Magnetic particle biosensors 5

Chapter Magnetic Twisting Cytometry Chapter MAGNETIC TWISTING CYTOMETRY 5.1 INTRODUCTION Cells, the smallest units of life, are complex living matter which comprise of numerous components capable of performing a vast variety of functions that include metabolism, control, sensing, communication, growth, remodeling, reproduction and apoptosis (programmed cell death) [1, 2]. In order to perform all these functions, cells undergo or control a series of intra- and extracellular events, many of which are directly or indirectly driven by certain mechanical phenomena (or forces) experienced by the cell [3]. As cells exhibit distinct mechanical characteristics which govern a host of essential biological processes that are crucial for cell survival, the study of cell mechanics will allow us to understand their mechanical properties, mechanical interactions with their environment, and ultimately the role of cell mechanics in biological function. A major field of interest in cell mechanics is the study of the cytoskeleton; a complex yet organized structure, whose mechanical properties are known to be a factor in the maintenance of cell shape and cellular functions such as cell migration, mechanosensing, and motility [4-6]. The innate relationship between the proper function of living cells and the mechanical stresses that it experiences, makes the study of mechanotransduction; the process by which mechanical forces are converted into changes 98 Chapter Magnetic Twisting Cytometry of intracellular biochemistry, important because of its significant implications in biotechnology and human health [7]. Although it is accepted that living cells sense and respond to mechanical stress, the molecular mechanism via which this process occurs is not well understood [8, 9]. Numerous computational models exist for cytoskeletal mechanics, and an even greater number of experimental techniques developed to try to quantify the process, most of which typically involve a mechanical perturbation of the cell in the form of either an imposed deformation or force, with subsequent observation of the static and dynamic responses of the cell [3]. Many experimental techniques have been used for the study of cell mechanics. atomic force microscope (AFM) [10], cell poking [11] and magnetic twisting cytometry (MTC) belong to a class where local probes are used to deform only a portion of the cell, whereas methods like micropipette aspiration and optical tweezers are used for the mechanical loading of an entire cell [12]. For applying mechanical stresses on entire population of cells, shear-flow methods and stretching devices have been used [13] (Fig. 5.1.1). As with the various methods of probing cell mechanics, many different models have been proposed to describe the mechanics of living cells, from modeling the cytoskeleton as a mechanical elastic [14], viscoelastic [15], porous gel or soft glassy material [16], to a network incorporating discrete structural element that bear compression [17]. This diversity in the proposed mechanical models stem from the fact that the length scale in which cellular mechanics is probed and the biomechanical issues that are targeted varies greatly from experiment to experiment. 99 Chapter (a) (c) Magnetic Twisting Cytometry AFM Micropipette (b) H applied (d) Magnetic bead Red blood cell Optical trap Silica bead (e) (f) Shear flow Cell stretching Focal adhesion sites Soft membrane substrate Fig. 5.1.1 Schematic showing the common methods for probing cell mechanics. (a) Atomic force microscopy. (b) Magnetic twisting cytometry. (c) Micropipette aspiration. (d) Optical tweezers. (e) Shear flow technique. (f) Substrate stretching device [1]. Based on the results obtained in Chapter 3, we have observed that the direction of magnetization of a bead can be calculated for any bead position above a sensor. Hence, in this Chapter we propose a modified version of MTC for the study of single cell mechanics, where a spin valve (SV) sensor is used to measure the angle of twist in a single ferromagnetic particle attached to a cell surface. This chapter will start off by describing the method of MTC. The fabrication methods adopted to use the SV sensors for cell culture will be discussed in section while section and will describe the cell culture protocols and experimental setup. 100 Chapter Magnetic Twisting Cytometry Section will cover our AFM calibration experiments, with our magnetic twisting results presented in section and 8. 