Introduction

Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder and the most common neurological disorder (~75% cases)1. In 2021, Word Health Organization (WHO) reported that more than 40 million people worldwide were affected by AD2. Two clinical biomarkers have been used to diagnose AD: Amyloid-β (Aβ) protein and Tau protein3. Aβ is the primary biomarker in AD4. Aβ monomers aggregate and form Aβ oligomers (AβO) and then protofibrils and fibrils. The further assemblies of amyloid fibrils form amyloid plaques5. These toxic aggregations disrupt communication between nerve cells and cause neuron cell death3. Aβ plaques were initially considered as a central factor in the brains of AD patients6. However, studies in AD animal models and humans have shown that AβO are most potent for AD pathogenesis, and the variation of AβO levels is one of the major biomarkers in AD progression7.

Microglia cells, as immune cells in the central nervous system (CNS), play a critical role in the clearance of AβO when they migrate to, adhere to, and phagocyte AβO8,9. However, in advanced AD, when microglia are stimulated by AβO, they release excessively inflammatory factors and cytotoxins, leading to neuronal injuries and death. Also, the interaction of Aβ aggregations to microglia receptors causes dysfunction and inflammation in the brain10. To detect and progress AβO in AD, immunosorbent assay (ELISA), western blotting, and capillary electrophoresis (CE) are the conventional techniques used11,12,13,14. However, the procedure of these methods has several steps and requires the preparation of reagents (antibodies, antigens, and gel) and labeling, making them laborious, time-consuming, and costly15,16.

Changing AβO levels can directly affect cellular behaviors in the brain17, and cell adhesion to the substrate as a mechanobiology property varies during pathological conditions18,19. We hypothesized that microglial activities in exposure to AβO, including migration, phagocytosis9, apoptosis20, and activating microglial receptors through Aβ21,22, could affect mechanobiology microglia adhesion to the substrate. Therefore, the quantification of microglia-substrate adhesion subjected to AβO can be defined as a physical or mechanobiological marker to show the progression of AD. This quantitative information can provide a better understanding of AD detection and progression by microglia cells-substrate adhesion, which can help develop immune cell-based AD prognostic strategies.

Microfluidic devices have been increasingly used to characterize cell adhesion due to their unique advantages, such as high controllability for fluid flow, portability, minimal reagent volume, ability to integrate a few components into a device, and reproducibility23,24,25,26,27. In order to study the cell-substrate adhesion for the cell population quantitatively using a microfluidic device, shear stress through fluid flow is applied over the attached cells on the substrate of the device. Then, the detached cells were quantified by real-time microscopy during the experiment. As a result, the microfluidic adhesion assay does not require labeling (label-free) and complex equipment and procedures to assess the cell-substrate adhesion. This approach has been expanded to study the cell-substrate adhesion by changing substrate topologies and coatings on various cell types, including fibroblast cells28, breast tumor cells, human lymphoma cells, keratinocytes cells29, fibroblast and bovine aortic endothelial cells30, vascular and valvular endothelial cells31, and cancer breast epithelial cells32. However, there is no report to study cell-substrate adhesion as a physical marker on-chip to detect a disease progression.

In this study, we presented a simple microfluidic device developed as an AD model to assess microglia BV2 cells-substrate adhesion without using labeling when the cells were exposed to AβO. The cells were subjected to fluid shear stresses to analyze microglia-substrate adhesion using microfluidic AD models to characterize the cell removal using time-lapse microscopy. Shear stresses of 3 Pa and 7.5 Pa were applied to the cells through the inlet flow at Reynolds number (Re) of 10 and 25, respectively when the cells were cultured in exposure to various AβO concentrations (1 µM, 2.5 µM, and 5 µM) for 1 h, 6 h, 12 h, and 24 h. It was shown that AβO in the microfluidic device were phagocyted by microglia BV2 cells using pHrodo dye assay. Statistical comparison between the studied cases was carried out using a fraction of the attached cells to the substrate under applied shear stress to the cells. The current investigation helps to better understand the strength of microglia-substrate interaction in an in vitro microenvironment of AD condition.

