We provide an evaluation for an electrically tunable lens (ETL), combined with a microscope system, from the viewpoint of tracking intracellular protein complexes. We measured the correlation between the quantitative axial focus shift and the control current for ETL, and determined the stabilization time for refocusing to evaluate the electrical focusing behaviour of our system. We also confirmed that the change of relative magnification by the lens and associated resolution does not influence the ability to find intracellular targets. By applying the ETL system to observe intracellular structures and protein complexes, we confirmed that this system can obtain 10 nm order z-stacks, within video rate, while maintaining the quality of images and that this system has sufficient optical performance to detect the molecules.

Advances in quantifying the dynamics of intracellular molecules have been realised through the development of live-cell imaging on a molecular level and algorithms required for tracking their behaviours.1,2 As a result, the parameters to describe molecular dynamics such as velocity, length, and lifetime are known to play an important role in revealing the molecular mechanisms that underpin dynamic movement.3,4 Most of these analyses had targeted two-dimensional image sequences; for example, tracking of transmembrane receptors, or microtubules in the peripheral region of interphase cells.3,5 For tracking molecules that are restricted in their movement within a 2-dimensional area, we do not need to speculate about the molecular movement towards the third dimension (z-axis). Consequently, 2-dimensional imaging is sufficient for the analysis of the behaviours of several intracellular molecules.

However, if we target for the molecular behaviours in mitotic cells, two-dimensional data are insufficient, as the rounded shape of the mitotic cells allows 3-dimensional movement. Even in the case of interphase cells, if our target molecules are observed near the nuclei, the thickness of the area is supposed to be more than 6–7 μm6 because of the existence of nuclei. This is a large space for a freely diffusing protein complex, which has a 3–500 nm diameter.7,8 In particular, observing 3-dimensional dynamics of the molecules is crucial for revealing the mechanisms of cell division, which has been a long-standing unresolved goal of biological and pathological research.9,10

One of the main challenges to achieve 3-dimensional analysis of intracellular molecules is that it requires a fast transition of the focal plane at a sufficient spatial and temporal resolution. Because these molecules move fast, about 250 nm/s,11 the transition of the focal plane must be achieved within a few milliseconds. At the same time, to make the voxel size even, spatial resolution must be about tens of nanometers. This harsh condition cannot be achieved even by a piezoelectric z-focusing device, which is the conventional way to obtain a three dimensional imaging sequence at high temporal resolution. The strategy of moving an objective revolver or a stage cannot avoid the effect of inertia; as a result, this effect restricts its focusing speed. In addition, the oscillation of samples or the objective lens, induced by the mechanical movement, makes it difficult to guarantee that there is no effect on the dynamics of molecules on a nano-scale.

One solution to overcome these limitations is using a focus tunable lens. The advantage of using a focus tunable lens is that it enables rapid focusing without the mechanical movement of a microscopic system. At present, many technologies are employed to take advantage of fast focus tuning without mechanical movement, such as electrowetting, liquid crystals, electroactive polymers (EAPs), servomotors, MEMS micropumps, and piezoelectric devices.12 

Here, we provide a potential solution to the 3-dimensional analysis of intracellular proteins with a high spatiotemporal resolution, by using a commercially available electrically tunable lens(ETL), EL-10-30 (Optotune, Switzerland). Among other studies that have already applied the ETL to various microscopic systems,13–15 this is the first report evaluating an ETL combined with a microscope system from the view point of 3-dimensional analyses of dynamic intracellular protein complexes. We attached the ETL to a fluorescence microscope and evaluated the optical properties of the assembly. We succeeded in acquiring reliable three-dimensional imaging sequences of intracellular structures with a high spatiotemporal resolution. We further confirmed that this system is applicable for detecting protein complexes.

We used an Olympus ix81, controlled with Micromanager16,17 for the evaluation of optical properties and the acquisition of intracellular images (Secs. III and IV). All images using ix81 were obtained by using 100 × oil-immersion microscope objective (Olympus UPlanSApo, 100 × Oil, NA 1.40), with ORCA-R2 (Hamamatsu Photonics). Olympus ix71 equipped DeltaVision Core system was used for the imaging of cell lines (Sec. IV) by using 100 × oil-immersion microscope objective (Olympus UPlanApo, 100 × Oil, NA 1.35). The images were acquired using a CoolSnap HQ Camera (Photometrics). We used originally modified Olympus FV1000 into ezDSLM18 to measure the settling time of the ETL. The images at ezDSLM were acquired using an ORCA-Flash4.0 V2 Camera (Hamamatsu Photonics) with a 20 × water immersion objective (Olympus UMPlanFLN, 20 × Water, NA 0.5) for excitation (Ex) of the fluorescent media and 40 × water immersion objective (Nikon CFI Apo NIR, 40 × Water, NA 0.8) for the emission.