5.2 MAGNETIC TWISTING CYTOMETRY (MTC) Magnetic methods for measuring intracellular viscoelasticity span back to as early as 1922, where Freundlich and Seifriz used a micromanipulator together with a microscope, to insert a magnetic particle into a large cell (sand dollar egg). Using a magnetic field gradient produced by an electromagnet, a force was applied on the bead and its displacement measured, giving rheological information of the cell [18]. This manipulation method is the basis of modern magnet-based microrheology systems, with many groups further modifying and enhancing the technique for the investigation on cell mechanics. In the study of cell mechanics, both superparamagnetic and ferromagnetic particles have been used, which can either be inserted or bound to cell surfaces. In an applied field, paramagnetic particles only experience a translational force in a field gradient (but no torque), while ferromagnetic particles can experience both a force and a torque [19, 20]. A problem faced in applying a force on a magnetic particle is the difficulty in constructing a well controlled field gradient. In order to avoid this complication, the method of applying a pure torque (twisting) to the magnetic particles has been adopted, as homogenous fields can be created rather easily [8, 21]. The method of applying a pure torque to magnetic particles is commonly known as magnetic twisting cytometry. In a typical twisting experiment (Fig. 5.2.1) a large number of ferromagnetic particles are attached to a large ensemble of cells (20,000 – 101 Chapter Magnetic Twisting Cytometry 40,000) and magnetized in a pulse of strong magnetic field (along direction x). A weaker twisting field orientated at 90º to the induced dipoles (z direction) is applied to rotate the beads, and the remanent magnetic fields measured in a lock-in mode with a magnetometer [8]. By continuously measuring the component of the remanent field in the horizontal plane (x), the angle of rotation can be obtained and used to calculate the equivalent Young’s modulus E of the intracellular medium [22]. (a) Before magnetization (b) After magnetization (c) During twist z x Magnetic particle ~1000G pulse Magnetometer Substrate Cell H 0-25 G twist (1 minute) Fig. 5.2.1 Schematic illustrating the general concept of MTC. (a) Ferromagnetic beads are bound to cell surfaces, with unbound beads washed off. (b) A brief but strong magnetic pulse is applied (~ 1000 G) in the x axis to fix the magnetization direction of the beads. (c) A smaller twisting field is applied in the z axis direction, and the change in the x component of the remanent magnetic field vector is measured with a magnetometer, giving the angle of bead rotation [8]. The advantage of MTC over optical tweezers and AFM in probing cell mechanics is that it is non-invasive. A disadvantage is that traditional MTC is only able to measure the twist of a large ensemble of beads [8], and measurements cannot be performed when the remanent magnetic field becomes too small [22]. This problem has since been overcome by using optical methods for the detection of single bead motion, with the 102 Chapter Magnetic Twisting Cytometry ability of measuring bead displacement with an accuracy of nm [16]. Other variations of magnetic methods for probing cell mechanics have also been introduced, with the likes of magnetic tweezers [23], electromagnetic pulling cytometry (EPC) [9], and laser tracking microrheology (LTM) [24]. The study of cell mechanics is extremely complex as the mechanical properties of cells differ greatly depending on the receptor probed for analysis. In typical MTC experiments where beads and cells are analyzed as an ensemble, certain measurements are often omitted or averaged over a large number of measurements, potentially excluding important data [23]. With regards to single cell mechanics, it is still not well known whether the models generated for cytoskeletal mechanics apply to the whole cell or to particular sub-cellular domains, as individual cells seem to exhibit more complex behavior than initially suggested [9]. In view of the importance of the study of single cell mechanics, we propose the use of a modified version of MTC for the measurement of the localized deformation of a single cell. An advantage over the traditional method of MTC is that we can study the twisting of a single bead attached to a single cell (as compared to an ensemble), with the angle of bead rotation detected by the SV sensor. The advantages of our proposed method is that twisting can be done regardless of whether the bead is attached to the surface or engulfed by the cell, and is not limited to the periphery of a cell (as in optical tweezers methods). Furthermore, it does not have the complication of cell heating as in laser detection methods [22, 24], and the angle of rotation of a bead twisting about the z axis can also be measured (which optical detection methods are not able to do). The only condition is that the bead has to be in the close vicinity of the SV sensor. 