Results

Mechanism of Microfluidic Adhesion Assay to Study of Microglia BV2 cells-Substrate Adhesion Exposed to AβO

In this study, a microfluidic device as an AD model was developed to measure microglia BV2 cells-substrate adhesion quantitively by exposing the cells to AβO. The microfluidic assay has a single channel with inlet and outlet reservoirs. (Fig. 1a, b). The PBS flow was pumped within the device using the syringe pump to exert the fluid shear stress on the cells (Fig. 1c–e). The cells under the shear stress start to detach from the substrate. The cell removal was monitored under the microscope to count the attached cells at each time frame until all cells in the region with a uniform shear stress of the device were washed. The region that the cells subjected to constant shear stress was determined when the shear stress distribution was measured in the cross-section of the channel middle at a height of 3 µm (Fig. 2a–c). Figure 2b shows that 75% of the channel has uniform shear stress of 3 Pa and 7.5 Pa at the Reynolds Number of Re=10 and 25, respectively, while the shear stress in 25% of the channel (side of the wall of the channel) varied, which was disregarded in the quantification of cell removal (Fig. 2c).

Fig. 1
figure 1

Schematic of microfluidic cell adhesion assay. a Photo of the microfluidic device; the microchannel and reservoirs were filled with red dye (scale bar= 25 mm). b dimensions of the microchannel and reservoirs (top view). c 3D Experimental setup and d side view of device when the PBS flow was pumped into the device to exert the flow shear stress on the cells. e Magnified shear stress on the cells, velocity streamlines, and velocity profile in the microchannel

Full size image
Fig. 2
figure 2

Shear stress distribution a of the cross-section at the middle of the channel and b at the height of 3 µm along width of the channel at inlet Re = 10 and Re = 25. c The area that the cells subjected to the uniform shear stress (central area between two dash lines) was shown. Scale bar = 450 µm

Full size image

Evaluation of Microglial BV2 Phagocytosis to AβO

In AD, microglial cells as immune cells are responsible for phagocytosis of AβO33. To evaluate phagocytosis of microglial BV2 cells to AβO in the microfluidic device, AβO were labeled with pHrodo Red dye (Invitrogen). pHrodo is a fluorogenic dye with no fluorescence at neutral pH; however, the fluorescence intensity of pHrodo Red dye increases significantly in acidic environments within the microglial cells, compared to the more neutral extracellular environment. As a result, when AβO are internalized through phagocytosis, the dye becomes highly fluorescent due to the acidic conditions inside the cells. Supplemantary Fig. 1a–d shows phagocyted AβO by microglia BV2 cells after 1 h, 6 h, 12 h, and 24 h of loading AβO 2.5 µM into the device. In Supplemantary Fig. 1a–d, AβO phagocyted into the cells were observed with red fluorescence, while AβO out of the cells did not represent red fluorescence. Therefore, the bright and fluorescent images showed that microglia BV2 cells in the microenvironment of the device involved to phagocyte AβO during 24 h.

Quantification of Microglial BV2 Adhesion in Exposure to AβO

In order to characterize the adhesion strength of microglial BV2 cells exposed to AβO (concentrations of 1 µM, 2.5 µM, and 5 µM), the fraction of the cell attachment (F) was quantified under PBS flow of inlet Re=10 and 25, then compared their results with the control conditions (Figs. 3, 5a–h). F was calculated for 1 h, 6 h, 12 h, and 24 h exposure of the cells to AβO, which their control conditions were equal to 5 h, 10 h, 16 h, and 28 h of cell culture into the device, respectively, because, after 4 h of cell loading to the device, AβO were added to the device. F in each condition was measured from Supplementary Movies 1-30, which shows cell removal as the cells were subjected to PBS flow.

