Nanomechanical Induction of Autophagy‐Related Fluorescence in Single Cells with Atomic Force Microscopy

Abstract Mechanistic understanding of how living systems sense, transduce, and respond to mechanical cues has important implications in development, physiology, and therapy. Here, the authors use an integrated atomic force microscope (AFM) and brightfield/epifluorescent microscope platform to precisely simulate living single cells or groups of cells under physiological conditions, in real time, concomitantly measuring the single‐cell autophagic response and its transmission to neighboring cells. Dual‐color fluorescence monitoring of the cellular autophagic response reveals the dynamics of autophagosome formation, degradation, and induction in neighboring contacting and noncontacting cells. Autophagosome formation is dependent on both the applied force and contact area of the AFM tip. More importantly, the enhancement of the autophagic responses in neighboring cells via a gap junction‐dependent mechanism is observed. This AFM‐based nanoacupuncture platform can serve as a tool for elucidating the primary mechanism underlying mechanical stimulation of living systems and other biomechanical therapeutics.


mCherry-GFP-LC3 transfection
2 × 10 4 HeLa cells were seeded in a standard glass-bottomed well plates and cultured in MEM medium for 24 h before transfection. Growth media were replaced before transfection with MEM medium containing G418 at a final concentration of 1 mg/mL. To each dish, 0.8 μL of Lipofectamine 3000 was added as described in the manufacturer's instructions using Opti-MEM, mixed with 300 ng mCherry-GFP-LC3 plasmids (Anne Brunet, Addgene plasmid #110060) and 25 μL incubation reagent. Transfected cells were incubated as described above and then cultured in MEM supplemented with 500 μg/mL G418.

Cx43-mCherry transfection
In order to study the role of gap junction communication for the transfer of an autophagic signal, we transfected endogenously low connexin expressing N2A cells with Cx43-mCherry.
Cx43-mCherry plasmid was constructed from pLPCX-Cx43-IRES-GFP from TrondAasen (Addgene plasmid #65433). 3 × 10 4 N2A-GFP-LC3 cells were seeded in a standard glassbottom well plate and cultured in MEM medium for 24 h. Growth media were replaced before transfection with MEM medium containing G418 at a final concentration of 1 mg/mL. To each dish, 500 ng Cx43-mCherry plasmid was added with 0.8 µl Lipofectamine 3000 and 25 μL incubation reagent according to the manufacturer's instructions. Transduced cells were grown in the medium, as described above. Adjacent contacting cells expressing Cx43-mCherry (with the formation of gap junction) were selected for nanoacupuncture experiments.

Cell staining
To study autophagy between N2A-GFP-LC3 and HeLa-GFP-LC3 cells, the nucleus in N2A cells were stained with Hoechst 33258 (Sigma-Aldrich, USA). 2 × 10 4 N2A-GFP-LC3 cells were seeded in a standard glass-bottomed well plate and cultured in MEM medium for 24 h, then incubated in MEM medium containing 5 μg/mL Hoechst 33258 for 10 min and washed for 3 times with PBS. Then to the confocal plates with Hoechst 33258 stained N2A-GFP-LC3 cells, 2 × 10 4 HeLa-GFP-LC3 cells were seeded, and the cell mixture was incubated in MEM medium for 12 h. The adjacent contacting cell pairs consisting of N2A-GFP-LC3 (nucleus stained with Hoechst) and HeLa-GFP-LC3 cells were selected for nanoacupuncture experiments.