103 Chapter Magnetic Twisting Cytometry In our twisting experiments, µm ferromagnetic particles (Spherotech CFM-805) were functionalized with collagen, and attached to the surface of human Mesenchymal Stem Cells (hMSCs) grown on a SV sensor chip in a PDMS well. With an optical microscope, a single µm bead was attached to a cell in the vicinity of a SV sensor and the sensor connected to external circuitry. A brief magnetizing field was applied to the magnetic particle, followed by a twisting field of ~20 G for ~ minutes. The twisting of the magnetic particle was then measured as a voltage change across the SV sensor because, as shown in Chapter 3, the SV response depends on the direction of bead magnetization with respect to the SV. 5.3 FABRICATION (a) Spin valve sensor 12 x µm2 SV structures were fabricated in the same method as described in Section 2.3(b), consisting of the thin Ta3/NiFe2/IrMn8/CoFe2/Ru0.8/CoFe3/Cu2.3/CoFe2.6/Cu1/Ta5 film (with multilayer thickness in nanometer) and capped with a 40 nm Al2O3 passivation layer. One modification for the MTC work was that a polymer (~ 150 nm) was spin coated onto the surface and heated on a hotplate at 95 ºC for minute. The polymer layer covering the region of the contact pads was removed with acetone, and the entire chip sterilized by dipping in 70% ethanol and air-dried in a biological safety cabinet (BSC). This procedure was done in order to prepare the sensor surface for cell plating as well as to further protect the SV sensor from the cell culture media (see Fig. 5.4.1(a)). 104 Chapter Magnetic Twisting Cytometry (b) PDMS wells To culture human mesenchymal stem cells (hMSCs) on our spin valve sensor, hMSCs were plated in a polydimethylsiloxane well (PDMS, Dow Corning Sylgard 184) attached to the surface of a SV chip. To fabricate the well, PDMS was mixed (ratio 1:11) and spread to a thickness of ~ mm on a glass slide and cured in an oven at 120ºC for 10 minutes. The mm thick sheet of PDMS (Fig. 5.3.1(a)) was then cut into a rectangular well of outer dimension 7.5 mm x 10 mm, and inner dimension 5.5 mm x mm (Fig. 5.3.1(c)). A cover glass slide (#00, Menzel Glaser) of dimension 7.5 mm x 10 mm was used for capping of the PDMS well, to enclose the cells and culture media within a sterile environment during experiments. Both the PDMS well and cover glass slide were sterilized in 70% ethanol and then air-dried in a BSC. (b) (a) PDMS block (~2 mm thick) (c) Inner edges of PDMS well Outer edges of PDMS well Fig. 5.3.1 Fabrication of PDMS wells. (a) PDMS block of mm thickness. (b) Outer edges of well cut to size. (c) Inner edges of well cut to size. 105 Chapter 5.4 CELL CULTURE PROTOCOLS (a) Cell culture Magnetic Twisting Cytometry Human MSCs (Poietics; Cambrex, East Rutherford, NJ) were cultured in 5% CO2 at 37ºC in T25 cell culture flasks in cell culture media made up of Dulbecco’s modified eagle’s media (Gibco, #12491015) supplemented with 10% fetal calf serum (Gibco, 26140079), mM L-glutamine (Gibco, 25030081), 100 U/ml penicillin and 100 µg/ml streptomycin (Gibco, #15140122). Passage 6-8 hMSCs were used in our twisting experiments. A SV sensor chip with PDMS well (SV wells) was assembled in the biological safety cabinet by placing the PDMS well structure on the SV chip and applying a slight pressure to create a tight seal between the sensor chip and PDMS. The assembled SV well (Fig. 5.4.1(a)) was then placed in a Petri dish. To plate the SV well, confluent hMSCs in a T25 cell culture flask were first rinsed with phosphate buffer solution (PBS) and then lifted by adding ml of trypsinEDTA (Gibco #25200056) and placed in the incubator (Forma Scientific) for minutes. ml of cell culture media was next added to the flask to deactivate the trypsin and create the cell suspension. 60 µl of the cell suspension was pipetted into the SV well, with cell culture media used to fill the remainder of the well till an inverted meniscus was obtained (Fig. 5.4.1(c)). The seeded SV chip was then incubated in a covered Petri dish at 37ºC (5% CO2). 106 Chapter Magnetic Twisting Cytometry (a) Polymer coated SV chip (b) Polymer over contact pads removed (c) PDMS well SV sensor chip (d) Cell culture Media inside PDMS well SV contact hMSC Fig. 5.4.1 (a) Polymer coated SV sensor chip. (b) SV sensor chip with attached PDMS well. (c) Inverted meniscus of cell culture for cell plating of sensor chip. (d) Typical hMSCs cultured on SV sensor chip (with attached beads). (b) Bead functionalization µm functionalized ferromagnetic with particles methylated (SPHEROTM collagen [25] CFM-80-5) using EDC were surface (1-ethyl-3(-3- dimethylaminopropyl) carbomiimide hydrochloride, Sigma, E7750) according to the vendor’s particle coating procedures [26]. 