Microscoop for Manipulation of Micro‐objects: Use of Fabricated Cantilever with Atomic Force Microscope

Fabricated cantilevers like pushpin removers used in an atomic force microscope allow the manipulation (picking up or putting down) of micro-objects. A customized cantilever with an appropriate spring constant also allows for the detection of the adhesion force between micro-objects and a surface, even if these objects are tightly bound. Developments and improvements in nanotechnology to realize means to freely manipulate nanometer- or micrometer-scale objects have provided great advantages in biophysics. The findings to date have revealed the significance of mechanical conditions in various biological phenomena, including how biological systems convert mechanical work into chemical energy and vice versa. Motor proteins and mechanosensory systems are one of the main targets in biophysics since they have a crucial role in energy conversion. Motor proteins drive all biological motion such as DNA replication, intracellular vesicle and organelle transport, and cell locomotion by molecular-scale forces generated by ATP (adenosine triphosphate) hydrolysis or by electrochemical potential.1 In the cell’s mechanosensory system, focal adhesion proteins that connect the extracellular matrix (ECM) to the cytoskeleton play a central role. The proteins are thought to change their structure and affinity toward their partner proteins in response to the surrounding mechanical tensile strength, which affects the subsequent biochemical signal transduction and finally cell differentiation.2-5 To investigate the biophysical properties of these proteins and the mechanics of the cell, quantitative manipulation of single molecules or single cells is required on a pico- to micronewton scale, as well as detection of the precise length of their extension or motion in response to the applied force. The most commonly used techniques are optical tweezers, magnetic tweezers, and atomic force microscopy (AFM).6 In a typical observation of a single molecule, one end of the mole­cule is attached to a surface either directly or indirectly via a partner molecule, and the free end is attached to a probe—an optically trapped bead, a magnetic bead, or a tip on an AFM cantilever—through which force is applied. These probes are also often used for single-cell observations, to manipulate the cell once the probes have adhered.7-11 The force ranges of conventional optical and electromagnetic tweezers cover from 0.1 to 100 pN, and from 0.01 pN to 10 nN, respectively:6 large enough to capture a single mole­cule. These two methods are also widely used in single-cell observations in combination with fluorescence microscopy to display how and where mechanical stress is transmitted and spread in a cell,12, 13 and to analyze the cell’s mechanical properties.11 It is suggested, however, that cells use multiple mechanisms to sense and respond to the level of applied mechanical stress.10, 11 Although cell culturing on silicone rubber is often used in molecular cell biological studies to apply a stretching force to cells, causing a large deformation of the cell, it cannot measure the strength of the applied force.14, 15 The conventional techniques cannot respond to the demand for larger force application and measurement. As such, there is a need for more sophisticated equipment to be developed to cover a wider range of forces (>10 nN). AFM allows the measurement of mechanical properties of various materials from single molecules to single cells in a larger force range by the choice of a cantilever with an appropriate spring constant, as well as allowing the capture of sample topography by scanning surfaces at the nanometer scale.16 AFM cantilevers come with various types of probes, such as probes with highly sharpened tips and probes with microspheres. The sharpened tip allows topographical profiles of a sample surface to be taken with a high spatial resolution, and limits the interaction between the probe and the sample to prevent multiple molecules on the surface from binding to the probe. A round tip is often used to estimate the elasticity of a soft material through force-indentation measurements.17 A spherical bead whose surface is modified with ligand mole­cules is often attached to the cantilever and used to interact with a number of receptors in the cell membrane.18 However, since the probe is fixed on the cantilever, the throughput of measurements is low, especially when it takes a long time to form a strong enough bond between the probe and the cell for later manipulation of the cell. Here, we introduce a new tool to manipulate micro-objects using AFM, which can cover a wide force range and yield high-throughput measurements. A novel feature of recovering intracellular components at the bead adhesion site is also described. An important aspect of the specially designed AFM cantilever introduced in this study is a semicircular cut at the edge, which will be called a hole hereafter. The hole was cut from commercially