Open Access
Add to BibSonomy
Abstract
Interferometric scattering (iSCAT) microscopy is a highly sensitive imaging technique that uses common-path interferometry to detect the linear scattering fields associated with samples. However, when measuring a complex sample, such as a biological cell, the superposition of the scattering signals from various sources, particularly those along the optical axis of the microscope objective, considerably complicates the data interpretation. Herein, we demonstrate high-speed, wide-field iSCAT microscopy in conjunction with confocal optical sectioning. Utilizing the multibeam scanning strategy of spinning disk confocal microscopy, our iSCAT confocal microscope acquires images at a rate of 1,000 frames per second (fps). The configurations of the spinning disk and the background correction procedures are described. The iSCAT confocal microscope is highly sensitive—individual 10 nm gold nanoparticles are successfully detected. Using high-speed iSCAT confocal imaging, we captured the rapid movements of single nanoparticles on the model membrane and single native vesicles in the living cells. Label-free iSCAT confocal imaging enables the detailed visualization of nanoscopic cell dynamics in their most native forms. This holds promise to unveil cell activities that are previously undescribed by fluorescence-based microscopy.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Label-free optical interference microscopy is an indispensable tool for biological studies [1]. Cell biologists rely on phase contrast and differential interference contrast (DIC) microscopes to monitor native cell morphology. Using quantitative phase imaging (QPI) techniques [2], the optical phase delay associated with a sample is measured quantitatively, providing a spatially resolved estimation of the biological dry mass. In addition to QPI, a variety of scattering-based interferometric microscope techniques have been demonstrated to meet the requirements of imaging various biological cell samples [3–7]. Furthermore, utilizing the machine learning approaches with fluorescence data, label-free interference microscopy currently provides organelle-specific and functional cell images [8–10]. These phase-sensitive interference techniques facilitate the monitoring of biological cells with high spatiotemporal resolutions, and most importantly without the perturbation introduced by labeling or staining. Long-term, noninvasive live cell imaging enables the investigation of complex cell dynamics over broad spatial and temporal scales, where photobleaching and phototoxicity limit fluorescence approaches.
Cell imaging requires high resolutions in all three dimensions owing to the essentially three-dimensional (3D) nature of biological cell structures. The 3D refractive index mapping with subcellular resolutions has been demonstrated using tomographic phase microscopy (TPM) [11,12]. In TPM, axial resolution is attained through the acquisition of multiple interferograms at various illumination angles, followed by numerical reconstruction. Another technique that produces 3D scattering images of biological cells is the full-field optical coherence microscopy (OCM), also referred to as full-field optical coherence tomography [13–16]. In full-field OCM, the optical sectioning is accomplished through coherence gating with a low-coherence source of light. Despite the great success of interference microscopy in label-free cell imaging, nanoscopic cell dynamics remain a challenge for the majority of existing technologies owing to weak signals. Moreover, illumination modulation and multiple frame acquisition are required for the aforementioned techniques to provide 3D resolutions, thereby limiting the imaging speed.
To capture the nanoscopic cell dynamics, it is necessary to fulfill several requirements. First is the detection sensitivity. Small biological entities scatter less light; therefore, a sensitive microscope is required for their detection. The photon shot noise determines the weakest signal that can be detected above the photon shot noise, which determines the sensitivity of interference microscopy. Thus, the microscope must be able to accommodate sufficient illumination intensity without saturating the detector. Low-coherence light sources (LEDs and incandescent lamps that are commonly used in QPI) have trouble delivering a high radiance to the sample, which ultimately limits the detection sensitivity. Second, the dynamic nature of the cell complicates the achievable detection sensitivity in live cell imaging applications. Driven by thermal fluctuation and cell activities, small biological entities (e.g., vesicles and filaments below 100 nm) move continuously in 3D space. The microscope must have a fast enough data acquisition rate to reduce the blurring caused by rapidly moving objects. Therefore, interference microscope techniques that require multiple frames for reconstruction, such as the TPM, face challenges in capturing the rapid dynamics of nano-sized objects. The need for rapid image acquisition also precludes the use of extensive frame averaging, a common method to virtually improve the signal-to-noise ratio (SNR). The third requirement for visualizing the nanoscopic cell dynamics is the scattering background, including the nonspecific scattering background created by the optical components and the out-of-focus signal of the specimen. In principle, one can effectively remove the scattering background by postprocessing the image if the background can be accurately measured or estimated [17]. Meanwhile, the out-of-focus scattering background of the sample is eliminated through optical sectioning (e.g., confocal-based detection).
The common-path wide-field interferometric scattering (iSCAT) microscopy meets the first two aforementioned requirements [18–20]. The iSCAT microscopy detects the backscattered light of the sample through interference where the reflection of the supporting coverglass serves as the reference beam. Using laser illumination, light is efficiently delivered across the field of view (FOV). With a high-speed camera, the interference image is captured at a high frame rate [21]. iSCAT microscopy resembles reflection interference contrast microscopy (RICM) that was originally demonstrated with low-coherence light [22]. Recent advances in iSCAT microscopy have made it possible to capture the nanoscopic motion of small nanoparticles with microsecond temporal resolution [23]. With proper signal integration and background correction, iSCAT microscopy enables the direct visualization of single unlabeled biological macromolecules on a clean coverglass [24,25].
