Open Access
Add to BibSonomy
Abstract
Biomedical researchers use optical coherence microscopy (OCM) for its high resolution in real-time label-free tomographic imaging. However, OCM lacks bioactivity-related functional contrast. We developed an OCM system that can measure changes in intracellular motility (indicating cellular process states) via pixel-wise calculations of intensity fluctuations from metabolic activity of intracellular components. To reduce image noise, the source spectrum is split into five using Gaussian windows with 50% of the full bandwidth. The technique verified that F-actin fiber inhibition by Y-27632 reduces intracellular motility. This finding could be used to search for other intracellular-motility-associated therapeutic strategies for cardiovascular diseases.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Intracellular motility plays a crucial role in various cellular processes, including the delivery of nutrients and signaling molecules, cell cycle differentiation, and pathological transformation [1–3]. As such, intracellular motility is an indicator of changes in the physiological and pathological characteristics of cellular processes. To improve the understanding of the factors that contribute to these cellular processes, two- and three-dimensional in vitro models that mimic in vivo cellular behavior have been developed and widely studied. However, observing changes in intracellular motility as they occur within cells remains a challenge.
Currently available tools used to observe changes in intracellular motility can be classified into two main categories: direct and indirect intervention techniques. Direct intervention techniques are methods in which molecular probes are directly introduced into cells to monitor changes in intracellular motility. One direct intervention technique uses micron-scale fluorescent microspheres, which are injected into the cytoplasm to extract its viscoelastic properties [2]. Another involves the use of molecular sensors to delineate malignant, differentiated, and apoptotic cancer cells; the sensors detect variations in the cellular viscoelastic properties by sensing changes in intracellular motility [4]. Although these methods have been successful in providing information about changes in intracellular motility, they may not be suitable for all types of biological samples [5]. Hence, it is desirable to have an indirect intervention technique that can measure changes in intracellular motility without directly intervening in normal cellular processes.
Optical coherence microscopy (OCM) is an indirect intervention technique that has the potential to successfully measure changes in intracellular motility. OCM is a version of optical coherence tomography (OCT) [6] that combines confocal gating with coherence gating to improve the lateral resolution, using a high-numerical-aperture objective lens without narrowing the depth of field to an extreme degree. OCM has been widely used in biomedical research because of its high spatial and temporal resolution and its real-time and label-free tomographic imaging capabilities. The label-free imaging capability of OCM originates in its ability to derive contrast from reflectance or refractive index variations within a sample. However, with samples having low reflectance or minute variations in refractive index, OCM tends to provide low contrast. Moreover, like OCT, OCM lacks the ability to provide bioactivity-associated functional contrast.
Scientific efforts to provide functional contrast in OCT systems began only a few years after the invention of the first OCT system. These efforts were directed toward providing OCT with a functionality that can both reveal the structures of blood vessels in the retina (up to capillary-length scales) and measure the blood flow within these vasculatures. OCT angiography (OCTA) [7,8] and Doppler OCT [9–11] were developed to provide these necessary functionalities. The concept behind OCTA is that the movements of red blood cells (RBCs) induce statistical changes in the intensity signal within a voxel, thereby allowing the possibility of distinguishing blood vessels from the surrounding static tissues [12]. In Doppler OCT, blood flow speed is measured by utilizing the Doppler principle to detect either spectral shifts or phase shifts [9]. Other research efforts were directed toward providing OCM with the functionality to compute the extracellular matrix deformations caused by cell-generated forces in 3D culture. With this functionality, OCM has been used successfully to delineate the dynamic forces exerted by NIH-3T3 fibroblasts [13]. In another study, OCM was provided with the functionality to measure the local axial strain of mechanically loaded tissues. Using OCM with this functionality, the mechanical heterogeneity of vascular smooth muscle cells and elastin sheets in freshly excised mouse aortas was revealed. This would have remained unresolved using typical lower-resolution optical coherence elastography systems [14].
Recently, research efforts have been directed toward providing OCT with the functionality to capture the dynamics of smaller scatterers present in constrained spaces, such as cell samples [15,16]. To provide this functionality to the OCT system, gold nanorods were utilized as diffusive probes to discriminate between various scatterers present in a sample. Both the time and polarization-dependent properties of the scatterers were utilized to demonstrate the functional capability of the OCT system to differentiate mammary epithelial cell (MEC) spheroids from the surrounding gold nanorods. This technique has succeeded in providing a more detailed MEC spheroid morphology, which is a hallmark of premalignancy in breast cancer cells [17]. However, the introduction of gold nanorods did not result in full utilization of the label-free imaging capability of OCT systems. To take full advantage of this capability, recent studies have provided functional contrast by analyzing the temporal fluctuations in signals caused by the metabolically driven activity of intracellular components at the subcellular-length scale. This technique was demonstrated using full-field optical coherence tomography (FFOCT) [18–20] and frequency-domain optical coherence tomography (FDOCT) [21,22]. Because these techniques are based on dynamic light scattering measurements, they are called dynamic OCT or dynamic contrast OCT [18]. These techniques have been successful in visualizing the temporal-dynamics contrast-based cellular structures of murine tongue and liver samples [21] and in performing reliable label-free segmentation of the layers in tissue samples [22] and cell cultures [23].