0.5 mg of freeze-dried methylated collagen was dissolved in 1ml of DI water, with 2.5 mg of EDC, mL of sodium acetate buffer (0.01 M, pH 5.0) and ml of 1% w/v µm magnetic particles added to the solution. The bead suspension was vortexed at room temperature for hours and the magnetic beads subsequently isolated using a permanent magnet. The supernatant was removed and beads re-suspended in PBS (rinsing step), with the rinsing step repeated times to give a final solution of collagen functionalized µm ferromagnetic particles. 107 Chapter Magnetic Twisting Cytometry Magnet on Magnet off Magnet on Magnet off SV Voltage Signal (mV) 35.20 35.10 35.00 (b) (c) (d) 34.90 34.80 34.70 (a) (e) 34.60 510 520 530 540 550 560 570 580 590 600 Time (s) Fig. 5.7.1 Response of a SV sensor due to an applied electromagnetic field (no bead present). (a) Electromagnets turned on. (b) SV sensor response goes to a stable value. (c) Electromagnetic turned off. (d) SV sensor response returns to a base value. The SV response time over region (b) is less than ~ 150 ms. (b) Bead positioning Initial experiments were done by locating a single magnetic particle which had simply settled on a cell above the SV sensor (Section 5.5(a)), which was a very tedious process as the chance of finding a single magnetic particle on a cell in the vicinity of a SV sensor was extremely small. An observation made was that for beads initially added (i.e. without enough time to bind to cell surfaces), we were able to make them move over cell surfaces by applying an external rotating field (i.e. a rotating bar magnet). With this method, we could selectively position individual ferromagnetic particles at the center of a SV sensor and hence bind to the cell surface at this position. Figure 5.7.2 shows images 127 Chapter Magnetic Twisting Cytometry of a single magnetic particle ‘rolling’ to a SV sensor, with the corresponding SV signal shown in Figure 5.7.2(i). (a) (b) (c) (d) (e) (f) (g) (h) (b) SV Voltage Signal (mV) (i) 34.448 (d) (e) 34.446 (h) 34.444 34.442 34.440 34.438 874 875 876 877 878 879 Time (s) Fig. 5.7.2 (a)-(d) Images showing a single magnetic particle ‘rolling’ to a SV sensor by applying a rotating magnetic field. The oscillations in the SV voltage signal are due to the applied rotating field. (e)-(h) The magnetic field is removed and the magnetic particle drifts towards the center of the SV sensor and settles in position shown in (h). (i) Real time 128 Chapter Magnetic Twisting Cytometry response of the SV due to the single µm magnetic particle at the different positions indicted by (b), (d), (e) and (h). (c) Magnetic twisting After ~2 days of cell incubation, the SV chip was checked for cells spreading over the SV sensor (Fig. 5.7.3). A low concentration of collagen functionalized µm ferromagnetic particles was added to the SV well and capped with a cover glass slide. The chip was next connected to external circuitry as described in Section 5.5(d), and a single magnetic particle ‘rolled’ onto the cell in the vicinity of the SV sensor. The chip was then left in ambient for 20 minutes to allow for binding to occur after which the sensor chip was magnetized in a brief field of ~ 1600 G along the z axis of the SV chip (Fig. 6.5.4(b)) so as to fix the magnetization direction of the µm ferromagnetic particle. The current through the sensor was subsequently allowed to stabilize for ~ minutes before a twisting field was applied, an important step to reduce drift in the SV signal. SV Sensor active area hMSC SV contacts Fig. 5.7.3 Image of a 12 x µm2 SV sensor used in magnetic twisting experiments. 129 Chapter Magnetic Twisting Cytometry Figure 5.7.4(a)-(h) shows images of a single µm magnetic particle twisting on the surface of a single hMSC. The corresponding voltage response across the SV sensor is shown in figure 5.7.4(i). (a) (b) (c) (d) (e) (f) (g) (h) A SV Voltage Signal (mV) A A B B 34.7 (i) A B B 34.6 34.5 34.4 (l) (k) (m) A – Electromagnet On B – Electromagnet Off 34.3 34.2 (j) 34.1 1200 1400 1600 1800 2000 2200 2400 Time (s) Fig. 5.7.4 (a)-(d) Images of a single magnetic particle twisting towards the right during the application of an electromagnetic field at time indicated by (k), with the rotation from (a)-(d) occurring within a time frame of less 130 Chapter Magnetic Twisting Cytometry than second. (e)-(h) Images of the same magnetic particle returning to its original position when the electromagnet is turned off, at time indicated by (m). (i) Response of SV sensor to a twisting µm particle, with twisting cycle repeated four times. In figure 5.7.4, the SV sensor response is seen to be stable (time(j)) before the application of the electromagnetic “twisting” field at time (k) similar to our observations in figure 5.7.1(a)). Figure 5.7.5 shows the SV sensor response during the application of the twisting field, extracted and magnified from figure 5.7.4(i) between the regions A to B. (e) SV Voltage Signal (mV) 34.630 34.625 (i) (f) 34.620 (g) (h) 34.615 34.610 (a) (b) (c) 34.605 1400 1600 1800 2000 (d) 2200 2400 Time (s) Fig. 5.7.5 SV sensor response during the application of ‘twisting’ field. (a)-(d) Four magnetic twisting cycles. (e)-(h) The sensor voltage signal measured with the initial application of each twisting field. The SV signal decreases with time due to the twisting of the single µm ferromagnetic particle. 