Fig. 3
figure 3

Fraction of the cell attachments under inlet Re=10 (3 Pa) by varying cell culture time and AβO concentrations. The fraction of the cell attachments for a control conditions after 5 h, 10 h, 16 h, and 28 h of cell culture and for various AβO concentrations including, b 1 µM, c 2.5 µM, d 5 µM after 1 h, 6 h, 12 h, and 24 h. The fraction of the cell attachments after e 1 h, f 6 h, g 12 h, and h 24 h exposure of the cells to AβO with concentrations of 1 µM, 2.5 µM, and 5 µM compared to the control conditions

Full size image

Figure 3a–d and Fig. 5a–d indicate the effect of culture time on the adhesion strength of microglial BV2 cells exposed to AβO and their control conditions by applying shear stress of 3 Pa (inlet Re=10) and 7.5 Pa (inlet Re=25) to the cells. In the control conditions under a shear stress of 3 Pa (Re=10, Fig. 3a) and 7.5 Pa (Re=25, Fig. 5a), the results showed that by increasing the time of cell culture in the device from 5 h to 28 h, there is a reduction trend in F, although the differences between the cases are minor, in which few time steps, their results overlapped. Similar to the control conditions, as the incubation time of the cells subjected to each AβO concentration increased from 1 h to 24 h, F decreased (Fig. 3b–d and Fig. 5b–d), which means the cell adhesion reduced. Despite the control conditions, by increasing the time of the cells exposed to AβO from 1 h to 24 h, there is a clear separation between F of the cases (no overlapped data).

In Fig. 3e–h, the adhesion of microglia BV2 cells was compared when the cells were under shear stresses of 3 Pa (inlet Re=10) at each time of 1 h, 6 h, 12 h, and 24 h of AβO loading into the device with concentrations of 1 µM, 2.5 µM, and 5 µM. Figure 3e indicated that as the cells were exposed to AβO for 1 h, 3 Pa shear stress (Re=10) was applied to the cells. It was found that the cells attached in the first 210 s were 65%, 40%, 27%, and 15% of the initial number of cells for control, 1 µM, 2.5 µM, and 5 µM, respectively. It shows that an increased concentration of AβO leads to losing microglial adhesion. Similarly, there is a reduction in microglial adhesion by exposing the cells to AβO for 6 h; when the cells were subjected to 3 Pa for 180 s, 60%, 34%, 11%, and 2.8% of cells remained for control, 1 µM, 2.5 µM, and 5 µM, respectively (Fig. 3f).

Increasing the incubation time of microglia exposed to AβO increases the differences between various AβO concentrations in F. For example, for microglia exposed to AβO for 12 h under 3 Pa, all the cells in AβO concentration of 5 µM were detached after 150 s under 3 Pa, while 85%, 49%, and 9% of the cells were attached in control and AβO concentrations of 1 µM and 2.5 µM, respectively (Fig. 3g). When the time of microglia exposed to AβO reached 24 h, the cell adhesion became zero, making them unlikely viable. (Fig. 4a, b), and for control, 1 µM, and 2.5 µM of AβO, F was 83%, 48%, and 20%, respectively after 120 s under 3 Pa (Fig. 3h).

Fig. 4
figure 4

Condition of microglia BV2 cells after 24 hours of exposure to 5 µM AβO in the microchannel: a 4X magnification and b 10X magnification. Microglial cell adhesion was reduced to zero, indicating likely loss of viability (cells appear as dots in the top view), as all cells were immediately washed away under PBS flow. Scale bar = 450 µm

Full size image

Figure 5e–h show the variation of F at each time of 1 h, 6 h, 12 h, and 24 h exposure of the cells to AβO concentrations of 1 µM, 2.5 µM, and 5 µM as the shear stress applied to the cells was increased from 3 Pa to 7.5 Pa. Our findings indicated a reduction trend in the strength of cell-substrate adhesion by increasing the time of microglia in exposure to AβO and AβO concentrations, similar to the results shown in the cells subjected to 3 Pa. However, as expected, the increase of the shear stress at 7.5 Pa leads to the cells being detached faster than 3 Pa. For example, after 1 h exposure of the cells to AβO, all cells under 7.5 Pa were removed in 220–240 s at AβO concentrations of 1 µM, 2.5 µM, and 5 µM, compared to those of 270-330 s under 3 Pa. It was found by increasing the time of the cells exposed to various AβO concentrations the difference of F between the cases increased (Fig. 5e-h).

Fig. 5
figure 5

Fraction of the cell attachments under inlet Re=25 (7.5 Pa) by varying cell culture time and AβO concentrations. The fraction of the cell attachments for a control conditions after 5 h, 10 h, 16 h, and 28 h of cell culture and for various AβO concentrations including, b 1 µM, c 2.5 µM, (d 5 µM after 1 h, 6 h, 12 h, and 24 h. The fraction of the cell attachments after e 1 h, f 6 h, g 12 h, and h 24 h exposure of the cells to AβO with concentrations of 1 µM, 2.5 µM, and 5 µM compared to the control conditions

Full size image

In order to characterize the strength of cell adhesions under various AD conditions, a nondimensional mechanobiological parameter of adhesive strength, Ø, was defined as

$$\O =\int \Delta {\rm{Fdt}}/{{\rm{T}}}_{\max }$$
(1)

Where ∫∆Fdt is the integration from the plot of F in respect to cell removal duration (t) for each condition (Figs. 3, 5), and Tmax is the maximum time for removing all cells in the area with uniform shear stress at each Re. Figure 6a, b shows Ø when 3 Pa and 7.5 Pa were applied to the cells after 1 h, 6 h, 12 h, and 24 h of cells exposed to AβO. For the cells subjected to 3 Pa, there is a significant difference (p < 0.05) of Ø between all cases at times of 12 h and 24 h of the cells exposed to AβO (Fig. 6a). However, by applying 7.5 Pa to the cells, a significant difference (p < 0.05) of Ø between all the cases was found only for 24 h exposure of the cells to AβO (Fig. 6b).

Fig. 6
figure 6

Ø (nondimensional parameter of cellular adhesive strength) after 1 h, 6 h, 12 h, and 24 h exposure to microglia BV2 cells to AβO 1 µM, 2.5 µM, and 5 µM under a Re=10 and b Re=25. Data represent the mean ± SD of three independent experiments (n = 3). ANOVA was used for statistical analysis, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Full size image

Discussion

The adhesion strength of microglia BV2 cells to the underlying substrate can be regulated in microenvironment stimuli of AD when the cells are exposed to AβO as a major pathological factor in the brains of AD patients7. Our report utilized a microfluidic assay to detect advancing AβO in exposure to microglia BV2 cells by quantitative comparison of the cell-substrate adhesion between various AβO concentrations. The microfluidic device was used because the procedure of characterization of the cell-substrate adhesion is label-free (without antibodies and antigens) when the cell removal from the substrate under the flow shear stress was quantified by real-time microscopy. Also, using the microfluidic device to measure cell-substrate adhesion requires only the tubing, syringe pump, microscope, and minimal cell media, making the experiment inexpensive and controllable.

In addition to these advantages, it is important to compare our label-free microfluidic assay with traditional methods that have been widely used in cellular analysis, such as immunosorbent assays (e.g., ELISA), western blotting, and capillary electrophoresis11,12,13,14. Traditional techniques typically rely on labeled antibodies or other markers, which can add significant cost and complexity to the procedure15. Moreover, these methods often provide only static, end-point measurements, limiting their ability to capture the dynamic nature of cell-substrate interactions, particularly in response to changing conditions16 such as AβO exposure. In contrast, our microfluidic platform offers continuous, real-time analysis of cell behavior without the need for labeling, reducing the time and cost associated with reagent preparation and handling. This dynamic monitoring enables a more comprehensive understanding of the temporal changes in cell-substrate adhesion and provides detailed insights into the effects of varying AβO concentrations. Furthermore, the simplicity and low cost of the microfluidic assay make it an ideal tool for high-throughput screening, which is difficult to achieve with more traditional, labor-intensive methods23,24,25,26,27.

To our knowledge, it is the first study to detect the progress of AβO by quantification of microglia cells-substrate adhesion as a mechanobiological marker. To show the strength of the cell adhesion, the cell attachment fraction (F) was measured during the applied shear stresses of 3 Pa (inlet Re=10) and 7.5 Pa (inlet Re=25) to the cells. The increased incubation time of the cells in vitro affects the reduction of cell viability and also cell-substrate adhesion34. The results in the control conditions indicated that after 5 h, 10 h, 16 h, and 28 h of the cell culture under 3 Pa and 7.5 Pa, there were overlapped data in F (Figs. 4, 5a), which means that over time of 24 h, no significant variations in the cell adhesion was observed. It shows that during 24 h, the cell viability rate in the control conditions was low. As a result, our microfluidic assay is a good alternative to culture the cells during the experiment. Our data presented that the cell adhesion was reduced by increasing two factors: the time of microglia BV2 exposure to AβO and AβO concentration. Because microglia cells in AD migrate to phagocyte AβO, in their migration, the cells release adhesions at the rear and form new adhesions at the front35, which means the cells in migration was attached to the substrate partially. Our numerical results showed that as a single cell attached to the substrate partially was subjected to the flow shear stress, the cells received higher mechanical force36, resulting in the detachment of the cells faster than the physiological condition. Also, the microglia body in phagocytosis of AβO (activated state) has a lower substrate area (lower adherent junctions) than the control condition37, which reduces adherent junctions of the cells to the substrate and, following that, reduction of cells-substrate adhesive strength. Furthermore, the increased time of microglia cells exposed to AβO can drop the microglial affinity to the substrate when AβO can induce microglial death20,38,39.

It was reported that the increase of AβO concentrations drops the strength of microglial attachment to the substrate. Because it is expected that microglial reactions, such as migration, phagocytosis, and apoptosis, were intensified in higher AβO concentrations, leading to losing microglial adhesion to the substrate. Also, the engulfing of AβO by the cells can largely change the organization of actin filamens that can affect the reduction of cell-substrate adhesion40. Regarding microglia BV2 cells exposed to AβO 5 µM, it was found that after 24 hours of exposure to AβO, microglial cell adhesion dropped to zero, indicating that the cells were likely no longer viable. (Fig. 4a, b). Since two causes of the reduction of cell-substrate adhesion, including the time of cells exposed to AβO and AβO concentration, reached the highest level compared to the studied cases.

Two shear stresses of 3 Pa and 7.5 Pa corresponding to the device inlet of Re at 10 and 25, respectively, were applied to the cells to quantify the cell adhesion. Our findings showed that the results of mechanobiological parameter of cellular adhesive strength (Ø) under 3 Pa have significant differences (p < 0.05) between all AβO concentrations for 12 h and 24 h exposure of microglia BV2 cells to AβO, while at Re=25, these significant differences were observed only for 24 h exposure of the cells to AβO. Under low Re (Re=10), the fluid loading on the cells is low, leading to the gradual weakening of cell adhesion to the substrate compared to higher Re (Re=25) condition. Hence, the quantification of cell-substrate adhesion using 3 Pa compared to 7.5 Pa had a clear effect showing advancing AβO in AD.

Interestingly, as shown in Fig. 6b (Re=25), under higher shear stress (7.5 Pa), the parameter Ø converges closer to the control after 6 hours of treatment with 1 μM AβO but diverges significantly at 12 and 24 hours. This transient behavior may result from the initial stabilizing effect of higher mechanical forces, temporarily counteracting the impact of AβO. However, as the exposure time increases, the toxic effects of AβO accumulate, leading to a reduction in adhesion strength. In contrast, Fig. 6a (Re=10, 3 Pa) shows a more uniform trend, with Ø diverging steadily over time, reflecting a gradual weakening of adhesion due to AβO. These observations suggest that Re=10 provides better conditions for capturing the effects of AβO on cell adhesion dynamics, enabling more consistent and reliable differentiation between AβO concentrations.