In this study, we extended these research efforts by providing a custom-made spectral-domain OCM system with the functionality for quantitatively measuring the changes in intracellular motility that occur when biochemical inhibitors are added to the external environment of the cells. We utilized this technique to elucidate the effects of Y-27632 [24,25]—a commonly used competitive antagonist for ATP in the kinase domain of the Rho/Rho-associated protein kinase (Rho/ROCK) pathway—on the intracellular motility of cardiac fibroblast cells. Rho/ROCK activity, which has a central role in the mechano-transduction of cells, has earned the spotlight in the urgent campaign to find a pharmacological therapy that can halt or reverse cardiac fibrosis, the major cause of cardiovascular disease–related deaths, because research suggests that increased Rho/ROCK activity may play a major role in the processes leading to cardiovascular diseases, such as hypertension, heart failure, and stroke [25]. Thus, Rho/ROCK activity has the potential to be classified as a biomarker for cardiovascular disease [26]. One possible approach for addressing the cardiovascular disease crisis is to identify inhibitors that could halt or reverse Rho/ROCK activity. Therefore, the method developed in this study may be useful as a tool in the search for activators and inhibitors of cellular processes, aiding in the treatment of various diseases.
2. Method
2.1 Live-cell spectral-domain optical coherence microscope
A custom-made spectral-domain optical coherence microscope (SD-OCM) system, the same one used in a previous study [27], was employed to collect cross-sectional images of cardiac fibroblast samples. A BX51 (Olympus, Japan) upright microscope serves as the skeleton of the system and is integrated with a two-photon microscope imaging system [28]. A schematic representation of the SD-OCM is shown in Fig. S1. A superluminescent diode (SLD) (T-850-HP-I, Superlum, USA) with a central wavelength of 850 ± 10 nm and a bandwidth of 165 nm serves as a light source and determines the axial resolution. With these specifications for the broad-bandwidth light source, an axial resolution of approximately 2.60 µm (in water) could be achieved experimentally. The spectrometer (Cobra-S, Wasatch Photonics, UT, USA), which has a 2048-pixel line-scan camera, provides an imaging depth of approximately 2 mm in air. The objective lens is a 20× water-immersion type (NA = 0.50; Olympus, Tokyo, Japan); However, by utilizing the underfilled back aperture of this lens, the SD-OCM system could achieve a transverse resolution of approximately 1.90 µm, which corresponds to 0.27 effective NA. With this transverse resolution, the depth of field is approximately 37.8 µm. Optical power was measured with power meter (C122C, Thorlabs, USA), as 2.55 mW after the objective lens. The transversal sampling rate was set between 600 and 900 points per scan, depends on experiment configuration. The A-scan rate was fixed to 100-k lines/sec, but due to the limited storage capacity and writing speed of the component, (NVMe interface solid-state disc, multiple vendors) 80 B-scans per second are set as default temporal sampling in this study.
Although the axial resolution of the SD-OCM system is improved by the utilization of a broad-bandwidth light source, the broader spectral bandwidth results in increased dispersion effects from the mismatch between the sample and reference arms. Hence, dispersion compensation is required to achieve the expected theoretical axial resolution [29,30]. Because it is more effective to compensate for the dispersion effects by directly assessing the image quality, a numerical dispersion algorithm based on optimizing the coefficient of variation for a cross-sectional image was applied to each set of images.
Moreover, because of the curvature of the optical path difference along the beam scanning or nonidealities in the imaging system design or component positioning, flat surfaces appear tilted or curved in the reconstructed images. The degree of tilting or curvature suffered by flat surfaces is not representative of the true structure of the biological sample. This erroneous representation of the true structure of the sample is called a coherence gate curvature (CGC) artifact. Because it could affect the accuracy of the insights inferred from the biological sample, it must be compensated for. Thus, before the dynamic contrast enhancement calculations are performed, CGC compensation is accomplished by performing a fourth-order polynomial fitting on a flat surface, such as the glass surface of a confocal dish. The fitted polynomial is used to correct the common geometry distortion.
To ensure thermal stabilization before acquisition of the images, stage-top heaters with humidified CO2 supply systems (Live Cell Imaging, South Korea) were integrated into the SD-OCM system’s stage, which is a custom-built three-axis mechanical stage with motorized actuators (LTA-HS, Newport, USA). This three-axis stage provides a three-dimensional degree of freedom for selecting an ideal location in the sample to be imaged. However, adjustments of the stage could potentially induce unwanted vibrations affecting the collected images; hence, after a sample was placed on the stage and an ideal location on the sample was identified for imaging, a wait time of at least 30 min was imposed to allow thermal and mechanical stabilization before image acquisition was performed. Images were acquired using a custom C++-based acquisition interface.
2.2 Temporal-dynamics contrast-based functional OCM
The functionality for quantitatively measuring changes in intracellular motility is provided by analyzing the changes to the temporal signal fluctuations caused by the scatterers present in a sample. To enable pixel-wise calculation of the temporal fluctuations in the signal, an M-scan is performed, acquiring a set of consecutive cross-sectional images from the same spatial region (Fig. 1(a)).
Fig. 1. (a) Schematic representation of the principle of temporal-dynamics contrast-based functional optical coherence microscopy (fOCM). (b) Schematic representation of split-spectrum-based fOCM. t : time domain, k : wavenumber domain.
The varied dynamics of the different scatterers in the sample causes varying fluctuations in the intensity signal, thereby allowing the differentiation of cells with high intracellular motility from the static scatterers. To quantify the fluctuations in the intensity signal, we began by assessing various functional OCM (fOCM) parameters that have previously been reported: standard deviation, motility [17], logarithmic intensity variance [31,32], and coefficient of variation. We assessed these candidate parameters for their capability of quantifying the minute fluctuations in the intensity signal, and after completing this assessment, we chose the motility parameter to use for the purposes of this study. Motility is an experimental measure of the intracellular motion arising from ATP-driven processes and suppresses information from the static parts of a sample [17].