131 Chapter Magnetic Twisting Cytometry Figure 5.7.5 shows that for each twisting cycle, the SV response immediately increases to a value corresponding to the SV signal rapidly adjusting to the external applied field. This value then gradually decreases with time, corresponding to the rotation of the µm ferromagnetic particle above the sensor. The decrease in the SV signal for the first twisting cycle (duration ~ 140 s) was measured to be ~13.2 µV (Fig. 5.7.5(a)). An interesting observation was that for the onset of the second twisting cycle, the SV sensor signal (f) did not return to the original value (e), but instead started at some intermediate value between the two extremes (e) and (i) reached in the previous cycle. This trend was observed for subsequent twisting cycles with the gradient of the SV signal response slowly decreasing with time, which may indicate that the µm ferromagnetic particle is gradually reaching a stable position on the cell. Cycle (d) certainly appears to have reached an equilibrium. This is not unexpected behaviour. On the initial application of the twisting field the particle may move slightly (in x, y or z) in addition to twisting. The cell material around the bead may also deform in a time dependent manner, i.e. the deformation is initially plastic, at the start of the twisting cycle (a), and then appears to equilibrate towards an elastic response (d). The physical cause of such a response could be the removal of liquid from the cell material near the bead, or an actual stiffening of the cell as it responses to the applied external force. The very long time response is compelling evidence that within a simple spring-dashpot model we are not singly measuring the stiffness of the cell (i.e. a spring) but also a large viscoelastic component (i.e. a dashpot). The data of figure 5.7.5 was only obtained after many trial and error experiments. If the bead is too strongly attached or embedded in the cell, or too far from the SV 132 Chapter Magnetic Twisting Cytometry surface, then no signal will be measured. If the bead is too weakly attached then application of the twisting field may translate rather than rotate the bead, resulting in jumps in the SV sensor response. The SV sensor response due to a twisting µm bead is entirely dependent on the change of magnetization direction as well as the position of the bead, which can result in either an increase or decrease in the SV sensor signal. Figure 5.7.6 shows the voltage response across the SV sensor due to a loosely bound µm ferromagnetic bead. The sensor response is seen to be stable time (a) before the application of the “twisting” field at time (b) similar to our observations in figure 5.7.1(a). B A B A SV Voltage Signal (mV) 34.15 34.10 (b) (c) (d) 34.05 A – Electromagnet On B – Electromagnet Off 34.00 (a) 33.95 384 386 388 390 392 394 Time (s) Fig. 5.7.6 Response of SV sensor to a twisting µm particle, with twisting cycle repeated two times. Figure 5.7.7 shows the SV sensor response during the application of the twisting field, extracted and magnified from figure 5.7.6 between regions A to B. 133 Chapter Magnetic Twisting Cytometry 34.160 (e) SV Voltage Signal (mV) 34.158 (d) 34.156 34.154 34.152 (c) 34.150 (a) (b) 34.148 384 386 388 390 392 394 Time (s) Fig. 5.7.7 SV sensor response during the application of ‘twisting’ field. (a)-(b) Two magnetic twisting cycles. The SV signal increases with time due to the twisting of the single µm ferromagnetic particle. Figure 5.7.7 shows that for each twisting cycle, the SV response immediately increased to a value and gradually continues to increase (as compared to Fig. 5.7.5), corresponding to the rotation of the µm ferromagnetic particle above the sensor. There were instances where jumps in the SV response was observed (i.e. (c), (d) and (e)), which we attribute to a translation in the magnetic particle. This shows that the SV sensor response is dependent on the change in the magnetization direction as well as the position of the magnetic particle, which can result in the increase or decrease in the SV response. 