this figure.
Figure 5: Cell morphology and phenotypic alterations in BAV-TAAD and TAV-TAAD patient-derived HASMC-OOC. Representative image of F-actin staining of BAV-TAAD (A)-derived primary HASMCs and TAV-TAAD (B)-derived primary HASMCs under static and strain conditions. SM22 immunofluorescence staining in BAV-TAAD (C)-derived primary HASMCs and TAV-TAAD (D)-derived primary HASMCs under static and strain conditions. Representative image of the immunofluorescence staining of CNN1 in BAV-TAAD (E)-derived primary HASMCs and TAV-TAAD (F)-derived primary HASMCs under static and strain conditions. (G) The relative intensity of SM22 and CNN1 immunofluorescence staining in BAV-TAAD-derived primary HASMCs. (H) SM22 and CNN1 mRNA expression in BAV-TAAD-derived primary HASMCs under static and strain conditions. (I) The relative intensity of SM22 and CNN1 immunofluorescence staining in TAV-TAAD-derived primary HASMCs. (J) SM22 and CNN1 mRNA expression in TAV-TAAD-derived primary HASMCs under static and strain conditions. n = 3, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, two-tailed Student's t-tests were used to compare the two groups. The data shows mean ± SD. Please click here to view a larger version of this figure.
Supplementary Figure 1: Schematic instruction of equipment preparation. Please click here to download this File.
With the rapid development of microfluidic technology, OOC models that can replicate the biological function and structure of one or more organs in vitro have emerged in recent years for applications in biology, medicine, and pharmacology15. OOC can simulate key functions of the human physiological microenvironment, essential for exploring disease mechanisms and promoting preclinical drug translation8,16. Although OOC is still in the early stages and more input is needed to optimize the design of OOC, much progress has been made in recent years17,18. Unlike other three-dimensional cell culture models, OOC can mimic the biomechanical parameters of the human body that are essential for tissue development and maintenance of organ function. Organs and tissues of the cardiovascular system are constantly subjected to fluid flow-induced shear and cyclic tension in vivo. Therefore, in vitro cellular experiments of the cardiovascular system need to be performed in an experimental platform that provides such biomechanical parameters to obtain realistic experimental results comparable to those obtained from in vivo studies.
In this study, we established a HASMC-OOC that can precisely control the parameters of the biomechanical strain to which HASMCs are subjected. It is possible to control the fluid velocity and fluid shear to which the cells are subjected, as well as to precisely control the amplitude, frequency, and rhythm of different patterns of cyclic strain. Of the many materials that have been used in the field of OOC, PDMS is one of the most widely adopted19,20,21. PDMS is transparent, flexible, biocompatible, and breathable, and these properties contribute to its wide usage in the field of microfluidic chips19. Applied pressure changes can deform PDMS to achieve the mechanical force of contraction and relaxation, and the permeability of PDMS ensures that the cell culture environment on the chip can be consistent with the incubation conditions of 5% CO2 at 37 °C in the incubator. Therefore, PDMS not only provides controlled mechanical force to the cells but also provides a culture environment suitable for normal cell growth. Our experimental results show that the morphology and activity of cells cultured in the PDMS chip are consistent with 2D culture conditions, indicating that PDMS is biocompatible.
Under biomimetic in vivo physiological conditions, HASMCs in the aortic wall are subjected to the cyclic strain of approximately 9%22. In pathological conditions, such as hypertension, aortic dilatation, and aortic aneurysm, HASMCs are subjected to cyclic strain greater than 16%23. We characterized the tensile properties of the PDMS chip, and the experimental results show that a vacuum pressure of 10 kPa induced 7.18 ± 0.44% strain and 15 kPa induced 17.28 ± 0.91% strain. In addition, the frequency of the strain and fluid shear stress can be controlled by this PDMS chip system. Using BAV-TAAD and TAV-TAAD patient-derived primary HASMCs, the molecular mechanisms and drug screening of these aortic diseases can be conducted using this HASMC-OOC in vitro model, which can replicate the disease pathogenesis of different aortic diseases. Using this system, we have established a disease model to study the pathogenesis of BAV-TAAD and revealed that mitochondrial fusion activator could partially rescue the mitochondrial dysfunction in diseased cells24.