In this study, the number M of cross-sectional images acquired from a single y-plane was set to 20. With the 80 B-scans per second of acquisition speeds, each set of M-scan records 0.25 seconds of dynamics per each y-plane. Ideally, it is desirable to acquire many cross-sectional images for the improvement in the signal-to-noise ratio (SNR) of the image that is reconstructed by averaging the M consecutive images. However, acquiring an extensively large number of images requires very high computer storage capacity, which could impose a significant burden, particularly for laboratories whose computer storage capacity is limited by cost or technical constraints. Hence, an optimal value for M was chosen to make maximum use of available computer storage capacity while still allowing the acquisition of images having sufficiently high-SNR intensity and motility. By performing an optimization calculation, we found that collecting 20 consecutive cross-sectional images from the same spatial location maximized our available storage capacity and provided images of sufficiently high quality for both intensity and motility (Fig. S2). However, it is advisable to take additional care when determining M values according to the cell type and physical condition of the sample. Even the same type of cells is used, the SNR estimated from the samples will be different if unclear criteria are used to determine signal and noise regions. All processing was performed using MATLAB R2021b (MathWorks, USA).
2.2.1 Split-spectrum-based fOCM
Because the shape of the SLD spectrum used in the SD-OCM system deviates from that of ideal Gaussian-function behavior, the experimental axial point spread function (PSF) suffers from sidelobe artifacts. If there is only a single scatterer in the sample, this is not a major concern. However, for samples with multiple scatterers, sidelobe artifacts significantly lower the image contrast and mask the intensity coming from scatterers with low reflectance or poor scattering properties, thereby potentially increasing the level of noise in the calculated temporal fluctuations in the intensity signal. Multi-apodization or spectral shaping techniques are typically used to address this problem. However, these techniques typically broaden the main lobe of the PSF and significantly reduce the axial resolution [33–35]. Hence, it is desirable to have a technique that can address this problem without overly compromising the axial resolution.
We consider a method in the field of OCTA that employs a split-spectrum technique to address the sidelobe artifact problem caused by the non-Gaussian shape of the source spectrum. This technique has been shown to improve flow-detection SNR, allowing the reconstruction of microcapillaries in the retina that were initially challenging to detect using conventional OCTA algorithms [36,37]. For this study, we adopted a split-spectrum image reconstruction technique wherein we split the full spectrum into smaller spectra using equally spaced Gaussian windows in the k-domain (Fig. 1(b)). Because this technique affords the possibility of generating multiple images from a single acquisition, the processed images are expected to have improved quality because of the enhancement of the intracellular-motility sensitivity and the suppression of background noise.
The motility images reconstructed from the individual narrower spectral bands are then averaged to fully utilize the speckle information in the entire spectrum. The averaged-motility parameter is given by
Because there is a trade-off between the enhancement in intracellular-motility sensitivity and axial resolution, the size of the Gaussian windows and the number of spectral splits must be carefully chosen. The size or bandwidth of the Gaussian windows is directly related to the axial resolution of the reconstructed OCM images. For this study, the bandwidth of the Gaussian windows was set to 50% of the full-spectrum bandwidth in order to maintain the resolution possible for cellular imaging. To determine the optimal number of spectral splits, we analyzed the image quality as a function of the number of spectral splits. We found that the image quality saturates to a certain value as the number of spectral splits is increased and that this saturation begins at five spectral splits (Fig. S3). Hence, we set five as the optimal number of spectral splits. All image processing procedures were performed using custom-written MATLAB script.
2.2.2 Preparation of phantom for comparison of split-spectrum and full-spectrum image reconstruction
To evaluate the effect of utilizing the split-spectrum image reconstruction technique compared with that of applying the full-spectrum technique, we applied both techniques to images acquired from a custom-made phantom containing vibrating magnetic particles that mimic the dynamics of the intracellular components contributing to the temporal fluctuations in the intensity signal. The vibration was controlled using a solenoid (10 mm in diameter) embedded in a polydimethylsiloxane (PDMS) block. Amino-functionalized magnetic particles (1.16 µm diameter, Spherotech, USA) were mixed with glycerin (Sam-Hyun Pharm, South Korea) at a 1:100 ratio; the mixture was then poured into a 3-mm hole in the PDMS block containing the solenoid. The vibration was induced on the solenoid using a function generator that produces a sine current having a 10-V peak-to-peak amplitude and a frequency of 100 Hz (Fig. 2(a)). From the intensity images collected for a given cross-sectional region, we confirmed that the magnetic particles were suspended throughout the glycerin solvent (Fig. 2(b)), thereby ensuring that the vibrations could be effectively induced on the particles.
Fig. 2. Validation of the advantage of the split-spectrum temporal-dynamics contrast-based fOCM. (a) Schematic representation of the phantom used for the validation experiment. (b) Log-transformed intensity image obtained for a cross-sectional region of the phantom. (c) Average normalized motility images of the phantom at 5-µm transversal thickness, processed using the full-spectrum technique (left) and split-spectrum technique (right). The motility values were fixed to the range [0, 0.5] arbitrary units (a.u.) for visualization purposes. Field of view: 150 × 225 (µm), Scalebar: 45 µm, FG: function generator; Sol: solenoid; MP: magnetic particle; FP: focal plane.