The wide-field iSCAT microscopy has excellent sensitivity and speed, allowing for resolving the dynamics of plasma membrane and intracellular vesicles [6,26,27]. Nevertheless, the wide-field iSCAT is not optimized for label-free cell imaging. The cell has a complex 3D structure that produces a great amount of out-of-focus background. Notably, iSCAT detects cell membrane reflection well, producing interference fringes that make intracellular organization difficult to visualize [1]. In an early work of iSCAT imaging of small nanoparticles, Sandoghdar et al. utilized a confocal-based configuration to reduce the out-of-focus scattering background [28]. Recently, the same group demonstrated laser scanning confocal-based iSCAT microscopy for the 3D imaging of biological cells [29]. The iSCAT confocal microscope is essentially a reflectance confocal microscope with a pinhole mask defining the detection volume [30,31]. Sandoghdar et al. showed that when using a reflectance confocal microscope to acquire images of cells close to a coverglass (within 1–2 μm), the reflection of the coverglass serves as a reference field that interferes with the backscattered signal of the cells. Interferometric detection enhances the signal contrast, enabling the detection of weakly scattering objects. In addition, the out-of-focus cell background was significantly reduced, and the nanoscopic cell dynamics can be visualized with in exquisite detail. However, the method of point scanning inherently limits the speed of image acquisition. With a typical pixel dwell time of 1 μs, it will take hundreds of milliseconds to capture a high-resolution image (e.g., 512 × 512 resolution at a frame rate of a few hertz).
Herein, we demonstrate a high-speed iSCAT confocal microscope that can capture hundreds of frames per second (fps), i.e., approximately two orders of magnitude faster than its point scanning counterpart. With a reduced FOV, an image acquisition rate of 1,000 fps is achieved. To accomplish this, a spinning disk confocal fluorescence microscope is modified into an iSCAT confocal microscope. Polarization optics are implemented to enhance the signal collection efficiency and reduce the ambient background noise. Using nanoparticle samples, we evaluated the detection sensitivity and speed of our iSCAT confocal microscope. Very small single gold nanoparticles (AuNPs), as small as 10 nm in diameter, are successfully detected. Moreover, we present high-speed 3D cell imaging with iSCAT confocal microscopy in which the nanoscopic diffusion of native vesicles can be resolved.
2. Results
2.1 Spinning disk iSCAT confocal microscopy
We developed a confocal-based iSCAT microscopy by utilizing a commercial spinning disk scanner unit (Fig. 1). A laser diode module with a 561 nm wavelength (OBIS 561, Coherent) was used as the light source. A Yokogawa confocal scanner unit (CSU-X1) was installed on an inverted microscope (Eclipse Ti2, Nikon). To convert the original fluorescence detection into scattering detection, polarization optics techniques were utilized. Specifically, we replaced the dichroic mirror with a custom-made plate polarization beamsplitter (PBS) (Control Optics Taiwan Inc.). The PBS was made via sputter coating on a fused silica substrate, with an extinction ratio of Tp:Ts > 1000:1 at an incidence angle of 45° at the wavelength of 561 nm. In addition, a quarter-wave plate (QWP) (WPQ10M-561, Thorlabs) was added to the filter cube of the turret. The combination of the PBS and the QWP considerably enhanced the collection efficiency of the iSCAT signal of the sample. Moreover, a linear polarizer (LPVISC100, Thorlabs) was installed in front of the camera to eliminate the nonspecific reflection background of the apparatus (mainly contributed by the reflection of the pinhole disk). A very weak reflective background of the pinhole disk remains, which is thought to be created by the depolarization effects of the backscattering of the pinholes. A high numerical aperture (NA), oil-immersion microscope objective (CFI Plan Apochromat Lambda 100X Oil, Nikon) was used. As the detector, a scientific CMOS camera (Zyla, Andor) allowed us to record high-speed iSCAT videos. Our confocal microscope could be converted from iSCAT to fluorescence imaging simply by replacing the PBS and linear polarizer with a dichroic beamsplitter and a fluorescence emission filter, respectively.
Fig. 1. Schematic diagrams of the spinning disk iSCAT confocal microscopy. PM: polarization maintaining fiber; ML: microlens array; PH: pinhole array; PBS: polarization beamsplitter; QWP: quarter waveplate; LP: linear polarizer, and OBJ: microscope objective. The inset depicts the iSCAT detection of the backward scattering of a cell that interferes with the reference field reflected from the water-coverglass interface.
We operated the spinning disk at its maximal rotation speed of 10,000 rpm. The Yokogawa scanner unit utilizes a Nipkow disk that scans the specimen once for every 30° of rotation. Thus, the duration of a single sample scan is 0.5 ms, and one disk rotation yields 12 sample scans. To stabilize the illumination intensity, the camera was set to global shutter mode with a frame time and an exposure time of multiples of 0.5 ms. For example, we set a frame time of 1 ms (1,000 fps) with an exposure time of 0.5 ms, a frame time of 2 ms (500 fps) with an exposure time of 1.5 ms, and a frame time of 6 ms (166.67 fps) with an exposure time of 5.5 ms. Using the aforementioned settings, we avoid the spatially moving fringes in a video caused by the frequency mismatch between the disk rotation and the image acquisition. Notably, we still observed a periodic variation of 10% peak-to-peak in the measured intensity every 1 ms. We interpret this periodic variation as the result of the imperfect manufacture of the spinning disk, yielding distinct but repeatable illumination patterns every 30° of rotation. To minimize the intensity variation between individual scans, we normalized each image of a video with its averaged spatial intensity. After normalization, the intensity fluctuation was ∼1%.