We combined convex ETL (EL-10-30-VIS-LD, Optotune AG, Switzerland) and a concave offset lens (OL) to generate a divergent or convergent light at the rear stop of the objective. ETL and the offset lens (ETL/OL) were mounted in a custom-built holder to which the microscope objective was attached. In order to assemble ETL/OL with the microscope, we originally prepared an adapter which can house the EL-10-30 (ETL) and an OL with defined distance and can attach the lens to the revolver and objectives which are a unified specification by Olympus. The holder was designed as the axial distances between the ETL and the plane surface of the offset lens would be 2.4 mm and 4.5–5.0 mm between the vertex of the concave surface of the offset lens and the mounting shoulder of the microscope objective. We used an offset lens -100 mm focal length, vis-nir coated plano-concave lens, 25 mm diameter (45-924-INK; Edmund TECHSPEC UK) for the evaluation of EL-10-30, and f = -100.0 mm, 1/2 UV Fused Silica Plano-Concave Lens, Uncoated (LC4232; Thorlab, US) for cell imaging. Figure 1(a) shows the design of the adapter of the ETL/OL to mount them on the lens revolver, and 1(b) shows the actual view of the assembly on Olympus ix81 microscope.

FIG. 1.

(a) The design of microscope adapter to mount and alignment of the ETL/OL assembly with respect to the microscope objective and excitation/detection pathways. The photograph in (b) shows the assembled whole system of the adapter and an objective lens at revolver.

FIG. 1.

(a) The design of microscope adapter to mount and alignment of the ETL/OL assembly with respect to the microscope objective and excitation/detection pathways. The photograph in (b) shows the assembled whole system of the adapter and an objective lens at revolver.

Close modal

The magnetic actuator of the ETL was controlled by Lens Driver 4, which delivered a stable current output from 0 to 298 mA with a precision of 12 bits, at a maximum of 5V, provided by Optotune (Switzerland). Our originally implemented software for the controller is written in C programming language and works on Unix based operating systems, for example, Mac OS X, FreeBSD, and Linux. All the source codes of the lens controller are available from the following URL (http://fun.bio.keio.ac.jp/software).

We estimated the axial range and optical resolution properties of the ETL-microscope objective combination by building simulation models including a 100 × objective using ZEMAX, the optical deign software (ZEMAX Development Corporation, Redmond, WA, USA). Prescription reference data for UPlanSApo, 100 × Oil, NA 1.40 were obtained from the patent literature.19 

Figure 2 shows the correlation between ETL control current mA and axial focus shift μm (a), relative magnification (b), and NA values simulated with ZEMAX (c). Simulated axial focus shift and relative magnification are fitted to equations z = 4.917 × 10−6i3 − 0.0004i2 + 0.2491i − 8.9443 with R2 = 0.999 and y = − 0.002i + 0.9967 with R2 = 0.997, respectively (z: axial focus shift μm, y: relative magnification, and i: current value mA). We used these simulation results to compare against the following experimental results in order to estimate if our setup of ETL is theoretically reasonable.

FIG. 2.

The graph (a) shows the simulated behavior on axial focus shift, (b) shows the relative magnification of the ETL/OL system in combination with the 100 × objective with ZEMAX. (c) The drawings show 2D ray tracing layout of the ETL/OL assembly with the microscope objective (OBJ) attached. The upper lane values indicate the focal position. The change in NA with axial focus shift was calculated with ZEMAX.

FIG. 2.

The graph (a) shows the simulated behavior on axial focus shift, (b) shows the relative magnification of the ETL/OL system in combination with the 100 × objective with ZEMAX. (c) The drawings show 2D ray tracing layout of the ETL/OL assembly with the microscope objective (OBJ) attached. The upper lane values indicate the focal position. The change in NA with axial focus shift was calculated with ZEMAX.

Close modal

We analysed axial focus shift by refocusing 100 nm fluorescent beads using a motorized z-stage with a built-in encoder of Olympus ix81, monitored by Micromanager. The experimental results showed in Figure 3(a) illustrate a characteristics of the dependency of axial focus shift on the control current of this assembled system. Zero focal position was set to the position when input current was zero. The focal position as a function of control current can be fitted to a polynomial equation z = 0.0007i2 + 0.1237i + 0.8545 with R2 = 0.9985, where z indicates the axial distance (μm) from the zero position and i is the input current (mA). We can control the focal position freely by using this equation. When we change the current in the finest way (0.07155 mA step), resulting focus shift was calculated to be 8.85 nm in minimum and 38.2 nm in maximum, which satisfies the required spatial resolution for detecting protein complexes in vivo. The results of the simulation and experiments are reasonably consistent (Figures 2(a) and 3(b)).