2.3 Sample preparation
The study used collagen gels and cardiac fibroblasts (CFs). Methacrylated collagen gels (PhotoCol, Advanced BioMatrix, USA) were prepared at a concentration of 4 mg/mL as static scatterers. To vary the stiffness of the gels, photo-crosslinking with UV irradiation was performed using a Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) photoinitiator.
Cardiac fibroblasts (5 × 104) were seeded on a confocal dish (35pi dish, Corning). To mitigate the intense specular reflection from the glass surface of the dish, a 6-kPa polyacrylamide gel with a coating of collagen approximately 100 µm in thickness were prepared on each confocal dish prior to cell seeding.
To vary the intracellular motility of the cardiac fibroblasts without changing the physical properties of the external environment of the cells, paraformaldehyde (PFA) and Y-27632 were used. PFA is commonly used in experiments involving cell fixation, a vital procedure in biology laboratories. Y-27632 is a well-known Rho/ROCK pathway inhibitor. In this experiment, 4% PFA and 50 µM Y-27632 were used.
2.4 Fluorescence staining and morphological analysis
Phalloidin was used to stain the distribution of F-actin fibers in the Y-27632-treated CFs. Fluorescence images were acquired with fluorescence microscope (Nikon Eclipse Ti-2, Japan). Qualitative directionality analysis of the F-actin images was performed using OrientationJ [38] with ImageJ [39] (NIH, USA). Local directionality was then quantified by measuring the texture polarizability at the center of randomly selected cell bodies (n = 50) using a custom-written MATLAB script. The local directionality of each cell was regularized by matching the direction of the principal axis with that of the x-axis before averaging.
3. Results
3.1 Comparison of split-spectrum and full-spectrum image reconstruction
We used both the split-spectrum method and the full-spectrum method to obtain motility images for a 20-µm-wide cross-sectional region centered on the focal plane of the prepared phantom (Fig. 2(b)). We found that the mean intensity using the full-spectrum method was 0.74 arbitrary units (a.u.), whereas that using the split-spectrum method was 0.51 (a.u.). This is a reasonable result because the split-spectrum method utilizes only a fraction of the bandwidth in the k-domain; hence, the intensity level of the processed image is expected to be less. Because the fOCM motility parameter is used to calculate the temporal fluctuations in the signal from the intensity information, it is predicted to follow the same behavior. Also, the motility value of a scatterer is affected by its position relative to the focal plane, even if it has the same temporal dynamics.
However, the key result from this experiment is the major difference between the functional images processed using the full- and split-spectrum reconstruction techniques. As shown in Fig. 2(c), use of the split-spectrum technique substantially reduced the noise level in the processed functional images.
In the phantom, the magnetic beads were suspended throughout the glycerin. Because glycerin is optically transparent in this system, the motility values derived from the glycerin-only region can be classified as the effect of noise. We found that the mean and standard deviation of the noise-induced motility using the full-spectrum method were 0.166 ± 0.163, which is even higher than the mean motility values of the feature (i.e., magnetic particle scatterers), at 0.161 ± 0.029. The feature-induced motility of split-spectrum method could have even lower values than the background motility originated from the statistical noise. This makes it difficult to distinguish the fine motility signals without prior knowledge of the object's geometry.
However, using the split-spectrum method, it involves averaging of motility images formed at different spectral bandwidth. This resulted as a washout of statistical noise but maintain feature-induced motility. As a result, the sensitivity of the minute motility are enhanced by lowering the noise induced motility. The mean and standard deviation of the noise-induced motility using the split-spectrum method (using five smaller spectra with Gaussian windows of 50% of the full-spectrum bandwidth) were 0.039 ± 0.006, and the mean motility value of the magnetic particles was 0.073 ± 0.045. Although the motility values for both the background noise and the magnetic particles were reduced, those from the noise region were decreased much more substantially. This considerable reduction in the noise information from the background makes it possible to visualize and quantify the temporal fluctuations in the signal caused by the vibration of the magnetic particles. This result suggests that the split-spectrum image reconstruction technique is advantageous in that it can reveal weaker signals coming from cells, which might otherwise be masked by noise.
3.2 Feasibility of using temporal-dynamics contrast-based fOCM for the targeted task
To validate the capability of the temporal-dynamics contrast-based fOCM method utilizing the split-spectrum image reconstruction technique to quantify the minute temporal fluctuations in the intensity signal caused by variations in the conditions of both static and motile scatterers, we applied the technique to two sets of samples: (1) collagen gels of differing stiffness and (2) cardiac fibroblast cells preserved using PFA. The collagen gels of differing stiffness were used to understand how variations in the conditions of static scatterers affect the temporal fluctuations in the intensity signal. The cardiac fibroblast cells preserved with PFA were used to investigate the feasibility of using the temporal-dynamics contrast-based fOCM method to accurately quantify the changes in intracellular motility that occur when intracellular components are cross-linked together.
3.2.1 Results using collagen gels of differing stiffness
Cells are three-dimensionally confined by an extracellular matrix (ECM), which is composed primarily of collagen fibers [40]. As our main interest was in quantifying the changes in intracellular motility that occur when the external environment of the cells is perturbed, we assessed how variations in the conditions of the surrounding ECM, a static scatterer, might affect the temporal fluctuations in the intensity signal. To mimic the in vivo ECM that confines the cells, gels were fabricated that had homogeneous reflectance but differed in stiffness (Fig. 3(a), left). Utilizing temporal-dynamics contrast-based fOCM, we found that both the intensity and the fOCM parameter, motility, are independent of the photo-crosslinking status of the collagen fibers (Fig. 3(b)). This result is important because it indicates that variations in the characteristics of static scatterers do not contribute significantly to temporal fluctuations in the intensity signal. Hence, it is established that any changes in the temporal fluctuations in the intensity signal are due primarily to changes in intracellular motility.