2.2 High-speed imaging and tracking of single nanoparticles
Beginning with gold nanoparticles (AuNPs), we characterized the imaging sensitivity of our iSCAT confocal microscope. Single 30-nm diameter AuNPs (BBI Solutions) were immobilized on a clean coverglass via spin coating and submerged in water for imaging (see the detailed protocol for nanoparticle sample preparation in Ref. [32]). When the sample particle was positioned at the focal plane of the microscope objective, the backscattered light of the particle and the reflection of the water-glass interface were collected and projected onto the camera. The measured intensity can be written as
where ${E_r}$ is the reflected reference field, ${E_s}$ is the backscattered signal field, and $\theta $ is the phase difference between the two fields. For weakly scattering objects, the ${|{{E_s}} |^2}$ is negligible compared to the other two terms in Eq. (1). The iSCAT contrast is defined as the interferometric visibility of the signal.Note that the interference contrast is determined by the ratio of the backscattered signal field to the reference field and the phase difference between the two fields. While both optical fields are spatially filtered by the confocal pinhole, these two fields still have different spatial modes (because the pinhole is not infinitely small) and thus they undergo different phase evolutions along the z direction (i.e., $\theta $ is a function of z). Figure 2(a) displays the iSCAT confocal contrast image of 30 nm AuNPs exhibiting destructive interference between the signal and the reference ($\cos \theta \approx{-} 1$; the particle appears as a dark spot in the image). The normalized iSCAT contrast of 30 nm AuNP is ∼0.19 (Fig. 2(b)), which is comparable to the contrast obtained using wide-field iSCAT microscopy [23]. The high contrast of AuNP indicates that our polarization optics successfully eliminates the nonspecific background reflection of the scanning unit. Conversely, when these polarization optics were removed and replaced with a non-polarizing beam-splitter, the nonspecific background considerably reduced the sensitivity (Fig. S1, Supplement 1).
Fig. 2. iSCAT confocal imaging and tracking of single 30 nm AuNPs. (a) Single 30 nm AuNPs immobilized on a coverglass. Inset: a closeup view of the iSCAT image of a 30 nm AuNP and its contrast line profile with a Gaussian fitting (red curve). (b) The histogram of iSCAT contrast of single 30 nm AuNPs. (c) Schematic diagram of AuNPs attaching to lipid molecules in the supported lipid bilayer membrane on a coverglass. (d)-(f) Diffusion trajectories of AuNPs on the membrane recorded at 166.67 fps (d), 500 fps (e), and 1,000 fps (f), respectively. See Visualization 1, Visualization 2, and Visualization 3 for the videos of AuNP diffusion on the membrane acquired at 166.67 fps, 500 fps, and 1,000 fps, respectively.
We emphasize that the accurate measurement of the reference intensity ${|{{E_r}} |^2}$ is essential for detecting the weakly scattering objects with a low iSCAT contrast. The reference intensity (also referred to as “background intensity”) is generally spatially heterogeneous owing to the nonuniform illumination and inevitable reflection from the optical elements. When the target of interest is in motion, we estimated the background intensity by calculating the temporal median image of an iSCAT video [17]. As discussed in Sec. 2.1, the spinning disk confocal microscope produced nonidentical but reproducible illumination backgrounds between individual scans (∼1% intensity variation). To further reduce the background fluctuation, we calculated the multiple temporal median background images for correction when the frame time is shorter than the time required for the spinning disk to make one revolution (i.e., 6 ms). For example, three background images were calculated for a frame time of 2 ms, while six background images were calculated for a frame time of 1 ms. For a static sample, we modulated the sample position laterally and extracted the static background intensity from the moving signal. The detailed procedures for background correction and its performance are discussed in Figs. S2 and S3 (Supplement 1).
The key advantage of the spinning disk iSCAT confocal microscope is its ability to acquire high-speed videos with optical sectioning. Herein, we demonstrate high-speed imaging of single nanoparticles diffusing on a supported lipid bilayer membrane. This model membrane system has been used as a platform to study single-molecule membrane dynamics [33,34]. An artificial lipid bilayer membrane was prepared on the coverglass. Biotinylated phospholipids (1 mol%) were incorporated into the membrane for AuNP labeling through streptavidin (see the schematic in Fig. 2(c)). After attachment, the AuNPs diffused laterally on the membrane whose motion was recorded using an iSCAT confocal microscope at 166.67 fps (Visualization 1). The diffusion trajectory was reconstructed by locating the AuNP in each image with a home-written MATLAB code followed by connecting the consecutive localizations (Fig. 2(d)). The SNR of 30 nm AuNP is ∼18, yielding a localization precision of ∼6.5 nm. It is possible to acquire images at a faster rate with a narrower FOV. Visualization 2 and Visualization 3 show the AuNP diffusion acquired at 500 fps and 1,000 fps, respectively. The SNR decreased slightly as the imaging speed was increased (SNR of ∼11 for 500 fps and ∼9 for 1,000 fps, respectively). We attribute the decrease in SNR to the twofold increase in the noise floor caused by the lower photon flux. The diffusion trajectories recorded at 500 and 1,000 fps are depicted in Figs. 2(e) and 2(f), respectively. The faster image acquisition rate enables us to resolve detailed diffusive movement at the nanoscale.