FIG. 3.

(a) The graph shows the electrical focusing behavior of the ETL/OL system in combination with the 100 × objective. (b) Stabilization time of the ETL was analysed with a light sheet microscope set up using a side-viewing microscope focused at the excited fluorescent spot in a cuvette containing a fluorescent solution. The images show the optical oscilloscope traces. Each condition of control current is in the table. We tested the effect of the current change to the focus plane stabilization from 0 to 17.9, 0 to 35.8, 0 to 64.4, and 71.5 to 114 mA, or the inverse. * These data were not fitted to Gaussian density function. We explored the maximum, 2nd and 3rd maximum values from the plot data list of each stacked image, and defined the stabilized timing for this case. (c) The graph shows relative change of the magnification with axial focus shift with respect to the magnification without ETL.

FIG. 3.

(a) The graph shows the electrical focusing behavior of the ETL/OL system in combination with the 100 × objective. (b) Stabilization time of the ETL was analysed with a light sheet microscope set up using a side-viewing microscope focused at the excited fluorescent spot in a cuvette containing a fluorescent solution. The images show the optical oscilloscope traces. Each condition of control current is in the table. We tested the effect of the current change to the focus plane stabilization from 0 to 17.9, 0 to 35.8, 0 to 64.4, and 71.5 to 114 mA, or the inverse. * These data were not fitted to Gaussian density function. We explored the maximum, 2nd and 3rd maximum values from the plot data list of each stacked image, and defined the stabilized timing for this case. (c) The graph shows relative change of the magnification with axial focus shift with respect to the magnification without ETL.

Close modal

To measure the stabilization time of the ETL, we used a light sheet microscope, which is equipped with an objective lens to view the focal volume from the side in a fluorescent medium, thus allowing us to visualize directly the focal spot of z-position over time and integrated the z-t trace (Figure 3(b)). Stabilization time was calculated by the following two steps; first, we count the slice number of the images between which the focal volume started to move and the first image at which the focal volume starts to stay within the defined range as described below. Next, the duration for taking one image (312 μsec) was multiplied by the slice number between the start of movement and the end. To define the position of the focal volume and its fluctuation range, we fitted the center line of each focal volume to Gaussian density function and defined the peak position as the focal position. We assumed that the focal position started to move when the sequential three positions were outside of the range of ±1% of the position at time 0, and that the focal position was stabilized when the sequential ten positions remained within ±1% of the position of the previous frame. The z-t traces showed fast initial oscillations with wide amplitude (Figure 3(b)). The observed oscillations were stabilized within 10 ms, for all cases varied at the start and end current values.

After confirming the mechanical performance of our ETL/OL assembly regarding its focus shift control and stabilization time, we next set to estimate whether we can obtain reliable images of intracellular structures and protein complexes using this system without compromising relative magnification and resolution.

ZEMAX simulation predicted that the magnification will change gradually with the transition of a focal plane. The reason of this change is described as follows. An objective lens is equipped inside an aperture stop between the convex 1st lens group and the concave second lens group. Attachment of the ETL/OL assembly to the rear stop of the objective is equivalent to equipping an additional lens in the second lens group; namely, to change the focal length of ETL means to change the focal length of the second lens group. Therefore, changing ETL curvature leads to the conversion of imaging size dependent on the characteristics of the other lenses. We confirmed the relative magnification of corresponding axial focus position. We observed Neubauer improved hemocytometer to quantitate the change of field of view (FOV) size with axial focus shift. We varied the control current from 0 to 286 mA by each 28 mA and compared with the size of control, which the ETL was not equipped in the microscope (Figure 3(c)). The relative magnification as a function of control current can be fitted to a linear equation y = − 0.016i + 1.0364 with R2 = 0.999, where y is the relative magnification and i is the control current. The results of the simulation and experiments are reasonably consistent (Figures 2(b) and 3(c)). The maximum thickness of a mammalian cell is approximately 10 μm as we had shown in Sec. I. When shifting the focal plane 10 μm, the relative magnification may change 8%, that is, 5 nm change in pixel, which can be ignored.

Our simulation predicted the change of NA value dependent on the curvature of ETL (Figure 2(c)). This prediction let us to estimate the further effects on the optical resolution of total ETL/OL assembly. Now, we evaluated the optical performance of this ETL/OL assembly by analysing point spread function (PSF), and estimated if our ETL/OL assembly is applicable for further intracellular imaging. Fluorescently labeled polystyrene microsphere, 100 nm in diameter (Fluoresbrite, Polyscience, Inc.) was used as point source. Data analysis was performed with MetroloJ,20 which is a plug-in software by Fiji.21 Lateral resolution was defined as FWHM of the PSF. Our application was tested with a conventional fluorescent (wide-field) microscope; therefore, we defined axial resolution as a depth of field. We set a threshold at 80% of the maximum intensity of the diffraction pattern and defined the width as the depth of field.