Fig. 3. (a) Schematic representation of samples used in the feasibility assessment of temporal-dynamics contrast-based fOCM method for the targeted task: collagen gels of differing stiffness (left) and cardiac fibroblasts preserved using paraformaldehyde (PFA) (right). PAA: polyacrylamide. (b) Temporal dynamics of methacrylated collagen gels with and without UV-induced cross-linking. Both intensity and motility are independent of the photo-crosslinking status of the methacrylated collagen fibers. ns: no significant difference. Box is 25%-75% interval and whisker are min-max values. (c) En face intensity and motility images of cardiac fibroblast cells before and after cell fixation. The areas in the green and red boxes indicate a selected cell body and are shown enlarged at right. Field of view: 850 µm, Enlarged area: 170 µm, Scalebar: 170 µm. (d) Histograms of the pixel values of intensity and motility for the selected cell body.
3.2.2 Results using cells fixed with PFA
Paraformaldehyde (PFA) is an excellent agent for preserving the morphological integrity of cell cultures [41]. We chose cells fixed with 4% PFA to investigate the feasibility of using the temporal-dynamics contrast-based fOCM method to accurately quantify the changes in intracellular motility that occur when all cell dynamics are halted.
In this experiment, cardiac fibroblasts (CFs) were cultured on a planar confocal dish with collagen-coated polyacrylamide (PAA) gel (Fig. 3(a), right). The PAA gel both served as a mechanical environment for the cells and acted as a barrier against the intense specular reflection from the glass surface of the confocal dish. After a sample was mounted on the stage, a set of images of the live CFs was collected. Cell fixation with PFA was performed in situ using a custom-made stage-top media exchange system, thereby ensuring that the succeeding images acquired were of the same location. After the addition of PFA, a further wait time of 1 hr was imposed before another set of images was taken, to allow time for the added PFA to cross-link the proteins or cytoskeletons in the CFs. After processing the sets of images, we confirmed that the temporal-dynamics contrast-based fOCM method was able to quantitatively measure changes in intracellular motility that could not be readily inferred from the conventional OCM intensity images (Fig. 3(c, d)).
The mean values in the histograms of intensity before and after the addition of PFA are 65.57 and 62.26 (a.u.), respectively, a difference of only approximately 5%. On the other hand, the mean values in the histograms of motility are 3.12 and 2.25 (a.u.), respectively, a decrease of ∼28%, which is substantial. This result successfully demonstrates the feasibility of using temporal-dynamics contrast-based fOCM as a method for quantifying changes in intracellular motility. It is also important to note that this technique effectively captures the blebbing phenomenon that arises after the addition of PFA [42] as a morphological feature of CFs undergoing apoptosis. The blebs appear as circular features in the motility en face images (Fig. 3(c): Motility After PFA). Interestingly, both full-spectrum-based motility and split-spectrum-based motility can capture the significant difference of the temporal dynamics after PFA treatment. It seems that the all-or-none changes in the metabolic state after cell fixation is too drastic to emphasis the efficacy.
3.3 Rho/ROCK-inhibitor-induced functional contrast
To verify that the changes in intracellular motility within the cytoplasm cause temporal fluctuations in the intensity signal, we applied the temporal-dynamics contrast-based fOCM method to cardiac fibroblast cells treated with Y-27632.
We cultured cardiac fibroblast (CF) cells with a setup that mimicked the inhibition of Rho/ROCK activity. After seeding the CFs on a planar confocal dish with collagen-coated PAA gel, we allowed the CFs a day to proliferate in the dish before performing the image acquisition. During the image acquisition, we allocated a period of 1 hr for stabilization after the mounting of the sample on the stage before collecting the set of images. After the control images were acquired, the media exchange system was used to remove the medium from the sample and to add the medium containing Y-27632. Then, a further wait time of 4 hrs was imposed to allow time for the added Y-27632 to inhibit the Rho/ROCK pathway.
Using the temporal-dynamics contrast-based fOCM method, we confirmed that the inhibition of Rho/ROCK activity leads to a reduction in intracellular motility within the cytoplasm. The decrease in intracellular motility causes minimal change to the temporal fluctuations in the intensity signal and thereby also significantly decreases the motility values (Fig. 4). Especially, split-spectrum based motility showed enhancement in intracellular motility changes. Figure 4(a) represents the pixelwise ratio of normalized Split-spectrum based motility (MotS) and Full-spectrum based motility (MotF). This revealed that the MotS in the living cell area exhibited a higher signal value (MotS / MotF > 1, indicated in red), and a lower value in a cell-free region (MotS / MotF < 1, indicated in blue). This indicates that the MotS has improved sensitivity to metabolism and suppressed noise. Figure 4(b) shows histograms of the MotS and MotF before and after the treatment of Y-27632. The shift of the mode value after the treatment, which could not be distinguished using the MotF, can be clearly distinguished by the MotS. This result supports the improved sensitivity of MotF to the changes in cellular metabolism due to Rho/ROCK activity.