As shown in Eq. (2), the iSCAT contrast of a particle is determined by the ratio of the signal and reference fields. Therefore, to detect very small nanoparticles, the iSCAT contrast of the particle can be enhanced by decreasing the intensity of the reference beam. This can be accomplished by replacing the sample immersion water with a high refractive index liquid, which decreases the reflectivity at the sample-coverglass interface and thus reduces the reference beam intensity. We replaced water with glycerol (G2025, Sigma-Aldrich) with a refractive index of 1.474. It reduced the reflectivity by a factor of 21.5 (from 0.43% to 0.02%, assuming the refractive indices of water and coverglass are 1.33 and 1.517, respectively). Meanwhile, exchanging the medium from water to glycerol also modified the scattering cross section of AuNP slightly (increased by ∼33%; the change is small because gold has a large complex refractive index [35]). Taken together, the medium exchange yields a $\sqrt {1.33 \times 21.5} \cong 5.3$ fold increase in iSCAT contrast. With contrast enhancement, we successfully detected single 10 nm AuNPs immobilized on the coverglass (Fig. 3). The enhanced iSCAT contrast of 10 nm AuNP is ∼0.04, yielding a SNR ∼4.
Fig. 3. iSCAT confocal image of single 10 nm AuNPs. The inset plots the iSCAT contrast line profile together with a Gaussian fitting (red curve). A spatial Gaussian filter with a standard deviation of 0.7 pixels is applied.
Finally, we verify that the measurement of our iSCAT confocal imaging is shot-noise-limited and thus the detection sensitivity can be improved by increasing the number of detected photons through averaging (see more discussion on the shot-noise-limited iSCAT imaging in Ref. [23]). Specifically, we acquire the iSCAT images of a clean coverglass at the maximal frame rate of 1,000 fps and analyze the fluctuation of its iSCAT contrast. We define the noise as the standard deviation of the contrast fluctuation. The dependency of the noise on the number of detected photons N follows the power law of $1/\sqrt N $ (Fig. S4, Supplement 1), indicating that the major noise is our iSCAT imaging is the photon shot noise at the highest image acquisition rate.
2.3 Live cell imaging by iSCAT confocal microscopy
We set out to examine the 3D imaging capabilities of iSCAT confocal microscopy using biological cells. Human bone osteosarcoma epithelial (U2OS) cells were cultured on a coverglass-bottom dish (WillCo Wells). We examined the optical sectioning capability of the iSCAT confocal by acquiring a z stack of images of a cell (Visualization 4). As the water-coverglass interface moves out of focus, the intensity of the reference beam decreases owing to the confocal sectioning (Fig. S5, Supplement 1). However, within an axial distance of ∼1 μm, the intensity of the reference beam is still sufficient for iSCAT imaging. The strongly attenuated reference beam makes it challenging to acquire iSCAT images deeper within the sample. Figure 4(a) displays the iSCAT confocal image of a cell with the focal plane positioned at the basal membrane of the cell (i.e., at the cell-coverglass interface). At this z position, we observed distinct iSCAT contrast variation due to the morphology of the basal membrane, reminiscent of the adhesion factors previously observed by the wide-field iSCAT microscopy [27]. When the focal plane was moved into the cell by 0.5 μm, distinct features became visible, and we were able to see the cell nucleus enclosed by the nuclear membrane (Fig. 4(b)). The great differences in the two iSCAT confocal images at the two axial positions illustrate the excellent optical sectioning capability of the microscope by eliminating the out-of-focus scattering signal.
Fig. 4. The comparison between the confocal and wide-field iSCAT images of a biological cell. (a),(b) The iSCAT confocal image of a U2OS cell at z = 0 μm (a) and z = 0.5 μm (b). See the z stack video of the cell in Visualization 4. (c),(d) The iSCAT wide-field images of the same cell shown in (a) and (b) acquired at the corresponding axial positions, i.e., z = 0 μm (c) and z = 0.5 μm (d). While the two confocal images (a) and (b) display distinct features because of the optical sectioning, the two wide-field images (c) and (d) show similar features as the result of a lower axial resolution. Please note that the figures in (c) and (d) are tiled images of 4 × 4 original images because the wide-field iSCAT microscope has a smaller FOV.
To illustrate the effect of optical sectioning on 3D cell imaging, we use a standard wide-field iSCAT microscope to acquire a z stack iSCAT images of the same cell as a comparison. The optical setup of the wide-field iSCAT microscopy has been reported previously [23], and we summarized it in Fig. S6 (Supplement 1). The wide-field iSCAT images at the two corresponding heights (z = 0 μm and z = 0.5 μm) are displayed in Figs. 4(c) and 4(d), respectively. Unlike the two iSCAT confocal images, the two wide-field iSCAT images show very similar features because the wide-field iSCAT microscope has a lower axial resolution. We also note that, in the wide-field iSCAT images, the concentric interference fringes appear on the cell periphery over a long axial range due to the reflection of the plasma membrane. Such interference rings are largely eliminated by the optical sectioning of the iSCAT confocal microscope.