To observe the PSF quality over focusing range, we measured the changes in resolution for a given control current. Changing the control current causes the change of the magnification, which affects the pixel size, so we calculated the pixel size in each current to measure accurate resolution. Figures 4(a) and 4(b) show the lateral resolution (a) and axial resolution (b) dependent on the control current mA.

FIG. 4.

PSF dependent on the control current. (a) Lateral resolution dependent on the ETL control current. (b) Axial resolution dependent on the ETL control current. (c) Lateral and axial resolution within the limited range of the control current (20–75 mA) which seems to be the best range of the resolution. (d) The measured lateral and axial resolution without ETL/OL assembly (control).

FIG. 4.

PSF dependent on the control current. (a) Lateral resolution dependent on the ETL control current. (b) Axial resolution dependent on the ETL control current. (c) Lateral and axial resolution within the limited range of the control current (20–75 mA) which seems to be the best range of the resolution. (d) The measured lateral and axial resolution without ETL/OL assembly (control).

Close modal

The observed changes in resolution are associated with the focusing-induced variation of the excitation NA (Figure 2(c)). The results showed that the best resolution both lateral and axial, which are close to the resolution without ETL (Figure 4(d)) is found in the current range of 0–100 mA. The detail behavior of the resolution of the range is shown in Figure 4(c). Although lateral resolution is almost constant over the range, axial resolution got worse in the range of 20-45 mA.

From these results, we concluded that to acquire a three-dimensional image sequence of our biological target, using the current between 50 and 75 mA would be appropriate.

We next applied our ETL/OL assembly to observe tubulin stained HeLa cells in order to validate whether we can observe intracellular structures at high spatiotemporal resolution. HeLa cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal calf serum (FCS), Penicillin, and Streptomycin for all the following experiments. Cells were plated onto plastic dishes for passages, and 0.15 mm thickness, glass bottomed 4 wells dishes (Nunc, LabTek) for microscopy observations. We stained HeLa cells with Oregon Green 488 paclitaxel (Invitrogen, OR, USA).

Briefly, 1 mM concentration of the stock solution was prepared with dimethyl sulfoxide (DMSO), aliquot, and stored at 20 °C freezer. Final concentration for culturing cells was 1 μM with culturing medium for HeLa cells (working solution). Cells were incubated at 37 °C for 30 min with working solution and washed before observation with 37 °C Opti MEM (Life Technologies, Tokyo, Japan). Fluorescence excitation of Oregon Green was performed with 490 nm, and Wide band Interference Blue Filter (WIB) was chosen for emission.

Figure 5 shows the result of axial scanning within a cell by using our ETL/OL assembly. We performed z-stack imaging by each 0.07 mA in the range of 28.6 mA–100 mA. 0.07 mA difference of current corresponds to 10–20 nm change of axial focus shift. The exact value of the focus shift at each current is described in Figure 3(a). Cells of 15.4 μm thickness were imaged in total. The figure shows the clear focus at intracellular structure at 70.0 mA and its surrounding areas. We can confirm the fiber of the microtubules and rounded shape of nuclei where microtubules do not exist. This result showed that we could obtain clear images of intracellular structure with fine resolution to achieve the same spatial resolution with x-y axes (66 nm) along with video rate. This specification is suitable for 4D tracking of plusTip comets.

FIG. 5.

Observation of microtubules in HeLa cells by using a focus tunable lens. The control current had been changed from 28.6 mA to 100 mA at the finest step. The correspondence depth-of-focus was 4.8–20.2 μm (diff = 15.4 μm). Images were taken at 26.7 fps. Bar = 5 μm.

FIG. 5.

Observation of microtubules in HeLa cells by using a focus tunable lens. The control current had been changed from 28.6 mA to 100 mA at the finest step. The correspondence depth-of-focus was 4.8–20.2 μm (diff = 15.4 μm). Images were taken at 26.7 fps. Bar = 5 μm.

Close modal

Although ETL and OL have high transmission range of about 95%, there is a 5% degression in total transmission, which worsens the S/N ratio (SNR). We evaluated whether the SNR of the images through ETL/OL is high enough to be able to detect intracellular molecules.