Fig. 4. Relationship between inhibition of the Rho/ROCK pathway by Y-27632 and the reduction in intracellular motility in the cytoplasm of cardiac fibroblast cells. (a) Pixelwise ratio-map between the normalized split- (MotS) and full- (MotF) spectrum motility to show the efficacy of the split-spectrum based motility. (b) Histogram of motility images. Split-spectrum based motility showed the left-shift of mode-peak which suggests the enhanced sensitivity of the split-spectrum based motility. (c) En face intensity and motility images of cardiac fibroblasts before and after addition of Y-27632. The enlargements of the green- and red-boxed areas show the difference between the control and the Y-27632-treated cardiac fibroblasts at the single-cell level. Field of view: 850 µm, Enlarged area: 300 µm, Scalebar: 300 µm. (d) Histograms of the intensity and motility for the representative single-cell region. (e) Paired plot for individual cells (n = 8) before and after addition of Y-27632. ****: p < 0.0001. (f) Paired plot for the areas of the cells (n = 19) before and after addition of Y-27632. *: p = 0.0320.
This relationship between the inhibition of Rho/ROCK activity and the reduction in intracellular motility in the cytoplasm could not be deduced merely by analyzing the intensity images (Fig. 4(e)). Moreover, inhibition of the Rho/ROCK pathway in CFs has distinct effects on cellular morphology. Using the segmented intensity images, we were able to track the change in the areas of individual cells that occurred upon the addition of Y-27632. As shown in Fig. 4(f), the cell areas in the Y-27632-treated CFs were substantially less than those in the controls.
To demonstrate that the observed relationship between the inhibition of ROCK activity and the reduction in intracellular motility in the cytoplasm was caused by the treatment with Y-27632, we captured F-actin-stained fluorescence images of control and Y-27632-treated CFs. The Y-27632-treated cells showed morphological changes such as elongated protrusions and reduced cytoplasmic regions, which is consistent with the results of previous studies involving treatment of cells with Y-27632 [43–45].
The F-actin fibers that were initially present in the control samples decreased in area after the treatment with Y-27632. The reduction in the distribution of F-actin fibers in the Y-27632-treated CFs was visualized via a directionality analysis of the entire image (OrientationJ, ImageJ, NIH, USA). Then, the local directionality was quantified via an analysis of the centers of randomly selected cell bodies (n = 50) using a custom-written MATLAB script. The local directionality of each cell was regularized by matching the direction of the principal axis with that of the x-axis before averaging. As shown in Fig. 5(b), we found that the ratios between the major polarized axis (a) and the minor polarized axis (b) for the control and Y-27632-treated CFs were 2.09 and 1.67, respectively. This confirms that the linear alignment of the F-actin fibers was reduced by approximately 20%.
Fig. 5. (a) F-actin fibers of control and Y-27632-treated cardiac fibroblast cells with corresponding directionality maps generated using OrientationJ (ImageJ). (b) Polar plots of angular directionality and the corresponding average normalized plot from the centers of randomly selected cell bodies (n = 50). a and b correspond to the major and minor axes of texture polarity, respectively. Under treatment with Y-27632, the polarity was reduced by approximately 20%. Field of view: 590 µm, Scalebar: 120 µm.
4. Discussion
The dynamics of cytoplasmic components is a leading source of parameters for representing processes in cellular physiology such as differentiation, migration, and apoptosis. The measurement of intracellular motility to monitor cytoplasm dynamics has enabled a better understanding of cellular functions. In this study, we developed a functional optical coherence microscopy (fOCM) imaging device for measuring the dynamics of intracellular activity. To measure temporal dynamics, an M-scan is performed to acquire a set of consecutive images for the same cross-sectional area, and the motility is calculated. To enhance the contrast of the images, a split-spectrum method is applied to reduce the noise in the static region, thereby revealing the weaker signals from cells. The split-spectrum image reconstruction method has allowed us to accurately measure the intracellular dynamics and deduce relevant information regarding cellular physiology. Using this fOCM method, we observed changes in the cytoplasmic motility of cardiac fibroblasts treated with modulators of cellular functions such as cellular metabolism and cytoskeletal dynamics.
In this study, we investigated Rho/ROCK activity in particular because its activation plays a significant role in processes leading to cardiovascular disease. Thus, having a tool to search for inhibitors of Rho/ROCK activation could help in the quest to understand and reduce the incidence of cardiovascular disease. Y-27632, a commonly used Rho/ROCK pathway inhibitor, is a competitive antagonist for ATP in the kinase domain [25,26]. It induces a decrease in ATP-driven cytoskeletal turnover. Using the fOCM method developed in this study, we have shown that the activation of Rho/ROCK activity by Y-27632 is associated with a decrease in intracellular motility. In analyzing the morphological characteristics of the Y-27632-treated cells, we observed a decrease in individual cell size and elongation of the cell body into a polarized star-shaped form. Moreover, we observed a decrease in the development of F-actin stress fibers and in the polarization of the fiber network (Fig. 5) in the center of the cell body. These observations are consistent with recent reports [46,47] that link the physical properties of cytoplasm to cellular morphology. Thus, we have demonstrated that this label-free fOCM method is highly capable of revealing the effects of a Rho/ROCK-pathway inhibitor on the intracellular dynamics of cardiac fibroblasts (which in previous studies required the use of chemical or fluorescent labeling [43–45]): increased overall cell proliferation, decreased size of individual cells, and decreased formation of F-actin stress fibers. Therefore, the fOCM method described herein has potential for use as a tool for monitoring cellular cytoplasm motility, which reflects disease states or the pathological status of cells and tissues.