134 Chapter 5.8 Magnetic Twisting Cytometry RESULTS: MODELING In order to correlate the change in the SV sensor signal to the angle of bead rotation, we first determined the position of the magnetic particle with respect to the SV sensor, and adopted the model as described in Section 5.6(d) to simulate the SV sensor response for a rotating bead at a specific height (z). In principle, the modeling procedure will yield accurate values of the twist cycle. However, for our particular experimental setup, two major problems were faced. The first was that the hMSC was extremely flat and transparent, making it difficult to accurately measure the thickness of the cell directly over the SV sensor. We estimated the cell thickness to be ~ 500 nm, as it was invisible when viewing through the 50x microscope objective. The second problem faced was that with the application of the twisting field. The particle rotated in the opposite direction expected. This we attribute to the fact that the magnetization was not done with a pulse magnetizer [17], but with a normal dipole magnet (duration ~ 3-5 s), which resulted in the magnetic particle rotating with the magnetizing field instead of fixing the magnetization direction of the particle. As a result, with the removal of the magnetizing field, the bead returned to its original magnetization direction. Based on the visual observations made during the twisting experiments, we can make an educated guess of the original magnetization direction of the magnetic particle. In the example of figure 5.7.5 the initial magnetization direction was assumed to be 115º with respect to the +y axis (Fig. 5.8.1 Case (115º)) and rotating clockwise towards the + y axis direction when the twisting field was applied. In our model, we attribute the slow varying signal change entirely to bead rotation and ignore any effects of bead translation. 135 Chapter Magnetic Twisting Cytometry This is a reasonable assumption as there was no visible translational movement of the bead during the twisting cycle after ~ second. Case (115º) y Case (90º) Case (65º) Case (43º) z hMSC m = m[−Cos (65°) y − Sin(65°) z ] m = −m z m = m[Cos(65°) y m = m[Cos (43°) y − Sin(65°) z ] − Sin (43°) z ] Fig. 5.8.1 Schematic of magnetization directions use to model the rotation of the particle used in Fig. 5.7.5. The magnetic particle is assumed to have an initial magnetization direction of ~115º with respect to the +y axis and then rotate in the +y direction when the twisting field is applied. Using a similar modeling method as described in 5.6(d) and taking the bead magnetization to be m ≈ 3.67 x 10-13 Am2, we are able to calculate the expected SV signal change when the µm ferromagnetic particle rotates (z = Al2O3+polymer+cell ≈ 0.69 µm). Figure 5.8.2 shows the calculated SV voltage signal with respect to bead position for the four magnetization directions considered in figure 5.8.1. 136 Chapter Magnetic Twisting Cytometry (a) (b) Calculated SV voltage signal (µV) 20 (c) (d) 10 (c) 10 20 30 40 50 (d) 90º -10 115º 65º -20 43º µm Fig. 5.8.2 Line scan profile showing the calculated sensor output as a function of bead position for the four different bead magnetization cases (115º)-(43º) as defined in Fig. 5.8.1. Representative line scans have been chosen across the SV sensor as indicated by the dotted line in (d). The dotted line as indicated in (c) represents the position of the magnetic particle with respect to the calculated sensor output. For the data of figure 5.7.5 the particle is at position (c) and we expect the sensor signal to shift from the red curve (115º) through to the black curve (43º). Taking cycle (d) in figure 5.7.5, which has reached an equilibrium, the change in SV signal is ~ 10 µV. This corresponds to an angle change of ~ (115º-65º) = 50º. Figure 5.8.2 shows that to achieve good sensitivity the bead must be positioned over the sensor between the signal maxima and minima. Based on our calculations and assuming the above conditions (cell thickness etc.), a µm ferromagnetic particle will generate a SV signal of ~ 0.2 µV per degree rotation. This is lower than the detection 137 Chapter Magnetic Twisting Cytometry resolution for single particle MTC using optical methods which can determine the bead position with an accuracy of ~ nm [16], corresponding to a rotation of ~ 0.1º. To get to this level we need an order of magnitude improvement in the signal to noise ratio, and this appears achievable given that we can have more gain, better magnet setup (different sets of magnets, and pulsed magnet), lock-in detection as well as on-chip electronics. Thus we believe a sensitivity of 0.1º is achievable and hence comparable to MTC using optical detection methods [16]. Together with the proper calibration for the values of the applied stresses acting on a free rotating particle, the Young’s modulus of the hMSC can be calculated [31]. We further believe that with the proper tools for the magnetization of the ferromagnetic particles, and the proper optical system (phase contrast or confocal microscope) to measure the cell thickness, we will be able to accurately measure the angle of rotation of a single ferromagnetic particle attached to a single cell. 5.9 CONCLUSION AND FURTHER WORK We have demonstrated a proof of principle of a modified