We used fluorescent microtubule-end associated plusTip protein complexes in this evaluation.22,23 We transfected commercially obtained mKate2-EB3 encoding plasmid vector from Evrogen(Russia) into HeLa cells and compared the ability to detect mKate2-EB3, that appear as comets in mitotic cells with and without ETL/OL (Figures 6(a) and 6(b)). We shifted the focal plane in nm step by moving the stage (a) and in mA step (b), and obtained the images every 5 s. We used the current from 0 to 146 mA and chose the focus plane at 54.6, 72.8, 91.0, 109.2, 127.4, and 145.6 mA. Each current step took 50 ms during this imaging. To confirm the presence of true comets among false signals, we performed comet detection by both manual and automatic methods (plusTip tracker24) and considered those comets that were recognized by both methods. We calculated the ratio of those confirmed comets among all signals detected by manual and automatic methods. We used the ratio as an indicator of how precisely we may find the targeted protein complexes independent from the finding strategy.

FIG. 6.

Percentage of confirmed comet by manual and automated detection. (a) The case without ETL. The top panels show the manually detected comets (red dots), the bottom panels show the automatically detected comets by cyan dots. The middle panels show the raw data without the labels for the detected comets. Bar = 10 μm. The time frame which the control current reached at 54.6 mA was set as time 0 s in the figure. (b) The case with ETL. The top panels show the manually detected comets (red dots), the bottom panels show the automatically detected comets by plusTip tracker (cyan dots). The middle panels show the raw data without the labels for the detected comets. Bar = 10 μm. (c) The table shows the exact numbers of detected comets by manual or automatic method, and the ratio of confirmed comets by both manual and automatic methods among the sum of comet numbers (100 × (2× confirmed comets) / (manually detected comets + automatically detected comets)). The box plots show the minimum, 1st percentile (Q1), median, 3rd percentile (Q3), maximum or a point beyond 1.5 × (Q3Q1) of the ratio.

FIG. 6.

Percentage of confirmed comet by manual and automated detection. (a) The case without ETL. The top panels show the manually detected comets (red dots), the bottom panels show the automatically detected comets by cyan dots. The middle panels show the raw data without the labels for the detected comets. Bar = 10 μm. The time frame which the control current reached at 54.6 mA was set as time 0 s in the figure. (b) The case with ETL. The top panels show the manually detected comets (red dots), the bottom panels show the automatically detected comets by plusTip tracker (cyan dots). The middle panels show the raw data without the labels for the detected comets. Bar = 10 μm. (c) The table shows the exact numbers of detected comets by manual or automatic method, and the ratio of confirmed comets by both manual and automatic methods among the sum of comet numbers (100 × (2× confirmed comets) / (manually detected comets + automatically detected comets)). The box plots show the minimum, 1st percentile (Q1), median, 3rd percentile (Q3), maximum or a point beyond 1.5 × (Q3Q1) of the ratio.

Close modal

Detailed results are shown in Figure 6(c). Between the current of 54.6 mA and 145.6 mA, the system with ETL/OL could exhibit more than 40% confirmed comets and showed better confirmation than the system without ETL/OL. The total result is shown in the box plot (Figure 6(c)). We cannot say which case is better for detecting the comets because the observed cells were different, but we can conclude that the result with ETL/OL is not worse than the result without ETL/OL. These results suggest that when we choose the appropriate range of control current, the ETL/OL assembly can maintain suitable SNR to allow the precise detection of protein complexes, and that the quality is robust independent from the strategy of detection.

We further confirmed that comets could be detected from the images through ETL/OL by observing the overlapping of plus end complex and small kinetochore associated protein (SKAP) (an end binding (EB)-protein interactor). We transfected human cDNA fragments encoding for SKAP (NM033286.2), which was subcloned into pEGFPC1 (Clontech) vectors, also commercially obtained mKate2-EB3. Stable isogenic HeLaGFPSKAP was established as described in the previous work, and tetracycline was used for inducing expression.25 EB3 and SKAP are known to exist in the same protein complex.26 We observed EB3 and SKAP almost simultaneously by changing the excitation wavelength.

Figure 7 shows the actual view of the cells at every 5 s with ETL/OL. We detected the comet by using plusTipTracker and calculated the distance between red (EB3) and green (SKAP) comets by using the nearest neighbor algorithm. Because the size of the comets could be 50–500 nm, when these two proteins are in the same comets, they may be found within 500 nm. Therefore, if the distance of the comets was within 500 nm, we defined these comets as ”overlapping comets.” As a result of calculation, we could find 140 overlapping comets. This methodology increases the reliability of comet detection.

FIG. 7.

Lens evaluation with a biological application. Observation of red and green comets overlapping with ETL-10-30 of a cell line, which overexpresses plusTips. Top panels and red comets in the merged images are EB3-mKate, middle panels, and the green comets in the merged images are green fluorescent protein (GFP)-SKAP. The observation timings were the same with the case of no-ETL/OL (0, 5, 10, 15, 20, and 25 s). Bar = 10 μm.

FIG. 7.

Lens evaluation with a biological application. Observation of red and green comets overlapping with ETL-10-30 of a cell line, which overexpresses plusTips. Top panels and red comets in the merged images are EB3-mKate, middle panels, and the green comets in the merged images are green fluorescent protein (GFP)-SKAP. The observation timings were the same with the case of no-ETL/OL (0, 5, 10, 15, 20, and 25 s). Bar = 10 μm.

Close modal

In this paper, we present the use of an ETL as a fast axial focusing device for imaging the intracellular proteins, which move at a rate of about 250 nm/s. The advantages of using ETL are its focusing speed combined with its fine control of the focal plane. In our case, temporal resolution was within 10 ms and spatial resolution was less than 40 nm, which satisfy the required conditions. The ETL we assembled in this study used the commercially available EL-10-30 which provides sufficient optical and mechanical performance. However, when combined with 100 × objective, there are certain degressions in optical performance because of the change of NA. We evaluated and confirmed that an ETL/OL assembly with 100 × objective is applicable to detect and track protein complexes.

To estimate the behavior of ETL/OL assembly with a 100 × objective, we simulated how it would work by ZEMAX. Our simulation results and experimental results were reasonably compatible. There was a small difference between the two results. The difference between the simulation and the experiments could be brought about by the difference between the basic information of the objective lens which we picked up from the patent information.19 

Our experimental results show that the ETL/OL with a 100 × objective can achieve high spatiotemporal resolution in axial focus shifts. If the faster transition is needed, we are able to shorten the stabilization time by using three low-pass filters (300 Hz) in series to a driving signal. For most of our intracellular observations, a small axial scanning range of (2 μm) is sufficient. In our system, we needed to change the ETL control current about 12 mA to achieve the 2 μm focus shift. If we choose 16.75–28.75 mA, the relative magnification will be 1.0096–0.9904; this can be translated into the change of the pixel size 65.4–66.6 nm. This is 1.8% change of the size from the top to bottom. Regarding quantity, a 90 nm difference could occur on a 5 μm spindle from the first focus plane to the last; this means 1 or 2 pixels difference during the single set of z-stacking. We can estimate this is a small effect as we expected at first. We may also recalculate precise, quantitative information for any case based on our results in this paper.

By applying the ETL to live-cell imaging, we could observe intracellular structures at high spatiotemporal resolution. We also confirmed that its optical performance was high enough to detect moving intracellular protein complexes. However, there remain some problems to be solved in the future when tracking the molecules in 3-dimension. The microscope we used was wide-field, but it is essential to employ a confocal microscope in three dimensional analyses. Moreover, improvements could be realized by using fluorescent probes that are bright enough to be observed at a video rate.

We revealed that the ETL is applicable for 3-dimensional analyses of dynamic intracellular protein complexes. Measuring the dynamic movement of intracellular proteins is essential to precisely determine the mechanisms of biological processes. In particular, we are interested in the dynamics of microtubule-ends in dividing human cells for modelling spindle orientation mechanisms that define the plane of cell division.9,27–30 This requires us to overcome a long-standing technical hurdle for acquiring appropriate quantitative data with improved speed and z-axis resolution. Our ETL/OL assembly system with 100 × objective which maintains high optical performance and an efficient stabilization time for quick focusing will be an efficient solution for capturing the dynamics of intracellular structures.

We declare that the research was conducted in the absence of any commercial or financial relation that could be construed as a conflict of interest.

We are deeply grateful to Professor Helge Ewers of Institute of Biochemistry, ETH Zurich, and Dr. Fabian Voigt, in the group of Fritjof Helmchen at the University of Zurich, who kindly offered much advice and scientific support to us. We also thank to Optotune AG (Switzerland) for their kind support. We thank Duccio Conti for helping with proof reading of manuscript.

This work is supported by a CASIO grant (Japan), Konica Minolta (Japan), award to N. Hiroi. This work was supported by Japan Society for the Promotion of Science (JSPS) Grants-in-Aid for Scientific Research (Kiban B) 24300112, and Grant-in-Aid for Challenging Exploratory Research (Chousenteki Houga) 26640132, award to A. Funahashi. This work was supported by a Cancer Research UK Career Development Award to V. M. Draviam. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

1.
E.
Meijering
,
O.
Dzyubachyk
,
I.
Smal
 et al, “
Methods for cell and particle tracking
,”
Methods Enzymol.
504
,
183
200
(
2012
).
2.
K.
Jaqaman
,
D.
Loerke
,
M.
Mettlen
,
H.
Kuwata
,
S.
Grinstein
,
S. L.
Schmid
, and
G.
Danuser
, “
Robust single-particle tracking in live-cell time-lapse sequences
,”
Nat. Methods
5
,
695
702
(
2008
).
3.
D.
Loerke
,
M.
Mettlen
,
D.
Yarar
,
K.
Jaqaman
,
H.
Jaqaman
,
G.
Danuser
, and
S. L.
Schmid
, “
Cargo and dynamin regulate clathrin-coated pit maturation
,”
PLoS Biol.
7
,
e1000057
(
2009
).
4.
A.
Matov
,
K.
Applegate
,
P.
Kumar
,
C.
Thoma
,
W.
Krek
,
G.
Danuser
, and
T.
Wittmann
, “
Analysis of microtubule dynamic instability using a plus end growth marker
,”
Nat. Methods
7
,
761
-
768
(
2010
).
5.
S.
Gierke
and
T.
Wittmann
, “
EB1-recruited microtubule +TIP complexes coordinate protrusion dynamics during 3D epithelial remodeling
,”
Curr. Biol.
22
,
753
762
(
2012
).
6.
A.
Fujioka
,
K.
Terai
,
R. E.
Itoh
,
K.
Aoki
,
T.
Nakamura
,
S.
Kuroda
,
E.
Nishida
, and
M.
Matsuda
, “
Dynamics of the Ras/ERK MAPK cascade as monitored by fluorescent probes
,”
J. Biol. Chem.
281
,
8917
8926
(
2006
).
7.
B. R.
Parry
,
I. V.
Surovtsev
,
M. T.
Cabeen
,
C. S.
O’Hern
,
E. R.
Dufresne
, and
C.
Jacobs-Wagner
, “
The bacterial cytoplasm has glass-like properties and is fluidized by metabolic activity
,”
Cell
156
,
183
194
(
2014
).
8.
L.
Sironi
,
J.
Solon
,
C.
Conrad
,
T. U.
Mayer
,
D.
Brunner
, and
J.
Ellenberg
, “
Automatic quantificatikon of microtubule dynamics enables RNAi-screening of new mitotic spindle regulators
,”
Cytoskeleton
68
,
266
278
(
2011
).
9.
H. M. S.
Chin
,
K.
Nandra
,
J.
Clark
, and
V. M.
Draviam
, “
Need for multi-scale systems to determine spindle orientation regulators relevant to the initiation, progression and promotion of cancers
,”
Front. Physiol.
5
,
278
(
2014
).
10.
V. M.
Draviam
,
I.
Shapiro
,
B.
Aldridge
, and
P. K.
Sorger
, “
Misorientation and reduced stretching of aligned sister kinetochores promote chromosome missegregation in EB1- or APC-depleted cells
,”
EMBO J.
25
,
2814
2827
(
2006
).
11.
A. M.
Corrigan
,
R. L.
Shrestha
,
I.
Zulkipli
,
N.
Hiroi
,
Y.
Liu
,
N.
Tamura
,
B.
Yang
,
J.
Patel
,
A.
Funahashi
,
A.
Donald
, and
V. M.
Draviam
, “
Automated tracking of mitotic spindle pole positions shows that LGN is required for spindle rotation but not orientation maintenance
,”
Cell Cycle
12
,
2643
2655
(
2013
).
12.
A.
Wilson
, “
Tunable optics
,”
Vision Systems DESIGN
(PennWell, Tulsa, OK, 2010).
13.
B. F.
Grewe
,
F. F.
Voigt
,
M.
van’t Hoff
, and
F.
Helmchen
, “
Fast two-layer two-photon imaging of neuronal cell populations using an electrically tunable lens
,”
Biomed. Opt. Express
2
,
2035
2046
(
2011
).
14.
F. O.
Fahrbach
,
F. F.
Voigt
,
B.
Schmid
,
F.
Helmchen
, and
J.
Huisken
, “
Rapid 3D light-sheet microscopy with a tunable lens
,”
Opt. Express
21
,
21010
21026
(
2013
).
15.
J. M.
Jabbour
,
B. H.
Malik
,
C.
Olsovsky
,
R.
Cuenca
,
S.
Cheng
