|Home | About | Journals | Submit | Contact Us | Français|
Programmed subcellular release is an in vitro technique for the quantitative study of cell detachment. The dynamics of cell contraction are measured by releasing cells from surfaces to which they are attached with spatial and temporal control. Release of subcellular regions of cells is achieved by plating cells on an electrode array created by standard microfabrication methods. The electrodes are then biochemically functionalized with an arginine-glycine-aspartic acid (RGD)-terminated thiol. Application of a voltage pulse results in electrochemical desorption of the RGD-terminated thiols, triggering cell detachment. This method allows for the study of the full cascade of events from detachment to subsequent subcellular reorganization. Fabrication of the electrode arrays may take 1–2 d. Preparation for experiments, including surface functionalization and cell plating, can be completed in 10 h. A series of cell release experiments on one device may last several hours.
Programmed subcellular release triggers detachment of distinct parts of a cell from a patterned substrate in a spatially and temporally controlled manner1. Subcellular release is accomplished by plating cells on a device with an array of gold electrodes, typically 1–10 μm wide. An adhesion-promoting arginine-glycine-aspartic acid (RGD) peptide sequence2 is attached to the gold electrodes by a thiol (Au-S-R) linkage (Fig. 1). Detachment of specific regions of an adherent cell is triggered by applying a sufficiently negative voltage pulse, resulting in rapid release of the RGD-terminated thiol3. The release process is an electrochemical reaction involving reductive desorption of the thiol (Au-S-R + H+ + e− → Au + HS-R). Reductive desorption has been used to release molecules4-6, fluorescently labeled molecules3, nanoparticles and proteins7. The regions on the glass slide between the electrodes can be modified with polyethylene glycol (PEG) to minimize focal adhesion formation8,9. Each stripe is electrically isolated so that the RGD-terminated thiols from a single electrode can be desorbed independently of adjacent electrodes. This design enables the release of a subcellular section of an adherent cell spanning multiple electrodes.
The steps involved in performing subcellular release experiments are as follows. (i) Microfabrication of an array of individually addressable gold electrodes on a glass slide using conventional photolithographic techniques. Electrical contact is made by attaching a wire to a contact pad at the end of each electrode. (ii) Biochemical functionalization of the gold electrode array by immersing the slide into a solution containing RGD-terminated thiol. The glass surface may also be chemically functionalized with PEG. (iii) Plating cells that will span multiple electrodes. (iv) Recording phase-contrast or fluorescence time-lapse movies of cells released under live cell conditions (37 ° C, 5% (vol/vol) CO2, at least 75 % humidity). (v) Analyzing cell contraction on phase-contrast images by measuring cell length (or area) and fitting the data to ΔL(t)/ΔLm = 1 − exp( − (t − t0)/τ), where ΔL (t) is the change in cell length at time t (that is Linital − L(t)) divided by the maximum change in cell length (Linital − Lfinal). The induction time before retraction of the cell, t0, and the characteristic contraction time, τ, are two parameters extracted from fitting the data.
Cell types showing a polarized morphology with few cell-cell adhesions are most suitable for programmed subcellular release. Examples of cells released by this method in our laboratory include NIH 3T3 fibroblasts10, mouse embryonic fibroblasts11, AG04151 (human skin fibroblasts), HT1080 (human fibrosarcoma cells)12 and MDA-MB-231 (human breast cancer cells)13. Epithelial and endothelial cells tend to form small clusters when plated on surfaces14,15 and do not release readily from our micropatterned electrodes. Cells with a circular morphology and cells that form multiple nuclei, such as U2OS (osteoblast) cells, do not release well or are not viable after release.
There are very few methods to study cell detachment. Other methods are based on cleavage of actin stress fibers or pharmacological treatment. Laser ablation has been used to cut stress fibers in cells transfected with yellow fluorescent protein–labeled G-actin, allowing quantitative analysis of cell contraction16. Chromophore-assisted laser inactivation of fluorescently labeled α-actinin has been used to observe cell contraction after detachment of stress fibers from focal adhesions17. These techniques can be used to study the dynamics of cell contraction but do not allow investigation of the full cascade of events from the detachment of integrins from the extracellular matrix to the reorganization of actin filaments and focal adhesion proteins. In a pharmacological approach, nocadozole has been used to induce bulk microtubule disassembly, such that subsequent washout allows visualization of microtubule targeting of focal adhesion disassembly18. A drawback of this method is that it can lead to uncontrolled off-target effects.