Our iSCAT confocal microscope is compatible with the confocal fluorescence imaging. As a demonstration, we labeled the DNA in the cell nucleus with the fluorescent dye DRAQ5 (ab108410, Abcam) and stained the cell membrane with a lipophilic dye (D3898, FAST DiO, Invitrogen). The iSCAT confocal image of a cell and the corresponding fluorescence images of the two dyes are shown in Fig. 5. We note that both the plasma membrane and the nuclear membrane produce strong iSCAT signals. Furthermore, we observed many small particles in the iSCAT channel that were colocalized with the DNA and lipophilic dyes (Insets of Fig. 5(a)–5(c)), indicating that they are involved in the uptake of the DNA dye and lipid-rich vesicles.
Fig. 5. iSCAT and fluorescence confocal images of a U2OS cell. (a) iSCAT confocal image of the cell. (b) Two-color fluorescence image of the cell. Fluorescence signals of the lipophilic dye and the DNA dye are displayed in green and red, respectively. (c) Overlay of the iSCAT image (a) and the fluorescence image (b). The insets display the close-up views of the cell vesicles. The magenta arrows indicate the vesicles that appear both in the iSCAT and fluorescence channels, whereas the blue arrows indicate the particles that are clearly detected in iSCAT but show low fluorescence intensity.
We observed through our iSCAT channel that many nanoparticles are in constant motion within living cells (Visualization 5). Using single-particle tracking and the fast image acquisition of our iSCAT confocal microscopy, we can measure the diffusive motion of these fast-moving vesicles [32]. Figure 6(a) plots a snapshot of the iSCAT image of multiple cells. Four regions of interest (ROIs) are indicated by the rectangular boxes whose closeup views are shown in Fig. 6(b)–6(e). Within these ROIs, nano-sized biological particles are detected and their diffusive motions are measured. To improve the localization accuracy, we measured the static cellular background using temporal median filtering and then removed the background from the raw image by division. The trajectories of the biological nanoparticles are shown in Fig. 6(b)–6(e). The majority of the particles diffuses within local confinements with a maximum displacement of less than 450 nm per second. These particles are confined within cytoskeleton meshworks. Occasionally, we observed some particles translocate over a longer distance of more than ∼1.5 μm through directional motion, which we interpret as the result of the active transport of the cell. The continuous single-particle tracking is often interrupted when the particle diffuses out of the focal plane. Indeed, the iSCAT confocal microscope has a rather thin detection volume due to its optical sectioning capability. While we demonstrate the particle tracking in 2D, the 3D particle tracking should be possible with proper modeling of the point-spread function and calibration [36].
Fig. 6. Diffusive motion of biological nanoparticles observed using high-speed iSCAT confocal microscopy. (a) A snapshot of iSCAT confocal video of multiple cells (see the full video in Visualization 5). The regions of nuclei are marked in red, corresponding to the fluorescence image of the DNA staining. (b)-(e) Closeup view of the four ROIs indicated in (a). The diffusion trajectories of the vesicles are plotted in red. Particles undergoing directional diffusion are indicated by magenta arrows. (b) ROI1 marks a region at the cell periphery where the particles move along the cell boundary. (c) ROI2 shows a region at the border of a nucleus where the directional motion of a particle is observed. (d) In ROI3, the association of a particle to a large cell structure (indicated by the yellow arrow) is monitored (see Visualization 6 for the closeup video). (e) ROI4 displays the random diffusion of particles in the cytoplasm.
3. Discussion
In this study, we report a high-speed iSCAT confocal imaging technique to detect the backscattered light of nano-objects. Our iSCAT confocal microscope shares the same working principle as a reflectance confocal microscope (RCM) by detecting the backscattered light with a confocal microscope system [37]. Notably, the RCM has been widely used for noninvasive tissue imaging and disease diagnostics [38]. Using near-infrared laser light, RCM is optimized for deep tissue imaging with a subcellular spatial resolution. Unlike the iSCAT confocal that measures the interference signal, RCM imaging generally lacks a well-defined reference field for interferometric detection. Thus, the RCM measures the backscattering intensity, and its image is dark-field-like (i.e., the object that backscatters light appears bright in the image with an overall dark background). Using a high-NA objective and a visible laser light for illumination, RCM generates dark-field-like backscattering cell images [30]. The achievable detection sensitivity of dark-field detection is typically lower than the interferometric detection because it is highly sensitive to the nonspecific background and readout noise (see more comparisons between iSCAT and dark-field imaging in Ref. [39]).
Our current iSCAT confocal microscope design operates close to the coverglass interface whose reflection serves as the reference field for interferometric detection. As the coverglass reflection disappears when the interface is no longer in focus, the imaging depth for iSCAT detection is effectively limited to a few micrometers. This imaging depth is generally sufficient for analyzing thin adherent cells. Within this detection volume, we demonstrate that iSCAT detection considerably improves imaging sensitivity and reveals weakly scattered signals from nano-objects. We note that in a recent demonstration of the laser scanning iSCAT confocal microscope, several cell organelles and nanoscopic cell structures are successfully resolved, including the endoplasmic reticulum and microtubules [29]. Because the strategy of achieving iSCAT confocal imaging is essentially the same (using the confocal pinhole), we believe our spinning disk iSCAT confocal microscope could also resolve these cell organelles or even mitochondria and lipid droplets that have been clearly visualized by other phase-sensitive interference phase microscopy [12]. We point out that the challenges of label-free visualization of cell organelles are often due to the lack of signal-to-background ratio (SBR), instead of the lack of detection sensitivity. This is because all cell structures scatter light and it becomes very difficult to distinguish them solely based on the iSCAT image when they are densely packed. We also note that it is relatively easy to resolve different cell organelles at the cell periphery where the cell structures are mostly 2D and spatially separated.