method of MTC able to measure the angle of twist in a single µm ferromagnetic particle attached to the surface of a hMSC, as well as an effective way of positioning a single magnetic particle on a cell. The advantage of this technique is that we are able to measure the rotation of a single magnetic particle attached to the surface of a single cell as compared to traditional methods of MTC where ~ 40,000 particles are used to twist an ensemble of cells. Our method will be able to provide a better insight into single cell mechanics which may have been overlooked in traditional MTC. 138 Chapter Magnetic Twisting Cytometry In comparison to MTC using optical detection techniques [15, 16] with similar ability to measure the angle of twist of a single bead attached to a single cell, our technique has the added advantage of being able to measure the angle of bead rotation about the z axis. In addition, we are able to measure the small gradual change in the SV response with respect to time (Fig. 5.7.5), which may provide information regarding the time dependent relationship between the stiffening of the cell surfaces with respect to applied stress. A disadvantage of our system is that our technique will not work for cells above a certain thickness, as the fields generated by the magnetic particle will be too small for the SV to detect. Nevertheless, one can envisage several approaches which can be adopted which would significantly improve and extend the applicability of our method. Subsequent work will be to incorporate of the use of lock-in amplifiers, apply appropriate gain factors, and use particles of higher magnetic content so as to achieve higher SV detection sensitivity. Separate magnetics for the SV sensor bias and twisting fields can be used, with the ability to apply twist fields in different directions. A pulsed magnetization of the bead is also desirable. Finally combining the SV technique with suitable optical methods would enable the height (z) of the bead to be found and also enable integration with the optical MTC technique, creating a powerful instrument platform for single cell characterization. 139 Chapter Magnetic Twisting Cytometry REFERENCES [1] G. Bao, S. Suresh, Nat. Mat. 2,715 (2003) [2] B. Alberts et. al. Molecular Biology of the Cell, 4th Ed. (Garland, New York, 2002) [3] R. F. Kamm, M. R. K. Mofrad, Cytoskeletal Mechanics. (Cambridge, 2006) [4] D.E. Ingber, D. Prusty, Z. Sun, H. Betensky, N. Wang, J. Biomech. 28, 1471 (1995) [5] T. P. Stossel, Science 260, 1086 (1993) [6] P. F. Davies, Arterioscler Thromb. Vasc. Biol. 22 (11) 1755 (2002) [7] T. P. Lele, C. K. Thodeti, D. E. Ingber, Journ. Cell. Bio. 97, 1175 (2006) [8] N. Wang, J. P. Butler, D. E. Ingber, Science 260, 1124 (1993) [9] D. R. Overby, B. D. Matthews, E. Alsberg, D. E. Ingber, Acta Biomaterialia 1, 295 (2005) [10] M. J. Rosenbluth, W. A. Lam, D. A. Fletcher, Biophy. J. 90, 2994 (2006) [11] B. Daily, E. L. Elson, G. I. Zahalak, Biophys. J. 45(4), 671 (1984) [12] C. T. Lim, E. H. Zhou, A. Li, S. R. K. Vedula, H. X. Fu, Mat. Sci. Eng. C 26(8), 1278 (2006) [13] S. Usami, H. H. Chen, Y. Zhao, S. Chien, R. Skalak, Ann. Biomed. Eng. 21, 77 (1993) [14] C. Storm, J. J. Pastore, F. C. Mackintosh, T. C. Lubensky, P A. Janmey, Nature 435, 191 (2005) [15] H. Karcher, J. Lammerding, H. Huang, R. T. Lee, R. D. Kamm, M. R. K. Mofrad, Biophys. J. 85, 3336 (2003) [16] B. Fabry, G. N. Maksym, J. P. Butler, M. Glogauer, D. Navajas, N. A. Taback, E. J. 140 Chapter Magnetic Twisting Cytometry Millet, J. J. Fredberg, Phys. Rev. E 68, 041914 (2003) [17] N. Wang, K. Naruse, D. Stamenovic, J. J. Fredberg, S. B. Majailovich, I. M. TolicNorrelykke, T. Polte, R. Mannix, D. E. Ingber, Proc. Natl. Acad. Sci. 98(14), 7765 (2001) [18] H. Freundlich, W. Seifriz, Phys. Chem. 104, 233 (1922) [19] A. R. Bausch, W. Moller, E. Sackman, Biophys. J. 76(1), 573 (1999) [20] M Keller, J. Schilling, E Sackman, Rev. Sci. Instr. 72(9), 3626 (2001) [21] P. A. Valberg, J. P. Butler, Biophys. J. 52(4) 537 (1987) [22] V. M. Laurent, S. Henon, E. Planus, R. Fodil, M. Balland, D. Isabey, F. Gallet, Journ. Biomed. Eng. 124(4), 408 (2002) [23] F. J. Alenghar, B. Fabry, K. Y. Tsai, W. H. Goldmann, D. E. Ingber, Biochem. Biophys. Res. Comm. 277, 93 (2000) [24] B. D. Hoffman, G. Massiera, K. M. Van Citters, J. C. Crocker, PNAS 103(27), 10259 (2006) [25] S. Z. Yow, C. H. Quek, E. K. F. Yim, C. T. Lim, K. W. Leong, Biomat. 30(6), 1133 (2009) [26] SPHEROTM Technical Note – Particle Coating Procedures. STN-1 Rev C. 041106 [27] D. K. Wood, K. K. Ni, D. R. Schmidt, and A. N. Cleland, Sens. Actuators, A 120, (2005) [28] P. P. Freitas, H. A. Ferreira, D. L. Graham, L. A. Clarke, M. D. Amaral, V. Martins, L. Fonseca, and J. M. S. Cabral, in Magnetoelectronics, edited by M. Johnson (Academic, New York, 2004) [29] M. Tondra, M. Porter, and R. J. Lipert, J. Vac. Sci. Technol. A 18, 1125, (2000). 