In this study, we present a method suitable for use in wet lab settings for measuring the temporal dynamics change of a single kind of cell. By acquiring a minimum of 20 images, this method allows for the statistical evaluation of dynamic contrast changes. However, it is important to note that the protocols for this method may need to be modified depending on the specific sample and goals of the experiment. For example, previous research [20] has demonstrated the effectiveness of dynamic contrast to visualize organoids, which are groups of differentiated cells. Each differentiated cell type exhibits unique temporal dynamics, requiring the use of multiple metrics for accurate visualization. The aforementioned study suggests extracting multiple dynamic contrast-related metrics from 512 consecutive images and mapping them in the color space (hue, saturation, and value).
In our method, it may be necessary to adjust imaging parameters, such as the number of consecutive scans (M-number) or the number of split-spectrum (N-number) based on the type of sample being used, to maintain the sensitivity for the temporal dynamics. To ensure the comparability of results across experiments, it is desirable to use the same image acquisition parameters in all related experiments.
This study has demonstrated the potential of functional OCM as a label-free tool to measure changes in intracellular motility upon treatment with chemical or biological agents in a 2D cell culture model. However, because there remains a plethora of physiological information that cannot be provided by 2D cell models, current research efforts in bioengineering are focused on studying multicellular, structured 3D culture models that are similar to the in vivo environment. Therefore, the development of advanced OCM systems that extend the depth of focus (DoF) while maintaining sharp lateral resolution (high NA) is mandatory. To address this problem, various approaches such as aperture engineering [48], computational adaptive optics [49], and machine learning [50] have been proposed. By extending the depth of field while maintaining its lateral resolution and functional contrast image acquisition capabilities, OCM may also be a promising tool for the long-term, label-free monitoring of intracellular dynamics in 3D biomimetic tissues with very little risk of causing damage to the cells.
5. Conclusion
We have developed a new approach for quantifying changes in intracellular behavior as they occur within the cells. To the best of our knowledge, this study is the first to comprehensively demonstrate and explicate the potential of a split-spectrum temporal-dynamics contrast-based functional OCM technique as a method for quantitatively monitoring the changes in intracellular behavior that occur when the external environment of the cells is perturbed. This method quantifies the changes in intracellular motility by performing pixel-wise calculations of the temporal fluctuations in the intensity signal. Moreover, when the method is utilized to elucidate the effect of a Rho/ROCK inhibitor on intracellular motility, a reduction in intracellular motility within the cytoplasm is observed. The relationship between the inhibition of the Rho/ROCK pathway and the decrease in intracellular motility in the cytoplasm could be exploited to monitor other inhibitors related to ATP-based metabolic activity. We have shown, therefore, that this label-free, non-toxic technique could be suitable for use in many biology laboratories to study changes in intracellular motilities as they occur within cells.
Funding
Basic Science Research Program through the National Research Foundation (NRF) of Korea, funded by the Ministry of Education (NRF-2021R1I1A1A01048370); Creative Materials Discovery Program through the National Research Foundation (NRF) of Korea Ministry of Science and ICT, South Korea (NRF-2019M3D1A1078940).
Acknowledgments
R. J. E. Canoy is supported by a Global Korea Scholarship. We thank Dr. Kwanjun Park at the Korea University Biophotonics Imaging Lab for valuable discussions.
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. J. F. Kelleher and M. A. Titus, “Intracellular motility: how can we all work together?” Curr. Biol. 8(11), R394–R397 (1998). [CrossRef]
2. T. P. Kole, Y. Tseng, I. Jiang, J. L. Katz, and D. Wirtz, “Intracellular mechanics of migrating fibroblasts,” Mol Biol Cell 16(1), 328–338 (2005). [CrossRef]
3. C. P. Brangwynne, G. H. Koenderink, F. C. MacKintosh, and D. A. Weitz, “Intracellular transport by active diffusion,” Trends Cell Biol. 19(9), 423–427 (2009). [CrossRef]
4. J. Mayank, A. Sindhu, N. Singh, N. Nayak, N. Garg, N. Kaur, and Singh, “Excited-state intramolecular hydrogen-bonding-assisted restricted rotation: a mechanism for monitoring intracellular viscosity and distinguishing malignant, differentiating, and apoptotic cancer cells,” ACS Appl. Bio Mater. 4(10), 7532–7541 (2021). [CrossRef]
5. E. O. Puchkov, “Intracellular viscosity: Methods of measurement and role in metabolism,” Biochem. Moscow Suppl. Ser. A 7(4), 270–279 (2013). [CrossRef]
6. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, “Optical coherence tomography,” Science 254(5035), 1178–1181 (1991). [CrossRef]
7. S. Makita, Y. Hong, M. Yamanari, T. Yatagai, and Y. Yasuno, “Optical coherence angiography,” Opt. Express 14(17), 7821–7840 (2006). [CrossRef]
8. L. An and R. K. Wang, “In vivo volumetric imaging of vascular perfusion within human retina and choroids with optical micro-angiography,” Opt. Express 16(15), 11438–11452 (2008). [CrossRef]
9. R. A. Leitgeb, R. M. Werkmeister, C. Blatter, and L. Schmetterer, “Doppler optical coherence tomography,” Prog. Retinal Eye Res. 41, 26–43 (2014). [CrossRef]
10. Z. Chen, T. E. Milner, S. Srinivas, X. Wang, A. Malekafzali, M. J. van Gemert, and J. S. Nelson, “Noninvasive imaging of in vivo blood flow velocity using optical Doppler tomography,” Opt. Lett. 22(14), 1119–1121 (1997). [CrossRef]
11. Y. Zhao, Z. Chen, C. Saxer, S. Xiang, J. F. de Boer, and J. S. Nelson, “Phase-resolved optical coherence tomography and optical Doppler tomography for imaging blood flow in human skin with fast scanning speed and high velocity sensitivity,” Opt. Lett. 25(2), 114–116 (2000). [CrossRef]
12. Y. Li, J. Chen, and Z. Chen, “Advances in Doppler optical coherence tomography and angiography,” Transl. Biophotonics1(1-2), (2019).