available arrowhead-shaped cantilevers by a focused ion beam such that its diameter was slightly smaller than the microspheres to be manipulated (Figure 1). The circular hole was intended to hook the micro-object and pick it up (Scheme 1). At first, the cantilever approaches the surface near the microsphere by the piezo extension (Scheme 1A→B). The cantilever then moves laterally and positions itself just under the micro-object (Scheme 1B→C). As the cantilever is moved to contact the surface, the pushing force of the piezo extension is regulated. This force is estimated from the cantilever deflection. After setting the cantilever under the microsphere, the microsphere is trapped in the hole of the cantilever and is picked up by retraction of the piezo (Scheme 1C→D). The hooked bead may be moved and placed in another position on the surface by reversing the scheme (Scheme 1D→C→B→A). Scanning electron micrographs of a “microscoop” cantilever-like pushpin remover. The edge of a fabricated cantilever (A) has a hole that is slightly smaller in diameter in comparison to a manipulated bead. The hole is cut from a commercially available arrowhead cantilever (B) by the focused ion beam technique. Inset: entire image of the arrowhead cantilever. Bar represents 10 μm in (A) and (B), and 50 μm in Inset. Experimental design for manipulation of the cantilever to pick up, move, and drop micro-objects by AFM. The microscoop cantilever deflection and Z position are detected and regulated by an optical lever system and a piezo scanner. During the picking-up operation (open arrows), the cantilever initially approaches the surface on which a bead lies, and comes into contact with the bead at a user-defined force (A→B). Next, the cantilever slides on the surface such that the hole at the edge of the cantilever comes under the bead (B→C). Finally, retraction of the piezo causes the bead to be hooked in the hole (C→D). The opposite operation allows the bead to be put down (D→C→B→A: indicated as solid arrows). We tested this procedure by selectively moving polystyrene microspheres around a surface, placing them in the shape of a pentagon using the fabricated AFM cantilever in phosphate buffered saline without divalent cations (PBS(-)) (Figure 2). Figure 2A shows the phase-contrast images of the beads before and after manipulation. The beads were placed at the apices of the pentagon in numerical order according to the arrows. Figure 2B shows the force on the cantilever (top) and the piezo position (bottom) over time during the operation. The piezo extension was regulated to maintain 50 nN of pushing force at the surface (0 μm Z position). Each bead was raised to a height of 10 μm, moved, and placed. The deflection of the cantilever with and without the attached bead showed little difference, suggesting that the net force acting on the single bead was negligible. Figure 2C shows the time-lapse image obtained during pick-up of the fourth bead. The bead was focused out after it was raised to 10 μm in height (at +9 s and +10 s). We thus confirm that the fabricated cantilevers are suitable for manipulating micro-objects. We also tested this technique in air and found that the cantilever could pick up dried glass or polystyrene beads from the glass surface (data not shown). Thus, the manipulation was observed to be effective both in air and liquid. Placement of microspheres in the shape of a pentagon using a microscoop cantilever. A) Phase contrast images before and after the placement of five beads in the shape of a pentagon. The numbers near the beads represent the order in which they were manipulated. B) The cantilever deflection (top) and the Z position (bottom) over time. The numbers above the deflection curve correspond to the number assigned to each bead. C) Time-lapse phase-contrast images of the fourth bead as it was picked up, indicated by a gray zone in (B). Bar represents 10 μm. Next, we analyzed the ECM adhesion force and the cell extension when the probe was detached from cell surface (Figure 3). Figure 3A depicts the design of the experiment. Rat fibroblasts were spread on a human fibronectin-coated glass surface. The cells’ stress fibers and focal adhesion were both marked with fluorescence proteins. Here, we marked β-actin, a main component of stress fibers, with an orange fluorescent protein (KO1: Kusabira Orange 1), and paxillin, one of the focal adhesion proteins, was marked with a green fluorescent protein (AG1: Azami-Green 1). Paxillin is one of the scaffold proteins in the focal adhesion, and has binding sites with diverse proteins including β-integrin, focal adhesion kinase, and vinculin which is one of the actin filament-binding proteins.19 It is thought to be a key factor in the mechanosensory system.20 Picking-up process of beads adhered on fibroblast cells. A) Schematic view of the manipulation of a hrFNIII7–10-coated bead on a cell. B) Micrographs of rat fibroblasts with hrFNIII7–10-coated beads adhered, expressing AG1-paxillin and KO1-β-actin with (e–h) or without (a–d) cytochalasin D. Images (a, e), (b, f) and (c, g) show phase contrast images (Ph), AG1-paxillin fluorescence, and KO1-β-actin fluorescence, respectively. Images (d) and (h) are merged from (a–c) and (e–g). Gray, green, and red in the merged images indicate the phase contrast, AG1-paxillin fluorescence, and KO1-β-actin fluorescence, respectively. The insets in (b), (c), (f), and (g) are magnified images around the bead as indicated by the boxes. C) Micrographs of the beads after they were picked up from the cells expressing AG1-paxillin and KO1-β-actin, with (d–f) or without (a–c) cytochalasin D. Images (a, d), (b, e), and (c, f) show phase contrast images, AG1-paxillin fluorescence, and KO1-β-actin fluorescence, respectively. The bar represents 10 μm. D) Force versus bead-sample separation curves derived from bead-picking-up experiments for intact cells (red) and cytochalasin D-treated cells (blue). Carboxylated, functionalized polystyrene microspheres were covalently modified with human recombinant fibronectin type III domains 7–10 (hrFNIII7–10) and allowed to adhere on the cell surface in advance of the manipulation experiment. We confirmed that the beads were bound to the cells via focal adhesion proteins and the cytoskeleton. As shown in Figure 3B, actin stress fibers and paxillin accumulated around the microsphere. Then the bead was lifted at 5 μm s−1 and completely separated from the cell. Images of a typical separated bead and the derived force versus bead-sample separation curves are shown in Figure 3C and D, respectively. The images reveal that the fluorescently labeled focal adhesion proteins and actin cytoskeleton successfully adhered to the bead surface. The images also show that the focal adhesion proteins co-localized with the actin cytoskeleton. This is the first clear demonstration of the recovery of focal adhesion proteins using an AFM cantilever. In the previous attempt to extract membrane proteins with AFM, intracellular components were not identified.21 We performed a similar experiment using cells that expressed fluorescently labeled αV-integrin, which forms heterodimers with one of several kinds of β-integrin, as an ECM receptor in the focal adhesion,22 and also confirmed the recovery of αV-integrin (data not shown). The force versus bead-sample separation curve shows the mechanical properties of the probe-adhering material. The pull of the probe bead initially increased internal tensile force in the cell. Then the deflection peaked (i.e., the maximum adhesion force was reached), followed by a decay indicating the destruction of the intracellular cytoskeletal network. Some of the force versus bead-sample separation curves have a short lag near the start of the picking-up operation, reflecting the process prior to the cantilever holding up the bead. Some of the other curves apparently have two peaks: the first peak, near the start of pulling, appears extremely sharp; it is supposed that the edge of the cantilever was stuck in the cell rather than set on the cell (Figure S1, Supporting Information). We pulled up beads from the cell surface one after another and made similar measurements for each bead, attaining a high overall throughput. To confirm that this technique can detect the difference in mechanical properties of probe-adhered materials, we examined the cells whose actin network was collapsed. Cytochalasin D, which is one of the inhibitors used to induce depolymerization of actin filaments, can disrupt the stress fibers in the cell, causing a decline in cellular stiffness (Figure 3B). The maximum adhesion force, work of detachment, and maximum sample extension were estimated from the curves (except for those that have an initial sharp peak larger than their second peak) (Figure 4 and Figure S1, Supporting Information). The variation in the data is partly attributed to the different contact areas between the beads and the cells. The different categories of actin stress fibers in the cell may also cause some variation in the adhesion force.23 However, there is a significant difference in the maximum sample extension between the intact cells and the cytochalasin-D-treated cells. As expected, the maximum sample extension from the cytochalasin D-treated cells was greater than that from the intact cells, reflecting a decrease in cellular stiffness due to cytochalasin D treatment. With the use of a standard linear solid model consisting of a spring with spring constant k1 parallel to a second spring with spring constant k2 and a dashpot with viscosity coefficient γ,24 the force-separation curves were fitted, excluding initial lags and large rupture events. This analysis showed the significant difference in stiffness and viscosity between the intact cells and the cytochalasin-D-treated cells. The distributions of the mechanical and viscoelastic parameters in the intact cells (open boxes) and the cytochalasin-D-treated cells (gray boxes). Box plots of the distribution together with the median of the respective parameter are shown. The p-value describes