141 Chapter Magnetic Twisting Cytometry [30] H. A. Ferreira, D. L. Graham, P. P. Freitas, J. M. S. Cabral, J. Appl. Phys. 93, 7281 (2003) [31] K.S. Zaner, P. A. Valberg, J. Cell. Bio. 109, 2233 (1989) 142 [...]... Chapter 5 Magnetic Twisting Cytometry Case (0º) (a) 50 40 30 20 10 0 0 10 20 30 40 50 Case (22º) (b) 50 40 30 20 10 0 0 10 20 30 40 50 (f) 50 40 30 20 10 0 0 10 20 30 40 50 (k) 15 10 5 10 20 30 40 50 50 40 30 20 10 0 0 10 20 30 40 50 µm 50 40 30 20 10 0 0 10 20 30 40 50 -5 50 40 30 20 10 0 0 10 20 30 40 50 µm 50 40 30 20 10 0 0 10 20 30 40 50 (j) 50 40 30 20 10 0 0 10 20 30 40 50 (o) (n) 20 10 10 5 10... (j) 50 40 30 50 40 30 50 40 30 50 40 30 50 40 30 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 (k) (l) 10 6 4 2 -2 10 5 5 10 20 30 40 50 µm -5 (m) 10 20 30 40 50 µm -5 10 5 10 5 10 20 30 40 50 µm (o) (n) -5 10 20 30 40 50 µm -5 -10 10 20 30 40 50 µm Fig 5. 6.9 Comparison of calculated voltage change (a)-(e) and experimental... proof of principle of the SV based magnetic twisting cytometry for single bead detection 126 Chapter 5 Magnetic Twisting Cytometry Magnet on Magnet off Magnet on Magnet off SV Voltage Signal (mV) 35. 20 35. 10 35. 00 (b) (c) (d) 34.90 34.80 34.70 (a) (e) 34.60 51 0 52 0 53 0 54 0 55 0 56 0 57 0 58 0 59 0 600 Time (s) Fig 5. 7.1 Response of a SV sensor due to an applied electromagnetic field (no bead present) (a)... shown in figure 5. 6.9(f)-(h) is set such that it matches that of the calculated voltage change Case (0º) (a) Case (22º) (b) Case (43º) (c) Case ( 65 ) (d) Case (90º) (e) 50 40 30 50 40 30 50 40 30 50 40 30 50 40 30 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 20 10 0 0 10 20 30 40 50 (g) (f) (h) (i) (j) 50 40 30 50 40 30 50 40 30 50 40 30 50 40 30 20... 10 10 5 10 20 30 40 50 Case (90º) (e) (i) (m) 10 5 -10 10 20 30 40 50 -20 Case ( 65 ) (d) (h) (l) 20 10 µm 50 40 30 20 10 0 0 10 20 30 40 50 (g) 50 40 30 20 10 0 0 10 20 30 40 50 -5 Case (43º) (c) -5 10 20 30 40 50 -10 -10 -20 µm 10 20 30 40 50 µm Fig 5. 6.6 Comparison of model voltage change (a)-(e) and experimental voltage change (f)-(j) on a 12 x 8 µm2 SV sensor due to a 8 µm ferromagnetic bead for... sensor after processing was ~2 .5% 119 Chapter 5 Magnetic Twisting Cytometry 37.2 Resistance (Ω) 37.0 36.8 36.6 36.4 36.2 36.0 -200 - 150 -100 -50 0 50 100 150 200 H Field (Oe) Fig 5. 6 .5 Transfer characteristic of a SV sensor (12 x 8 µm2 active area) to an external applied field applied along the y-axis direction Magnetoresistance = 2 .5 % For the experimental data shown in figure 5. 6.3, the bead was always... Chapter 5 Magnetic Twisting Cytometry hMSC 8 µm ferromagnetic particles (a) (b) Fig 5. 5.2 Optical images showing 8 µm magnetic particles attached to hMSCs (a) Magnetic particles attached to cells, with the microscope focused on the beads (b) Similar image with the microscope focused on the cell surface The position of the cell is indicated by the arrow (c) Bead manipulation To manipulate a single magnetic. .. CD-ROM] 110 Chapter 5 Magnetic Twisting Cytometry Edge of hMSC (a) (b) (c) (e) (f) hMSC (d) Fig 5. 5.3 Images showing the manipulation of a magnetic particle attached to a single cell (a) A single 8 µm particle attached to the edge of a cell (b) Zoomed in imaged of the single 8 µm ferromagnetic particle Red line indicates the rest position of the bead with no applied magnetic field (c) The magnetic field... points with higher SV voltage output Case (90º) Case ( 65 ) Case (43º) Case (22º) Case (0º) Fig 5. 6.8 SV voltage scans of a single 8 µm ferromagnetic bead over a 12 x 8 µm2 SV sensor at height z ~ 5. 99 µm for the five different magnetization cases (90º)-(0º) as defined in Fig 5. 6.2 Scan size: 56 x 56 µm2 123 Chapter 5 Magnetic Twisting Cytometry Figure 5. 6.9 show the 2D density plots for the calculated... magnetization cases considered (see Fig 5. 6.2); Case (90º): m = m z ; By = μ0 − 3myz [ ], 4π ( x 2 + y 2 + z 2 ) 5 2 (5. 2) Case ( 65 ): m = m[−Cos( 65 ) y + Sin( 65 ) z ] ; μ 0 2.718myz + 0.422m( x 2 − 2 y 2 + z 2 ) [ ], By = 5 4π (x 2 + y 2 + z 2 ) 2 (5. 3) Case (43º): m = m[−Cos(43°) y + Sin(43°) z ] ; By = μ 0 + 2.046myz + 0.731m( x 2 − 2 y 2 + z 2 ) [ ], 5 2 2 2 4π 2 (x + y + z ) (5. 4) Case (22º): m = m[−Cos(22°) . 109 Chapter 5 Magnetic Twisting Cytometry (a) 8 µm ferromagnetic particles (b) hMSC Fig. 5. 5.2 Optical images showing 8 µm magnetic particles attached to hMSCs. (a) Magnetic particles. give a final solution of collagen functionalized 8 µm ferromagnetic particles. 107 Chapter 5 Magnetic Twisting Cytometry 5. 5 EXPERIMENTAL SETUP (a) Cell-bead preparation Prior to reaching. Magnetometer ~1000G Magnetic particle pulse Substrate 0- 25 G twist Cell (1 minute) H 102 Chapter 5 Magnetic Twisting Cytometry ability of measuring bead displacement with an accuracy of 5 nm [16].