Using spinning disk confocal microscopy and a high-speed camera, our current iSCAT confocal imaging achieves a rate of 1,000 fps. An improved frame rate can be achieved with a faster rotating spinning disk and a faster camera. We also note that the spinning disk iSCAT confocal microscope has a potential advantage of achieving a large FOV at a high speed. In the operation of a Yokogawa spinning disk confocal microscope, around 1,000 laser foci are scanned across sample at a maximal speed of ∼1,000 Hz. In comparison, the wide-field iSCAT microscopes typically illuminate the sample by scanning a single beam with a pair of acousto-optic deflectors (AODs) at a maximal scanning speed of ∼100 kHz [23]. Thus, it appears that the multi-beam scanning of the spinning disk confocal could cover a larger FOV within a shorter frame time. To increase the imaging depth of a thick 3D sample, it would be necessary to generate a steady reference beam, possibly with Michelson and Mirau objectives [40,41]. With a steady reference field, the axial range of iSCAT confocal imaging would be determined by the working distance of the objective.
While we demonstrated iSCAT confocal microscopy in reflection in this study, it appears that transmission geometry would offer several advantages. In transmission, the nonscattered transmitted beam readily serves as a robust reference beam on the common path. Therefore, the imaging depth would not be restricted as in the case of reflection. In addition, unlike the reflection geometry that detects much of the membrane reflection, transmission imaging is more effective at detecting the intracellular signals [1]. Wide-field iSCAT microscopy in transmission, referred to as coherent brightfield (COBRI) microscopy, has been demonstrated to resolve intracellular nanoscopic cell dynamics, including vesicle transportation, chromatin remodeling, and virus-membrane interactions [5,9,42]. Transmission confocal microscopy is a less developed technology compared to its reflection counterpart because of the difficulty of descanning the transmitted beam with the synchronization of the incident beam. The descanning of the transmitted beam can be achieved by sending it back to the original optical paths [43] or, in theory, adding a synchronized descanning unit in transmission can descan the transmitted beam.
4. Conclusion
We demonstrate a high-speed spinning disk iSCAT confocal microscope for wide-field iSCAT imaging with optical sectioning. The operational conditions, including polarization optics and the synchronization of the disk rotation and image acquisition, are described. The iSCAT confocal microscope enabled the visualization of single nanoparticles as small as 10 nm AuNPs. In addition, the high-speed tracking of the nanoscopic motion of a single AuNP (up to 1,000 fps) was demonstrated. Moreover, the iSCAT confocal imaging of living cells was demonstrated. The plasma membrane and nuclear membrane were clearly visualized and resolved with optical sectioning. Furthermore, numerous nano-sized cell vesicles diffusing in the cytoplasm in 3D were observed. In conjunction with multicolor fluorescence confocal imaging, our iSCAT confocal imaging provides rich structural and dynamic information about the cell sample. Thus, we anticipate that iSCAT confocal microscopy will become the standard imaging technique for confocal microscopy when examining cell samples.
Funding
National Science and Technology Council (109-2112-M-002-026-MY3, 111-2112-M-001-051-MY5, 111-2119-M-002-013-MBK, 111-2321-B-002-016); Academia Sinica (AS-CDA-107-M06, AS-iMATE-111-35).
Acknowledgments
The authors thank the Biophysics Core Facility at the Institute of Atomic and Molecular Sciences (IAMS), Academia Sinica for the support. The authors thank the supports from the NTHU Brain Research Center and NTU-110L8809 under the Higher Education Sprout Project funded by the Ministry of Science and Technology and Ministry of Education in Taiwan.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. Y.-T. Hsiao, C.-N. Tsai, C.-Y. Cheng, and C.-L. Hsieh, “Molecularly specific and functional live cell imaging by label-free interference microscopy,” ACS Photonics 9(7), 2237–2245 (2022). [CrossRef]
2. Y. Park, C. Depeursinge, and G. Popescu, “Quantitative phase imaging in biomedicine,” Nat. Photonics 12(10), 578–589 (2018). [CrossRef]
3. F. Jünger, D. Ruh, D. Strobel, R. Michiels, D. Huber, A. Brandel, J. Madl, A. Gavrilov, M. Mihlan, C. C. Daller, E. A. Rog-Zielinska, W. Römer, T. Lämmermann, and A. Rohrbach, “100 Hz ROCS microscopy correlated with fluorescence reveals cellular dynamics on different spatiotemporal scales,” Nat. Commun. 13(1), 1758 (2022). [CrossRef]
4. L. M. Almassalha, G. M. Bauer, J. E. Chandler, S. Gladstein, L. Cherkezyan, Y. Stypula-Cyrus, S. Weinberg, D. Zhang, P. T. Ruhoff, H. K. Roy, H. Subramanian, N. S. Chandel, I. Szleifer, and V. Backman, “Label-free imaging of the native, living cellular nanoarchitecture using partial-wave spectroscopic microscopy,” Proc. Natl. Acad. Sci. 113(42), E6372–E6381 (2016). [CrossRef]