Programmed subcellular release overcomes many of the limitations associated with existing methods. Triggering cell detachment by the release of the RGD-terminated thiol is sufficiently upstream to study the full cascade of events involved in cell detachment. Using a fluorescent probe, we have shown that electrochemical desorption of thiols occurs on time scales less than a few milliseconds3, much faster than processes associated with cell detachment and contraction. Spatial control is achieved, for a cell that spans multiple electrodes, by triggering detachment from one electrode. For example, an NIH 3T3 fibroblast spanning three electrodes can be released by one of the outermost electrodes. Programmed subcellular release has also been shown to mimic the spontaneous cell contraction or ‘tail snap’ of cells moving on a glass surface1,19.
Programmed subcellular release can be combined with other techniques, such as immunofluorescence microscopy, pharmacological inhibition studies and real-time live cell imaging of fluorescently labeled proteins to study biochemical and biophysical aspects of cell detachment. Immunofluorescence staining and statistical analysis of ensemble averages can be used to study the effects of cell detachment on focal adhesions20 and cytoskeletal proteins21. Molecular inhibition studies can be performed to deduce the role of a particular molecule or protein in the signaling cascade22,23. For example, cells incubated with 50 μM blebbistatin, a drug that inhibits myosin II and is known to influence cell contractility24,25, result in more than a 100-fold increase in contraction time. Realtime live cell imaging provides a method to probe the kinetics of detachment26. For example, we have imaged the dynamics of actin stress fibers during contraction by transfection with GFP-labeled actin27,28. Programmed subcellular release can also be used as a versatile tool to study the mechanical properties of normal and diseased cells11,29 and could be readily combined with complementary cell-mechanics methods, including particle-tracking microrheology30. The ability to combine programmed subcellular release with other tools and techniques provides new opportunities to study cell motility, embryonic development, the inflammatory response, wound healing and metastasis of cancerous cells.
Both device fabrication and surface modification are straightforward. Device fabrication involves standard microfabrication methods. The RGD-thiol coupling chemistry and surface functionalization are undemanding and all reagents are commercially available. The live cell chamber on the microscope may need to be modified to accommodate the electrode connections.
The volume of reagent required must be sufficient to submerge the electrode array in the well used for subcellular release. The volumes noted in this protocol are based on a well with an o-ring diameter of 19 mm (see Fig. 2). About 150 μl of 2 mM cyclo RGDfK and 150 μl of 2 mM NHS-thiol solutions should be prepared in DMSO under argon. Aliquots are stored in 0.5-ml Eppendorf tubes at − 20 °C as reaction of the succinimide group to the side chain amine of the lysine is moisture sensitive.
Live cell imaging can be performed by transiently transfecting cells with DNA plasmids that encode a GFP-labeled protein of interest and replating cells on release device 12–36 h after transfection. Cells can be released as described by this protocol with the consideration that many GFP-labeled proteins (with the exception of Lifeact-GFP actin) photobleach within seconds. Therefore, the exposure time and frequency of image acquisition must be adjusted accordingly. Insert the following steps between Steps 9 and 10 in the standard protocol.
These may be conducted in conjunction with cell release experiments by fixing and staining released cells after Step 16 as follows:
Troubleshooting advice can be found in Table 1
An example of release of an NIH 3T3 fibroblast cell is shown in Figure 4a. The red arrow indicates the release electrode on which the thiol molecules have been electrochemically desorbed at t = 0 s. After a brief induction time, t0 = 37.1 s, the cell begins to contract with a time constant τ = 34.2 s. The end-to-end distance of the cell along the contraction axis was measured and plotted as normalized contraction versus time as in Figure 4b. Figure 4c shows the normalized contraction for 22 cells with an average induction time of 57 ± 14 s and average contraction time of 39 ± 7 s.
Programmed subcellular release can be combined with other methods, such as immunofluorescence staining, molecular inhibition studies and real-time live cell imaging of fluorescently labeled proteins. An example of a cell that was released and stained for paxillin, actin and DAPI is given in Figure 5. Figure 6 reveals that releasing cells in the presence of myosin II inhibitor, blebbistatin significantly increases both the induction and contraction times. Real-time live cell imaging of fluorescently labeled proteins can be used to study the effects of cell release on cytoplasmic proteins such as actin and paxillin. Figure 7 depicts a cell release experiment conducted on a cell transiently transfected with Lifeact-GFP actin.
This work was supported in part by NIH Grants R21EB008259 and U54CA143868. B.W. acknowledges support from the Achievement Awards for College Scientists (ARCS) Foundation. We thank members of the Wirtz and Searson labs for technical advice and reagents.
AUTHOR CONTRIBUTIONS B.W. conducted experiments. B.W., D.W. and P.C.S. designed experiments, analyzed results and wrote the paper.
COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests.