2.1.

Optical Setup

Multiple optical point traps are generated by using a 532-nm green laser (Nd:YAG 5 W, Spectra-Physics, Newport Corporation, Irvine, California) coupled with a Nikon inverted light microscope and integrated with a Biorryx system (Arryx Inc., Chicago, Illinois). Though the 532-nm green laser is known to be harmful to cells, it presents a worst-case scenario that makes it easier to detect differences in photodamage with different manipulation methods. Our indirect pushing method keeps cells viable even when hundreds of milliwatts of laser power at 532 nm are used for manipulation. A spatial light modulator has been built on this setup to customize the multitrap arrangement within 3-D spaces. With a Nikon 60X NA 1.4 oil immersion objective, the Biorryx system can arrange up to 100 point traps within the operating region of about $100\times 100{\mu \mathrm{m}}^2$ and about 10 μm above or below the focal plane. The same objective was used for both optical trapping and imaging. The cell manipulation process and long-term cell shape dynamics were imaged at 15 frames per second by a CCD camera (Foculus, Aegis Electronic Group, Gilbert, Arizona) and at 1 frame per second by a CCD camera (Ueye, IDS GmbH, Obersulm, Germany), respectively.

2.2.

Preparations of Dictyostelium Cell, Human MCF-10A Mammary Epithelial Cells, and Silica Beads

D. discoideum lives as a single-cell amoeba when there is sufficient nutrition in the environment. Upon starvation, wild Dictyostelium is able to sense its neighboring cells and collectively migrate to an aggregation center, and eventually forms a slug, which helps the whole cell group move faster during the food searching. The wild Dictyostelium (AX3) cells were grown in an HL-5 medium at concentrations no higher than $5·{10}^6\text{cells}/\mathrm{mL}$ at 21°C.³⁶ In these experiments, we starved and developed cells for five hours in a development buffer (DB: 5 mM ${\mathrm{KH}}_2{\mathrm{PO}}_4$; 5 mM ${\mathrm{Na}}_2{\mathrm{HPO}}_4·7{\mathrm{H}}_2\mathrm{O}$; 2 mM ${\mathrm{MgSO}}_4$; 0.2 mM ${\mathrm{CaCl}}_2$) with pulses of 75 nM of cAMP every 6 min, as described in other papers.^37,38 Developed cells were harvested after 5 h by centrifuging 500 μL of liquid from a develop flask at 9000 rpm for 3 min. The cell pellets were dissolved in 500 μL of distilled water or phosphate buffer (PB: 5 mM ${\mathrm{KH}}_2{\mathrm{PO}}_4$; 5 mM ${\mathrm{Na}}_2{\mathrm{HPO}}_4·7{\mathrm{H}}_2\mathrm{O}$).

The MCF-10A cells were grown in an incubator in which the humidified atmosphere was kept at 37°C and 5% ${\mathrm{CO}}_2$, in the DMEM/F12 media with 5% horse serum, $10\mu \text{g}/\mathrm{mL}$ insulin (Invitrogen), $10\mathrm{ng}/\mathrm{mL}$ epidermal growth factor (EGF) (Peprotech, Rocky Hill, New Jersey), $0.5\mu \text{g}/\mathrm{mL}$ hydrocortisone (Sigma, St. Louis, Missouri), and $100\mathrm{ng}/\mathrm{mL}$ cholera toxin (Sigma, St. Louis, Missouri). For these experiments, cells were harvested from the culture flask by adding Trypsin (Invitrogen, Life Technologies, Carlsbad, California), and the supernatant were centrifuged at 350 rpm for 5 min. The cell pellet was dissolved in 6 mL growth media, and 10 μL cells were added to a well of a μ-slide eight-well chamber (Ibidi, Martinsried, Germany)

Silica beads (5 μm) are used as optical trapping assistants in indirect manipulation approaches. A microbead solution was prepared by mixing 2 μL of an original silica microspheres solution (Bangs Laboratories, Fishers, Indiana) with 1 mL distilled water.

2.3.

Suspended Cells and Beads

In order to have enough time to manipulate cells and beads, the electrostatic force is utilized in experiments with Dictyostelium cells to keep cells and beads from sticking to the glass surface. The surface of cover glass, Dictyostelium cells, and silica beads are all negatively charged, so they repel each other because of the Coulomb force when they get close together. After adding cells to a normal buffer, as in a standard experiment protocol, cells gradually settle to the surface from suspension. Once placed on the surface, they adhere and start to spread; i.e., they form a growing contact area with the surface. The reason that they are able to overcome the repellent Coulomb force is that the cations in the buffer screen out the negative charge on cell membranes and surfaces.³⁹ In this study, we add cells and beads to distilled water so that the Coulomb force between cells and a glass surface can keep them suspended for manipulation. When studying cell–cell interaction, the lack of charge screening may also lead to repulsion between negatively charged cells. Thus, polyethyleneglycol (PEG) coated glass slides (from MicroSurface Inc., St. Louis, Missouri) are also used in the experiment shown in Sec. 3.3 to keep cells from adhering to the surface.

3.1.

Viability of Cells After Optical Manipulation

We used three manipulation approaches on Dictyostelium cells: direct trapping, direct pushing, and indirect pushing. To show the configurations of different manipulation approaches and illustrate how the focus laser cones intersect with trapped beads or cells, we gave schematic figures of these three methods, as shown in Fig. 1(a)–1(c). The diameter of the silica beads is 5 μm. The cells are assumed to be spherical, with similar size to the silica beads. The angular aperture (the angle between the cone and the direction of the laser beam) is 67° for the objective (comparable to other high NA objectives needed for optical trapping) used in this study. Unlike direct trapping [Fig. 1(a); see Video 1 for a representative direct trapping process], where the laser beam directly focuses on the target cell, the direct pushing approach [Fig. 1(b)] limits a cell’s exposure to light (see Video 2 for a representative direct pushing process). As shown in Fig. 1(b), in equilibrium, with the trapped bead at the focal point of the laser, 6.7% (in volume) of the target cell is exposed to the laser cone. However, this is the lower limit of the volume exposed to laser light. If cells are larger than beads (5 μm in diameter) or can deform their shape, the exposed volume can be higher. For example, Dictyostelium cells and epithelial cells are typically over 15 μm in diameter, and the Dictyostelium cells actively deform their shapes. Using a bigger bead (at least 8 μm in diameter in this example case) could possibly keep the cell out of the laser cone, but bigger beads cannot be stably trapped against its gravitional force with our optical tweezer system. The third method, indirect pushing, is shown in Fig. 1(c): The target cell is pushed by an intermediate bead, and the intermediate bead is pushed by another bead that is directly trapped by the laser beam. These two beads and the cell do not stick to each other (e.g., if the manipulation involves putting cells together into larger collections, beads have to be removed from the cell group). The directions of the forces that the directly trapped bead applies to the intermediate bead are tuned depending on the relative position of the intermediate bead and the target cell. How to push indirectly on the cells in a reliable manner has been described elsewhere.⁴⁰ When the target cell is pushed indirectly, the cell remains at least several microns away from the directly trapped bead and thus is not directly exposed to the light cone of the trapping laser (see Video 3 for a representative indirect pushing process).

Video 1

Direct pushing on Dictyostelium cells (MOV, 3.0 MB) [URL: http://dx.doi.org/10.1117/1.JBO.18.4.045001.1].

Video 2

Indirect pushing on Dictyostelium cells (MOV, 10.5 MB) [URL: http://dx.doi.org/10.1117/1.JBO.18.4.045001.2].

Video 3

Direct trapping on Dictyostelium cells (MOV, 7.08 MB) [URL: http://dx.doi.org/10.1117/1.JBO.18.4.045001.3].

Viability tests of these three manipulation methods were performed on a set of suspended Dictyostelium cells (21 cells for direct trapping, 18 cells for direct pushing, and 26 cells for indirect pushing). Videos 1–3 are representative videos that present the survival of unmanipulated cells with the manipulated cells in the same field of view. As shown in these videos, the cells were active before manipulation. To maintain constant conditions, the output laser power was set to 0.2 W, and the manipulation times remained between 40 and 50 s [i.e., $(\text{laser power})\times (\text{exposure time})=8\mathrm{t}\mathrm{o}10\mathrm{J}$]. Four criteria were used here to estimate the photodamage [Fig. 1(d)–1(f)]: survival, the generation of blebs on the membrane, the generation of protrusions during manipulation, and the generation of protrusions after manipulation.

Note that the 532-nm green laser that we used is known to be harmful to cells. Thus, our study presents a worst-case scenario that makes it easier to detect differences in photodamage with different manipulation methods.

Dead Dictyostelium cells [shown in Fig. 1(g)] are easy to distinguish from living cells because they do not keep their membrane intact. All cells survived after all three manipulation approaches (data not shown). However, some living cells generate membrane blebs during manipulations, which usually indicates that cells cannot retain their cortical tension and are unhealthy.^41–43 A representative cell with a membrane bleb is shown in Fig. 1(h). As shown in Fig. 1(d)–1(f), the percentage of cells that generate blebs during trapping drops from 67% to 0% from a direct trapping approach to an indirect pushing approach.

Dictyostelium cells migrate, and they need to extend protrusions in order to move forward. When Dictyostelium cells migrate on a surface, they generate new protrusions every 20 s on average.²⁷ Thus, during the 45-second manipulation in this viability test, a healthy cell normally will be able to generate protrusions. But, as shown in Fig. 1(d)–1(f), only 9% of the cells under direct trapping extend protrusion during the manipulation, and this number increases to 50% and 80% when using the direct pushing and indirect pushing methods, which reveals that only indirect pushing approach does not have significant negative impact on cells in terms of generating protrusions in a short time of optical trapping.

Nevertheless, photodamage could have a long-term effect on cells. To study that, cells have been observed for at least 5 min after each manipulation. As shown in Fig. 1(d)–1(f), only after indirect pushing, most of the cells (77%) are found to be able to generate protrusions in 5 min after manipulation, while only 9% of the cells could generate protrusions after the other two types of manipulations. Thus, the indirect pushing approach shows no significant long-term photodamage on Dictyostelium cells.

3.2.

Shape Analysis of Cells After Optical Manipulation

The viability test indicates that the indirect pushing method does not cause long-term photodamage in the generation of new protrusions, but the shape dynamics of the cell could be changed. As we learned from crawling Dictyostelium cells, protrusions should not only be generated, but also travel on cell membranes like a wave. Thus, a shape dynamics analysis method that we previously developed has been used to analyze shapes of cells after optical trapping (see details in Ref. 27).

The shapes of cells after direct or indirect pushing are analyzed as shown in Fig. 2. In Fig. 2(b), indirect pushing of a cell toward another cell is shown. In Fig. 2(a) and 2(b), Dictyostelium cells maintain a polarized, elongated shape. The polarized end that generates new protrusions is considered as the “front” (i.e., “head”) of the cell, and the other end is the “back” (i.e., “tail”) of the cell.

Fig. 2

Shape analysis of Dictyostelium cells after optical manipulation. (a) A cell that is pushed directly by a silica bead. (b) A cell that is pushed indirectly by a silica bead through an intermediate bead. (c) Extracted outlines of a representative polarized cell (red: positive curvature; blue: negative curvature). (d) Overlaid shapes of a cell manipulated by direct pushing (20 s apart; red: initial shape). (e) Overlaid shapes of a cell manipulated by indirect pushing (2 s apart; red: initial shape). (f) Curvature versus time plot of a cell manipulated by direct pushing. (g) Curvature versus time plot of a cell manipulated by indirect pushing. Scale bars in (a)–(c) are 5 μm.

To analyze shape dynamics, 400 points on the cell boundary are extracted by an active contour algorithm.⁴⁴ The curvature along each point on the boundary is calculated and used as a metric for shape. Figure 2(c) shows an image of a representative cell with its extracted boundary, colored with curvature value. Two main high-curvature regions are seen (red indicates high-positive curvature): one in the front (black arrow), and the other in the back (white arrow). In addition, there are small high-curvature regions (small red regions, red arrows).

To visualize shape dynamics, we overlay cell shape outlines from different points in time. Figure 2(d) shows overlaid cell shapes after direct pushing [the cell is the same as in Fig. 2(a)]. As we can see, although the cell retains its polarity and looks similar to a healthy cell [shown in Fig. 2(c)], its overall shape does not change notably in 80 s, indicating that the cell’s active shape changes after direct pushing. After indirect pushing, however [the cell in the Fig. 2(b)], cells continue to change shape [Fig. 2(e)]. Cell shape dynamics do not seem to be affected by indirect pushing. Localized protrusions travel from the front to the back of the cell (blue arrow), consistent with our prior observations.

To visualize and quantify shape dynamics over minutes, we plot a kymograph (space-time plot) where color indicates local curvature (red: positive curvature; blue: negative curvature; curvature value shows in the color bar). Each vertical line in the kymograph represents one shape outline at a different time, for a total of 4.5 min [Fig. 2(f) and 2(g)]. After direct pushing, the shape of the cell remains unchanged with time: The front and back, as well as additional small protrusions form multiple straight red lines in Fig. 2(f). After indirect pushing [Fig. 2(g)], tilted red lines are visible that start at the front of the cell, indicating protrusions that travel as a wave from the front of the cell to the back (white dotted lines), similar to what we reported previously on healthy Dictyostelium cells.³⁴

3.3.

Indirect Pushing Allows for Studies of Cell–Cell Adhesion

Next, we demonstrate an example of how indirect pushing may be used to address a biological question: namely, how cell–cell contact affects cell behavior. Often, cell–cell contact is studied when cells are anchored to a surface, and therefore, cell–cell contact competes with cell-surface contact. However, many cells can be brought into suspension or be prevented from anchoring to the surface, at least temporarily. In that case, they also lack the ability to push off the surface and move, and thus they need to be manipulated in order to get in contact with other cells in a controlled way.

We first investigated cell–cell adhesion in cells suspended in distilled water. Unlike mammalian cells, the amoeboid cells we investigate (Dictyostelium) have special organelle that allow them to survive in distilled water. As shown in Video 3, when two cells were pushed together, we find that they did not form stable cell–cell adhesion. Instead, after indirect pushing stopped, they moved apart immediately, though cells were viable by our strictest standards. They maintained their shape dynamics until they moved out of the field of view 19 min after manipulation.

To assess whether this lack of cell–cell adhesion was related to the absence of a buffer medium that screens charges and limits repulsive forces between cells, we used cells in a buffer, but on PEG-coated glass slides to prevent cell-surface adhesion. We find that cells in buffer are able to stick to others when they are pushed into contact. This process is shown in Fig. 3. In Fig. 3(a) and 3(b), two suspended cells are generating new protrusions (black arrows) before manipulation. A bead (white arrow) is directly trapped by an optical trap and moved toward an intermediate bead and cell A. The indirect pushing process is shown in Fig. 3(b)–3(d) as these two cells are arranged along their polarization direction (head to tail) and are pushed together to form direct cell–cell contact. This manipulation process lasts for about 40 s. After the manipulation stops, these two cells still extend protrusions (black arrows) and retain dynamic shapes. The cells remain in contact and eventually form a clump several minutes after manipulation, as shown in Fig. 3(h). We have carried out this procedure successfully eight times, demonstrating that it is possible to generate cell–cell contact with controlled polarity for biophysical investigations.

Fig. 3

Indirect pushing of two Dictyostelium cells allows testing of cell–cell adhesion with controlled cell polarity. (a) and (b) A bead is directly trapped by the optical trap and moved toward another bead and cell A. (c) and (d) Cell A is pushed indirectly by the trapped bead through the intermediate bead. (e) The manipulation stops after 1 min, and the two Dictyostelium cells are able to stick together. (f)–(h) The two adhered Dictyostelium cells extend protrusions and form a clump in the following 8 min. Scale bar: 5 μm.

3.4.

Manipulating Epithelial Cells via Indirect Pushing

The indirect pushing method introduced in this paper is useful for nonadherent cells. But it also can be used on adherent cells, taking advantage of the slow process of cell-substrate adhesion. Mammalian cells usually adhere to surfaces or an extracellular matrix, and utilize cell-surface adhesion to migrate. When adhered to a surface, these cells cannot be manipulated with optical tweezers since adhesion forces are in the nN range, which is stronger than optical forces. However, there can be a delay of up to half an hour between the time that cells settle to the surface and the time that they start to adhere, as reported in Ref. 45, among others. The delay time can be adjusted by varying the surface coating concentration. For example, we studied MCF-10A human breast cells (see Fig. 4). After placement in an eight-well chamber (Ibidi), it was possible to push cells indirectly with silica beads before the cells stick on the surface for more than 30 min.

Fig. 4

Indirect pushing of an MCF-10A cell. (a) A bead is directly trapped directly by the optical trap and pushed to another bead, and moves toward an MCF-10A cell. (b)–(e) Cell is pushed indirectly from a different direction by the trapped bead through the intermediate bead. (f) After manipulation, the trapped bead is released from trapping. Scale bar in (a) is 10 μm.