5. Y.-F. Huang, G.-Y. Zhuo, C.-Y. Chou, C.-H. Lin, W. Chang, and C.-L. Hsieh, “Coherent brightfield microscopy provides the spatiotemporal resolution to study early stage viral infection in live cells,” ACS Nano 11(3), 2575–2585 (2017). [CrossRef]
6. J.-S. Park, I.-B. Lee, H.-M. Moon, J.-H. Joo, K.-H. Kim, S.-C. Hong, and M. Cho, “Label-free and live cell imaging by interferometric scattering microscopy,” Chem. Sci. 9(10), 2690–2697 (2018). [CrossRef]
7. R. Imai, T. Nozaki, T. Tani, K. Kaizu, K. Hibino, S. Ide, S. Tamura, K. Takahashi, M. Shribak, and K. Maeshima, “Density imaging of heterochromatin in live cells using orientation-independent-DIC microscopy,” Mol. Biol. Cell 28(23), 3349–3359 (2017). [CrossRef]
8. Y. Jo, H. Cho, S. Y. Lee, G. Choi, G. Kim, H. S. Min, and Y. Park, “Quantitative phase imaging and artificial intelligence: A review,” IEEE J. Sel. Top. Quantum Electron. 25(1), 1–14 (2019). [CrossRef]
9. Y.-T. Hsiao, C.-N. Tsai, T.-H. Chen, and C.-L. Hsieh, “Label-free dynamic imaging of chromatin in live cell nuclei by high-speed scattering-based interference microscopy,” ACS Nano 16(2), 2774–2788 (2022). [CrossRef]
10. C. Ounkomol, S. Seshamani, M. M. Maleckar, F. Collman, and G. R. Johnson, “Label-free prediction of three-dimensional fluorescence images from transmitted-light microscopy,” Nat. Methods 15(11), 917–920 (2018). [CrossRef]
11. W. Choi, C. Fang-Yen, K. Badizadegan, S. Oh, N. Lue, R. R. Dasari, and M. S. Feld, “Tomographic phase microscopy,” Nat. Methods 4(9), 717–719 (2007). [CrossRef]
12. P. A. Sandoz, C. Tremblay, F. G. van der Goot, and M. Frechin, “Image-based analysis of living mammalian cells using label-free 3D refractive index maps reveals new organelle dynamics and dry mass flux,” PLoS Biol. 17(12), e3000553 (2019). [CrossRef]
13. C. Apelian, F. Harms, O. Thouvenin, and A. C. Boccara, “Dynamic full field optical coherence tomography: subcellular metabolic contrast revealed in tissues by interferometric signals temporal analysis,” Biomed. Opt. Express 7(4), 1511–1524 (2016). [CrossRef]
14. C.-Y. Tsai, C.-H. Shih, H.-S. Chu, Y.-T. Hsieh, S.-L. Huang, and W.-L. Chen, “Submicron spatial resolution optical coherence tomography for visualising the 3D structures of cells cultivated in complex culture systems,” Sci. Rep. 11(1), 3492 (2021). [CrossRef]
15. J. Scholler, K. Groux, O. Goureau, J.-A. Sahel, M. Fink, S. Reichman, C. Boccara, and K. Grieve, “Dynamic full-field optical coherence tomography: 3D live-imaging of retinal organoids,” Light: Sci. Appl. 9(1), 140 (2020). [CrossRef]
16. E. Beaurepaire, A. C. Boccara, M. Lebec, L. Blanchot, and H. Saint-Jalmes, “Full-field optical coherence microscopy,” Opt. Lett. 23(4), 244–246 (1998). [CrossRef]
17. C.-Y. Cheng and C.-L. Hsieh, “Background estimation and correction for high-precision localization microscopy,” ACS Photonics 4(7), 1730–1739 (2017). [CrossRef]
18. J. Ortega-Arroyo and P. Kukura, “Interferometric scattering microscopy (iSCAT): new frontiers in ultrafast and ultrasensitive optical microscopy,” Phys. Chem. Chem. Phys. 14(45), 15625–15636 (2012). [CrossRef]
19. R. W. Taylor and V. Sandoghdar, “Interferometric scattering microscopy: seeing single nanoparticles and molecules via rayleigh scattering,” Nano Lett. 19(8), 4827–4835 (2019). [CrossRef]
20. C.-L. Hsieh, “Label-free, ultrasensitive, ultrahigh-speed scattering-based interferometric imaging,” Opt. Commun. 422, 69–74 (2018). [CrossRef]
21. C.-Y. Cheng, Y.-H. Liao, and C.-L. Hsieh, “High-speed imaging and tracking of very small single nanoparticles by contrast enhanced microscopy,” Nanoscale 11(2), 568–577 (2019). [CrossRef]
22. A. S. G. Curtis, “The mechanism of adhesion of cells to glass: a study by interference reflection microscopy,” J. Cell Biol. 20(2), 199–215 (1964). [CrossRef]
23. Y.-H. Lin, W.-L. Chang, and C.-L. Hsieh, “Shot-noise limited localization of single 20 nm gold particles with nanometer spatial precision within microseconds,” Opt. Express 22(8), 9159–9170 (2014). [CrossRef]
24. G. Young, N. Hundt, D. Cole, A. Fineberg, J. Andrecka, A. Tyler, A. Olerinyova, A. Ansari, E. G. Marklund, M. P. Collier, S. A. Chandler, O. Tkachenko, J. Allen, M. Crispin, N. Billington, Y. Takagi, J. R. Sellers, C. Eichmann, P. Selenko, L. Frey, R. Riek, M. R. Galpin, W. B. Struwe, J. L. P. Benesch, and P. Kukura, “Quantitative mass imaging of single biological macromolecules,” Science 360(6387), 423–427 (2018). [CrossRef]
25. M. Piliarik and V. Sandoghdar, “Direct optical sensing of single unlabelled proteins and super-resolution imaging of their binding sites,” Nat. Commun. 5(1), 4495 (2014). [CrossRef]
26. J.-S. Park, I.-B. Lee, H.-M. Moon, S.-C. Hong, and M. Cho, “Long-term cellular cargo tracking reveals intricate trafficking through active cytoskeletal networks,” bioRxiv, 2022.2003.2023.485568 (2022).