Conclusion

In the current study, a microfluidic device was developed as an AD model on-chip to study AβO progression in AD by quantifying the strength of microglia BV2 cells-substrate adhesion as a mechanobiological marker. The fraction of cells attached to the substrate was counted when the cells were subjected to the flow shear stress at different time steps during 24 h to characterize cell-substrate adhesion exposed to AβO. Our data showed the quantification of the cell-substrate adhesion could successfully represent various conditions created by increasing AβO concentrations exposed to the cells. It was revealed that by increasing the duration of the cells exposed to AβO, the binding strength of cells to the substrate was reduced. Our present study demonstrated that advancing AβO concentrations caused losing the strength of cell-substrate interaction. The result showed after 24 h incubation of microglia BV2 cells exposed to AβO 5 µM; the cell adhesion became zero, making them unlikely viable. The shear stress of 3 Pa (device inlet Re=10) compared to 7.5 Pa (device inlet Re=25) was useful to study conditions created in advancing AD with AβO concentrations. This quantitative information can provide a better understanding of AD detection and progression by microglia cells-substrate adhesion, which can help develop immune cell-based AD prediagnostic and diagnostic strategies.

Materials and Methods

Device Design and Fabrication

The microfluidic device consists of a single channel (length=8 mm, Width=900 µm, and Height=100 µm), and there are the inlet and outlet reservoirs (diameter=6 mm and Height=7 mm) at the terminals of the channel (Fig. 1a, b). The devices were fabricated in polydimethylsiloxane (PDMS) using the standard soft lithography technique. The master mold on the silicon wafer was created using photolithography of the negative photoresist (SU8- 2075, MicroChem, USA). The PDMS, a mixture of the elastomer base (Sylgard 184, Dow Corning, USA) and the elastomer curing agent (Sylgard 184, Dow Corning, USA) at a ratio of 10:1, was poured over the master. After incubating at 70 °C for 2 h, the PDMS was peeled off, and then, two terminals of the channel were punched with a biopsy punch (diameter of 6 mm) to create the reservoirs. The PDMS was bonded irreversibly to the glass slide by treating oxygen plasma for 1 min. Next, the device was incubated at 50 °C for 15 min to make a strong bonding between the PDMS and glass.

Experiment Setup to Quantification of the Cell Removal into the Device

The cell adhesion was quantified when the cell detachment under fluid shear stress was monitored using time-lapse microscopy. The experiment steps were carried out to subject the cells to the shear stress as follows: After leaving the device under the microscope, the syringe (1 ml, Becton, Dickinson and Company, USA) with PBS 0.01X was mounted onto a syringe pump (Legato 111, Kd Scientific, USA). The syringe was connected to the inlet reservoir with microfluidic tubing (ID: 2 mm, OD: 5 mm, Halmar, UK). Using the syringe pump, PBS 0.01X flow was pumped through the microchannel at different velocities corresponding to inlet Re of 10 and 25 (Fig. 1c–e).

Prediction of Flow Shear Stress over the Cells

The shear stress in the channel was estimated through finite element analysis of fluid flow to obtain the region with uniform shear stress over the cells (Fig. 1). 3D flow in the microfluidic channel was predicted using 3D module (Laminar Flow) of COMSOL Multiphysics 5.5. The steady-state incompressible Naiver-Stokes equations were used to describe the flow of fluid. The flow was assumed steady, laminar, Newtonian, and viscous. In the simulation, the fluid was considered deionized water (DIW) with dynamic viscosity and density of 0.00845 Pa.s and 996 kg/m3, respectively. The boundary conditions of outlet and channel walls were defined as no-slip and atmospheric pressure with normal flow, respectively. The inlet boundary condition was kept at inflow velocities for Re=10 and 25.

The shear stress (τ) in the device was calculated as follows36,41,

$${\rm{\tau }}=\frac{6{\rm{\mu U}}{\rm{ave}}}{h}(1-\frac{2y}{h})$$
(2)

where Uave was the average velocity, µ was the fluidic viscosity, h was the height, and y was the distance from the bottom plate. To predict the shear stress acting on the cells attached to the substrate, leading to cell removal, the shear stress at the height of 3 µm was studied, which is assumed to be the height of microglial cells42.