the probability that the distributions are identical in both setups, calculated using the Mann–Whitney test. All parameters here are estimated from the force versus bead-sample separation curves (Figure S1, Supporting Information). In summary, a new microscoop AFM cantilever was used to manipulate microspheres and detect binding forces on the order of micronewtons between the probes and the cells. This method is a high-throughput approach for analyzing the binding force in comparison to that of probe-glued cantilevers because the probes are separated from the microscoop cantilever in advance. The technique in which the cantilever directly picks up the probe allows various kinds of probes to be manipulated and was performed in two different environments in our case. However, the cantilever dimension, particularly its thickness, must be considered: it is expected to be difficult for a thicker cantilever to approach from beneath the micro-object and to pick it up. To obtain a larger spring constant of the customized cantilever, its width and length may be respectively increased and decreased while its thickness is kept small.25 We also demonstrated that the fabricated cantilever allows the recovery of proteins interacting with ligands on a bead. This will help to extract and identify proteins that respond to the ligand on the probe. Cantilever Fabrication: Commercial cantilevers (TL-CONT, TL-FM, ATEC-FM) were all from Nanosensors, Switzerland. The edge of these cantilevers had a hole (≈9 μm in diameter) that was cut using a FB2100 FIB instrument (Hitachi High-Technologies Corporation). Bead Preparation: Polystyrene beads modified with carboxyl groups (10 μm in diameter; Polysciences, Inc.) and glass beads (10–30 μm; Polysciences, Inc.) were used for the demonstration of bead manipulation in Figure 2. For the manipulation of beads adhered to cells, the carboxylate beads were modified with hrFNIII7–10 by a cross-linking reaction, as described in the Supporting Information. Bead-Cell Binding Assay: Tested cells all expressed either AG1-paxillin and KO1-β-actin, or AcGFP1-β-actin and αv-integrin-TagRFP. The cells were harvested by 0.05% trypsin-0.2 mg mL−1 EDTA(ethylenediaminetetraacetic acid) (Invitrogen) and suspended in a CO2-independent medium (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen) and 4 mM L-glutamine (basal medium), together with the hrFNIII7–10-coated beads described above. The conjugate was transferred to a glass-bottom dish (IWAKI) covered with human fibronectin (CalBioChem), and then incubated at 37 °C, 5% CO2 for 2 h. During incubation, the cells were allowed to adhere on the fibronectin-coated glass surface and to bind to the hrFNIII7–10-coated beads. After incubation, the medium was replaced with fresh medium to remove unbound beads. After further incubation for more than 1 h, bead mani­pulation was performed. For preparation of cytochalasin-D-treated cells, the cells were further incubated in basal medium containing 1 μM cytochalasin D (Wako). Microsphere Manipulation with AFM: A NanoWizard II Atomic Force Microscope (JPK Instruments) mounted on an Axio-observer D1 inverted microscope (Carl Zeiss) was used for the mechanical measurements and optical observations throughout this work. The AFM head was equipped with a 15 μm Z-range linearized piezoelectric ceramic scanner and an infrared laser in the optical lever system for the detection of cantilever deflection. For long Z-range scanning, a 100 μm Z-range linearized piezoelectric scanner mounted in the sample stage, CellHesion Module (JPK Instruments), was used. The sensitivity of the optical lever system was calibrated and the cantilever spring constant was determined in situ before every experiment, using the thermal noise method.26, 27 Within the reported uncertainty of this method (≈10%), the cantilever spring constants were found to agree with the manufacturer’s specifications. A closed-loop system was used to control the speed and positioning of the vertical piezo elements.28 Force spectroscopy experiments were performed at room temperature. The piezo-traveling distance (Z position) and cantilever deflection were recorded in the force mode operation of the AFM. The result was converted to the relation between force and bead-sample separation, where force was calculated as cantilever deflection [V] × cantilever sensitivity [m V−1] × spring constant [N m−1], and bead-sample separation was calculated by subtracting cantilever deflection [m] (cantilever deflection [V] × cantilever sensitivity [m V−1]) from piezo-traveling distance [m]. Supporting information is available from the Wiley Online Library or from the author. The authors thank Dr. I. Harada and Dr. H. P. Erickson for supplying plasmid DNA for this research. Technical support for the focused-ion beam (FIB) technique was provided by the foundry staff of the International Center for at the for This work was by a for to of to are as are not or are made available as by the The is not for the or of information by the than be to the for the