27. J.-S. Park, I.-B. Lee, H.-M. Moon, J.-S. Ryu, S.-Y. Kong, S.-C. Hong, and M. Cho, “Fluorescence-combined interferometric scattering imaging reveals nanoscale dynamic events of single nascent adhesions in living cells,” J. Phys. Chem. Lett. 11(23), 10233–10241 (2020). [CrossRef]
28. K. Lindfors, T. Kalkbrenner, P. Stoller, and V. Sandoghdar, “Detection and spectroscopy of gold nanoparticles using supercontinuum white light confocal microscopy,” Phys. Rev. Lett. 93(3), 037401 (2004). [CrossRef]
29. V. Sandoghdar, M. Kueppers, D. Albrecht, A. Kashkanova, and J. Luehr, “Confocal interferometric scattering microscopy reveals 3D nanoscopic structure and dynamics in live cells,” Research Square (2022).
30. N. Lue, W. Choi, G. Popescu, Z. Yaqoob, K. Badizadegan, R. R. Dasari, and M. S. Feld, “Live cell refractometry using hilbert phase microscopy and confocal reflectance microscopy,” J. Phys. Chem. A 113(47), 13327–13330 (2009). [CrossRef]
31. C. Smithpeter, A. Dunn, R. Drezek, T. Collier, and R. Richards-Kortum, “Near real time confocal microscopy of cultured amelanotic cells: sources of signal, contrast agents and limits of contrast,” J. Biomed. Opt. 3(4), 429 (1998). [CrossRef]
32. C.-L. Hsieh, S. Spindler, J. Ehrig, and V. Sandoghdar, “Tracking single particles on supported lipid membranes: multimobility diffusion and nanoscopic confinement,” J. Phys. Chem. B 118(6), 1545–1554 (2014). [CrossRef]
33. H.-M. Wu, Y.-H. Lin, T.-C. Yen, and C.-L. Hsieh, “Nanoscopic substructures of raft-mimetic liquid-ordered membrane domains revealed by high-speed single-particle tracking,” Sci. Rep. 6(1), 20542 (2016). [CrossRef]
34. Y.-H. Liao, C.-H. Lin, C.-Y. Cheng, W. C. Wong, J.-Y. Juo, and C.-L. Hsieh, “Monovalent and oriented labeling of gold nanoprobes for the high-resolution tracking of a single-membrane molecule,” ACS Nano 13(10), 10918–10928 (2019). [CrossRef]
35. P. B. Johnson and R. W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]
36. R. Gholami Mahmoodabadi, R. W. Taylor, M. Kaller, S. Spindler, M. Mazaheri, K. Kasaian, and V. Sandoghdar, “Point spread function in interferometric scattering microscopy (iSCAT). Part I: aberrations in defocusing and axial localization,” Opt. Express 28(18), 25969–25988 (2020). [CrossRef]
37. L. A. Amos and W. B. Amos, “The bending of sliding microtubules imaged by confocal light microscopy and negative stain electron microscopy,” J. Cell Sci. 1991(Supplement_14), 95–101 (1991). [CrossRef]
38. P. Calzavara-Pinton, C. Longo, M. Venturini, R. Sala, and G. Pellacani, “Reflectance confocal microscopy for in vivo skin imaging,” Photochem. Photobiol. 84(6), 1421–1430 (2008). [CrossRef]
39. L. Priest, J. S. Peters, and P. Kukura, “Scattering-based light microscopy: from metal nanoparticles to single proteins,” Chem. Rev. 121(19), 11937–11970 (2021). [CrossRef]
40. O. V. Lyulko, G. Randers-Pehrson, and D. J. Brenner, “Simultaneous immersion Mirau interferometry,” Rev. Sci. Instrum. 84(5), 053701 (2013). [CrossRef]
41. A. Dubois, W. Xue, O. Levecq, P. Bulkin, A.-L. Coutrot, and J. Ogien, “Mirau-based line-field confocal optical coherence tomography,” Opt. Express 28(6), 7918–7927 (2020). [CrossRef]
42. Y.-F. Huang, G.-Y. Zhuo, C.-Y. Chou, C.-H. Lin, and C.-L. Hsieh, “Label-free, ultrahigh-speed, 3D observation of bidirectional and correlated intracellular cargo transport by coherent brightfield microscopy,” Nanoscale 9(19), 6567–6574 (2017). [CrossRef]
43. A.E. Dixon, S. Damaskinos, and M.R. Atkinson, “A scanning confocal microscope for transmission and reflection imaging,” Nature 351(6327), 551–553 (1991). [CrossRef]