Atomic force microscopy and fluorescence microscopy
A Bioscope Resolve AFM (Bruker, USA) and fluorescence microscope (Leica DMI 3000 B, Germany) were used in the present study. Fluorescence images were acquired by oilimmersion 63× (NA1.4, Leica) with 380 nm, 480 nm, and 520 nm laser lines. All living cell AFM experiments were performed in Leibovitz's L-15 cell culture medium at 37 °C and 40-50% environmental humidity.
The tip radius for the DNP-S10 cantilever was calculated by applied deconvolution on image results for the RS-12M tip-checker sample (Bruker). Other cantilevers' tip radii were calibrated by their manufacturers. Spring constants for DNP-S10 and silicon nitride probes were calibrated by the thermal noise method in Bruker Nanoscope 9.3 software.
The force on the cells was loaded in the PFQNM ramp mode. The relaxation and forcedistance (F-D) curve measurements were obtained with a 4.0 μm/s ramp rate, 2.0 μm ramp size, and holding time of 2700 s or 4500 s (45 min or 75 min). The loading force applied to the samples varied between 25 pN to 5 nN, according to the different experimental protocols.
For the nano-dissection, approximately 7-10 nN force was applied in PeakForce tapping mode.
Indentation depths of cells were determined from F-D curves and Young's modulus of cells was calculated using the JKR model. According to the JKR model, the applied force F and the indentation depth h are related by: Where F is the force applied on the cell by the spherical AFM tip, Δγ is the work of adhesion, E is the elastic modulus of the cell, ν (with ν = 0.5) is the Poisson ratio of the cell, R is the tip radius, a is the contact radius, and h is the indentation depth.
For the interacting area between the cell membrane and a spherical tip, different equations were used to calculate due to different indentation depth. When indentation depth (h) is larger than the radius of the tip (R) (a), the interacting area (IA) between the cell membrane and a spherical tip is calculated following the equation IA=2πR 2 . When indentation depth (h) is smaller than the radius of the tip (R) (b), the interacting area (IA) between the cell membrane and a spherical tip is calculated following the equation IA=2πRh. Schematic illustration of the correlation between the contact area, and indentation on a cell membrane using a diameter spherical-shaped AFM tip was shown as follows.
The morphological images of cells were acquired with a tapping frequency of 0.5-1.0 kHz using DNP-S10 tip (diameter of 20 nm). The PeakForce tapping amplitude was 300 nm with the setpoint 1.0 nN. Measured heights of cells before nanoacupuncture were determined from sectional profiles of three types of Cells. All AFM images and F-D curves were analyzed by NanoScope Analysis 2.0 (Bruker).
Change of occupied area of cells on Petri dishes before and after nanoacupuncture were captured using fluorescence imaging and then calculated using Image J software.

Nano-dissection of a connected cell pair
Two adjacent cells was selected with the optical microscope, and the connected area between them was scanned with AFM imaging (PeakForce tapping mode) to confirm if there was a direct physical connection between these two selected cells. To cut the connection, the coneshaped AFM tip (20 nm diameter) was placed perpendicularly with a loading force >8 nN and a scan aspect ratio of 1/64 frame. Typically, the nano-dissection was conducted on a small scan-area (1000 nm x 15 nm) with a scanning speed of ~1 Hz, and the whole dissection process was generally longer than 5 min. After the dissection, the connected area between the two cells was scanned with AFM imaging (PeakForce tapping mode) again to confirm the success.          (middle) and N2A cell (right) and corresponding section profiles of the three cells. b) Based on AFM sectional images, the thickness of untreated cells is determined to be 5.4 ± 0.3 μm, 4.5 ± 0.2 μm, and 7.8 ± 0.5 μm, for HeLa, PC12 and N2A cells, respectively. n = 10 for HeLa and N2A cells, n = 8 for PC12 cells. Data were presented as mean ±SD. c) Based on fluorescence images, occupied area of untreated cells on Petri dishes is determined to be 593.9 ± 57.3, 479.7 ± 46.6, and 336.8 ± 23.1 μm 2 , for HeLa, PC12 and N2A cells, respectively. n = 10 cells for each group. Data were presented as mean ±SD. ns = no significant difference, *p<0.05, ***p<0.001, Student's t-test. The occupied area of cells on Petri dishes before and after nanoacupuncture revealed by fluorescence imaging results. n = 5 cells for each group. Data were presented as mean ±SD. Nanoacupuncture time, 45 min; loading force, 5 nN; sphere-shaped AFM tip diameter, 2.5 µm. ns = no significant difference; *p<0.05, ***p<0.001; Student's t-test.   HeLa-HeLa, N2A-HeLa, N2A-N2A, and HeLa-N2A, respectively). Data were presented as mean ±SD. Autophagy response could be observed in unstimulated cells connected to cells directly stimulated via nanoacupuncture. g) The average number of new puncta in each responsive cell in pairings of cells. The stimulated (press) group is in red and the connection group is in blue (n = 6, 4, 5 and 4, for HeLa-HeLa, N2A-HeLa, N2A-N2A, and HeLa-N2A, respectively). Data were presented as mean ±SD. Cone-shaped AFM tip diameter: 20 nm; nanoacupuncture time: 45 min; loading force: 5 nN. *p<0.05, ns: not significant, Student's ttest. Cell pairs without connections (-, n = 6; data were presented as mean ± SD); cell pairs with connections (+, n = 6; data were presented as mean ± SD); cell pairs with connections resected ((-), n = 3; data were presented as mean ± SD). Loading force: 5 nN; cone-shaped AFM tip diameter: 20 nm; nanoacupuncture time: 45 min. **p<0.01, Student's t-test.  Note：1) for the SEM images, Scale bar:1 μm. 2) we define the calculation formula under two different conditions. For the case of indentation depth ≤ probe diameter, contact area = 2πRh; for the case of indentation depth > probe diameter, contact area = 2πR 2 .