,
J. A.
Jo
,
Y. S. L.
Cheng
,
J. M.
Wright
, and
K. C.
Maitland
, “
Optical axial scanning in confocal microscopy using an electrically tunable lens
,”
Biomed. Opt. Express
5
,
645
652
(
2014
).
16.
A.
Edelstein
,
N.
Amodaj
,
K.
Hoover
,
R.
Vale
, and
N.
Stuurman
, “
Computer control of microscopes using μ Manager
,”
Curr. Protoc. Mol. Biol.
14
,
1
17
(
2010
).
17.
N.
Stuurman
,
N.
Amodaj
, and
R. D.
Vale
, “
Micro-manager: Open source software for light microscope imaging
,”
Microsc. Today
15
(
3
),
42
43
(
2007
), see http://www.micro-manager.org/.
18.
D.
Takao
,
A.
Taniguchi
,
T.
Takeda
,
S.
Sonobe
, and
S.
Nonaka
, “
High-speed imaging of amoeboid movement using light-sheet microscope
,”
PLoS One
7
,
e50846
(
2012
).
19.
Y.
Fujimoto
,
M.
Tokunaga
, and
K.
Abe
, “
Microscope objective lens
,”
Japanese patent
2007-034020
(8 February 2007).
20.
C.
Mattews
and
F. P.
Cordelières
, “
MetroloJ: Ann ImageJ plugin to help mon itor microscopes health
,” in
Proceedings of the ImageJ User and Developer Conference 2010
, pp.
1
6
, available online at http://imagejdocu.tudor.lu/lib/exe/fetch.php?media=plugin:analysis:metroloj:matthews_cordelieres_- _imagej_user_developer_conference_-_2010.pdf.
21.
J.
Schindelin
,
I.
Arganda-Carreras
,
E.
Frise
,
V.
Kaynig
,
M.
Longair
,
T.
Pietzsch
,
S.
Preibisch
,
C.
Rueden
,
S.
Saalfeld
,
B.
Schmid
,
J. Y.
Tinevez
,
D. J.
White
,
V.
Hartenstein
,
K.
Eliceiri
,
P.
Tomancak
, and
A.
Cardona
, “
Fiji: An open-source platform for biological-image analysis
,”
Nat. Methods
9
,
676
682
(
2012
).
22.
S. C.
Schuyler
and
D.
Pellman
, “
Microtubule ‘plus-end-tracking proteins’: The end is just the beginning
,”
Cell
105
,
421
424
(
2001
).
23.
A.
Akhmanova
and
M. O.
Steinmetz
, “
Tracking the ends: A dynamic protein network controls the fate of microtubule tips
,”
Nat. Rev. Mol. Cell. Biol.
9
,
309
322
(
2008
).
24.
K. T.
Applegate
,
S.
Besson
,
A.
Matov
,
M. H.
Bagonis
,
K.
Jaqaman
, and
G.
Danuser
, “
plusTipTracker: Quantitative image analysis software for the measurement of microtubule dynamics
,”
J. Struct. Biol.
176
,
168
184
(
2011
).
25.
R. L.
Shrestha
,
N.
Tamura
,
A.
Fries
,
N.
Levin
,
J.
Clark
, and
V. M.
Draviam
, “
TAO1 kinase maintains chromosomal stability by facilitating proper congression of chromosomes
,”
Open Biol.
4
,
130108
(
2014
).
26.
X. W.
Wang
,
X. X.
Zhuang
,
D.
Cao
,
Y. J.
Chu
,
P.
Yao
,
W.
Liu
,
L.
Liu
,
G.
Adams
,
G.
Fang
,
Z.
Dou
,
X.
Ding
,
Y.
Huang
,
D.
Wang
, and
X.
Yao
, “
Mitotic regulator SKAP forms a link between kinetochore core complex KMN and dynamic spindle microtubules
,”
J. Biol. Chem.
287
,
39380
39390
(
2012
).
27.
J. S.
Trinauer
,
J. C.
Canman
,
E. D.
Salmon
, and
T. J.
Mitchison
, “
EB1 targets to kinetochores with attached, polymerizing microtubules
,”
Mol. Biol. Cell
13
,
4308
4316
(
2002
).
28.
A. J.
Dunsch
,
E.
Linnane
,
F. A.
Barr
, and
U.
Gruneberg
, “
The astrin–kinastrin/SKAP complex localizes to microtubule plus ends and facilitates chromosome alignment
,”
J. Cell Biol.
192
,
959
968
(
2011
).
29.
J. G.
Ferreira
,
A. J.
Pereira
,
A.
Akhmanova
, and
H.
Maiato
, “
Aurora B spatially regulates EB3 phosphorylation to coordinate daughter cell adhesion with cytokinesis
,”
J. Cell Biol.
201
,
709
724
(
2013
).
30.
A.
Brüning-Richardson
,
K. J.
Langford
,
P.
Ruane
,
T.
Lee
,
J. M.
Askham
, and
E. E.
Morrison
, “
EB1 is required for spindle symmetry in mammalian mitosis
,”
PLoS One
6
,
e28884
(
2011
).