The HASMC-OOC was designed based on the inspiration and reference of lung-on-a-chip from Ingber's work9. Their device replicated physiological breathing movements of the alveoli in the lung, and the chip consisted of three layers: upper and lower medium channels, a porous PDMS membrane, and two side vacuum chambers. Side vacuum chambers can only be made in specialized laboratories with advanced microfluidic equipment. To enable assembly in more general laboratory conditions, we placed the vacuum chambers on the top and bottom of the chip. Both stretching approaches have similar functions, so the chip structure can be chosen according to the laboratory conditions and the research needs so thatselected. In addition, to structurally establish the tubular structure of the human aorta and obtain more biological samples from the cells for conducting further complex biological analysis and experimental assays, we established a five-layered PDMS chip with larger channels, which consisted of a middle medium channel, two PDMS membranes, and top and bottom vacuum chambers. Two PDMS membranes and the large channels of the PDMS chip increased the cell culture area up to 8.4 cm2, providing enough samples for protein analysis. The previously reported platforms mainly focused on investigating minor artery diseases, such as a cerebral artery, peripheral vessel, or other arterial diseases25,26. The aorta is a large vessel with a diameter of 25-35 mm, and the tunica media of the aorta is mainly composed of smooth muscle cells subject to high tensile strain. The pathology of aortopathy is associated with remodeling of the aortic wall, smooth muscle cell apoptosis, and phenotypic switching, all of which can be regulated by mechanical stress. Thus, patient-derived HASMCs and tensile strain are essential components for the success of HASMC-OOC. However, there are some limitations to this study. HASMC-OOC did not utilize cocultured vascular cells, including HASMCs, aortic endothelial cells, and fibroblasts, which might better simulate the in vivo microenvironment of the human aorta and can be used to study the interaction between these cells. Although HASMCs experience rhythmic strain, the cells are still cultured on a flat PDMS membrane that fails to mimic the extracellular matrix. Ia n future studies, we will overcome these limitations by combining microfluidic chips with 3D bioprinting technology.
The authors have nothing to disclose.
The authors acknowledge that this work was supported by grants from the Science and Technology Commission of Shanghai Municipality (20ZR1411700), the National Natural Science Foundation of China (81771971), and Shanghai Sailing Program (22YF1406600).
4% paraformaldehyde | Beyotime | P0099-100ml | Used for cell immobilization |
Alexa Fluor 350-labeled Goat Anti-Rabbit IgG | Beyotime | A0408 | Antibodies used for immunostaining |
Bovine serum albumin | Beyotime | ST025-20g | |
Calcium AM/PI | Invitrogen | L3224 | |
Cell culture flask | Corning | 430639 | |
CNN1 | Abcam | Ab46794 | |
Commercial flexible PDMS membrane |
Hangzhou Bald Advanced Materials | KYQ-200 | |
F-actin | Invitrogen | R415 | |
FBS | Sigma | M8318 | |
Hoses | Runze Fluid | 96410 | 1 mm inner diameter; 3 mm outer diameter; 1 mm wall thickness; Official website address: https://www.runzefluidsystem.com |
Human aortic smooth muscle cell line CRL1999 |
ATCC | Lot Number:70019189 | |
Image J | Imagej.net/fiji/downloads | Free Download: https://fiji.sc | Imaging platform that is used to identify fluorescence intensity |
Incubator | Thermo Fisher Scientific | Ensures that the temperature, humidity, and light exposure is exactly the same throughout experiment. |
|
Luer | Runze Fluid | RH-M016 | Official website address: https://www.runzefluidsystem.com. |
Microscope | Olympus | ||
mouse collagen | Sigma | C7661 | |
Oxygen plasma | Changzhou Hongming Instrument | HM-Plasma5L | |
Pasteur pipette | Biologix | 30-0138A1 | |
PBS | Beyotime | C0221A | |
Pen-Strep | Sigma | P4458-100ml | Antibiodics used to prevent bacterial contamination of cells during culture. |
peristaltic pump | Kamoer | F01A-STP-B046 | |
Petri dish | Corning | 430167 | |
PLC controller | Zhejiang Jun Teng (BenT) CNC factory | BR010-11T8X2M | The detailed program setting can be found in supplementary. Official website address: files.http://www.btcnc.net |
polydimethylsiloxane (PDMS) | Dow Corning | Sylgard 184 | |
SM22 | Abcam | ab14106 | |
SMCM | ScienCell | Cat 1101 | |
solenoid valve | SMC (China) | VQZ300 | |
Syringe | Becton,Dickinson and Company | 300841 | |
Triton-X 100 | Beyotime | ST795 | To penetrate cell membranes |
Trizol | Invitrogen | 10296010 | Used for RNA extraction |
trypsin | Sigma | 15400054 | |
vacuum filter | SMC (China) | ZFC5-6 | Official website address: https://www.smc.com.cn |
vacuum pump | Kamoer | KVP15-KL-S | |
vacuum regulator | AirTAC | GVR-200-06 | |
Primers | |||
Primer Name | Forward (5’ to 3’) | Reverse (5’ to 3’) | |
SM22 | CCGTGGAGATCCCAACTGG | CCATCTGAAGGCCAATGACAT | |
CNN1 | CTCCATTGACTCGAACGACTC | CAGGTCTGCGAAACTTCTTAGA |