Cell Culture

The murine microglial BV2 cells were gifted by Dr. Carol Colton from Duke University. The murine microglial BV2 cells (~5×104 cells/ml concentration) were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Wisent Technologies, Canada) supplemented with 10% fetal bovine serum (FBS, Invitrogen), 1% Penicillin-streptomycin (Wisent Technologies, Canada). The cells (passages 10-12) were incubated at 37°C in 5% CO2 for two days, and each 24 h, the cell media was replaced with fresh media until the cells reached 70% confluency.

Preparation of Amyloid Beta Oligomers

Aβ1-42 peptides were dissolved in hexafluoro isopropanol (HFIP; Sigma-Aldrich). The HFIP was evaporated with a vacuum using a SpeedVac Concentrator (Thermo Fisher Scientific) and then dried in an airing cupboard. Then, Aβ monomers were dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) at the concentration of 2.2 mM. To prepare Aβ oligomers, Aβ was dissolved in DMSO and then incubated overnight at room temperature43,44.

Labeled Amyloid Beta Oligomers (AβO)

AβO solution was centrifuged at 16000 × g for 1 min to discard the supernatant, and then 1 ml of Hank’s Balanced Salt Solution (HBSS) was added and pipetted to rinse the AβO solution. To collect AβO, the AβO solution was centrifuged at 16000 × g for 1 min, and the supernatant was aspirated. 200 µl of sodium bicarbonate (0.1 M) was added to resuspended AβO by pipetting several times. 1 mg of pHrodo Red dye was diluted with 100 µl of DMSO, and then 36 µl of 10 mg/ml pHrodo Red dye was added to AβO suspension to make an initial labeling reaction. For 1 h at room temperature, the reaction tube was incubated in the dark. Next, the reaction tube was centrifuged at 16,000× g for 1 min to discard the supernatant. 1 ml of methanol was added to the reaction tube to remove excess dye. The reaction tube was centrifuged at 16,000× g for 1 min, and the supernatant was discarded. The following process was repeated four times: 1 ml of HBSS was added to the reaction tube, and then for 10 s, the tube was vortexed. The reaction tube was centrifuged at 16,000× g for 1 min, and the supernatant was discarded. After the final wash with HBSS, the supernatant was discarded, and 200 µl of HBSS was added. To improve the consistency of the Aβ suspension, the tube was left in a sonicator bath at room temperature for 10 min. The pHrodo Red-labeled Aβ was stored at -80 °C.

Cells and AβO Loading into the Device

The cells were harvested by treating with 0.05% trypsin (Wisent Technologies, Canada). After centrifuging and removing the supernatant, the cells were resuspended with a fresh media. The cells (~ 4×105 cells/ml) were added at the same time to the inlet reservoir (105 µl) and outlet reservoir (95 µl). The device was left in the incubator (37°C and 5% CO2). After 4 h of cell loading into the device, the cells became stable and attached to the device’s substrate, and then AβO stock was diluted with the cell media to obtain our desired concentration, 1 µM, 2.5 µM, and 5 µM. The media in the reservoirs was removed, and then AβO was loaded into the device with the same method used for cell loading (at the same time, 105 µl and 95 µl of AβO solution were added to the inlet and outlet reservoirs, respectively).

Data and Statistical Analysis

The images were taken when the cells were subjected to flow shear stress after 1 h, 6 h, 12 h, and 24 h of loading AβO into the device at concentrations of 1 µM, 2.5 µM, and 5 µM using an inverted fluorescence microscope (4X magnification objective, AMG EVOS FL). During cell removal, the microscope was set to take images from the center of the channel (equal distance from the inlet and outlet) every 10 s until all cells in the region with uniform shear stress were washed. The number of cells in the images was counted using ImageJ software.

During cell removal, the fraction of the cell attachment (F) at each time of image frame (tF) was measured to assess the strength of cell-substrate adhesion quantitively. F at each tF was obtained by dividing the number of the cell-attached at tF by the number of initial cells at t = 029,30.

Three experiments (three devices, n = 3) were performed for each condition. The quantified data were expressed as means ± standard deviation (SD). The statistical analysis, ANOVA, was performed using GraphPad PRISM software. P < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001) was considered statistically significant.