13. J. A. Mulligan, F. Bordeleau, C. A. Reinhart-King, and S. G. Adie, “Measurement of dynamic cell-induced 3D displacement fields in vitro for traction force optical coherence microscopy,” Biomed. Opt. Express 8(2), 1152–1171 (2017). [CrossRef]
14. A. Curatolo, M. Villiger, D. Lorenser, P. Wijesinghe, A. Fritz, B. F. Kennedy, and D. D. Sampson, “Ultrahigh-resolution optical coherence elastography,” Opt. Lett. 41(1), 21–24 (2016). [CrossRef]
15. G. Farhat, A. Mariampillai, V. X. Yang, G. J. Czarnota, and M. C. Kolios, “Detecting apoptosis using dynamic light scattering with optical coherence tomography,” J. Biomed. Opt. 16(7), 070505 (2011). [CrossRef]
16. J. Lee, W. Wu, J. Y. Jiang, B. Zhu, and D. A. Boas, “Dynamic light scattering optical coherence tomography,” Opt. Express 20(20), 22262–22277 (2012). [CrossRef]
17. A. L. Oldenburg, R. K. Chhetri, J. M. Cooper, W. C. Wu, M. A. Troester, and J. B. Tracy, “Motility-, autocorrelation-, and polarization-sensitive optical coherence tomography discriminates cells and gold nanorods within 3D tissue cultures,” Opt. Lett. 38(15), 2923–2926 (2013). [CrossRef]
18. 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]
19. C. E. Leroux, F. Bertillot, O. Thouvenin, and A. C. Boccara, “Intracellular dynamics measurements with full field optical coherence tomography suggest hindering effect of actomyosin contractility on organelle transport,” Biomed. Opt. Express 7(11), 4501–4513 (2016). [CrossRef]
20. 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]
21. M. Munter, M. Vom Endt, M. Pieper, M. Casper, M. Ahrens, T. Kohlfaerber, R. Rahmanzadeh, P. Konig, G. Huttmann, and H. Schulz-Hildebrandt, “Dynamic contrast in scanning microscopic OCT,” Opt. Lett.x 45(17), 4766–4769 (2020). [CrossRef]
22. H. M. Leung, M. L. Wang, H. Osman, E. Abouei, C. MacAulay, M. Follen, J. A. Gardecki, and G. J. Tearney, “Imaging intracellular motion with dynamic micro-optical coherence tomography,” Biomed. Opt. Express 11(5), 2768–2778 (2020). [CrossRef]
23. J. S. Lee, K. Eom, C. Polucha, and J. Lee, “Standard-unit measurement of cellular viability using dynamic light scattering optical coherence microscopy,” Biomed. Opt. Express 9(11), 5227–5239 (2018). [CrossRef]
24. S. Narumiya, T. Ishizaki, and M. Ufhata, “Use and properties of ROCK-specific inhibitor Y-27632,” in Methods in Enzymology, W. E. Balch, C. J. Der, and A. Hall, eds. (Academic Press, 2000), pp. 273–284.
25. M. Uehata, T. Ishizaki, H. Satoh, T. Ono, T. Kawahara, T. Morishita, H. Tamakawa, K. Yamagami, J. Inui, M. Maekawa, and S. Narumiya, “Calcium sensitization of smooth muscle mediated by a Rho-associated protein kinase in hypertension,” Nature 389(6654), 990–994 (1997). [CrossRef]
26. S. Hartmann, A. J. Ridley, and S. Lutz, “The Function of Rho-Associated Kinases ROCK1 and ROCK2 in the Pathogenesis of Cardiovascular Disease,” Front. Pharmacol. 6, 276 (2015). [CrossRef]
27. Y. G. Kang, H. Jang, Y. Park, and B. M. Kim, “Development of a 3-D Physical Dynamics Monitoring System Using OCM with DVC for Quantification of Sprouting Endothelial Cells Interacting with a Collagen Matrix,” Materials 13(12), 2693 (2020). [CrossRef]
28. Y. G. Kang, H. Jang, T. D. Yang, J. Notbohm, Y. Choi, Y. Park, and B. M. Kim, “Quantification of focal adhesion dynamics of cell movement based on cell-induced collagen matrix deformation using second-harmonic generation microscopy,” J. Biomed. Opt. 23(06), 1–8 (2018). [CrossRef]
29. M. Wojtkowski, V. Srinivasan, T. Ko, J. Fujimoto, A. Kowalczyk, and J. Duker, “Ultrahigh-resolution, high-speed, Fourier domain optical coherence tomography and methods for dispersion compensation,” Opt. Express 12(11), 2404–2422 (2004). [CrossRef]
30. G. Ni, J. Zhang, L. Liu, X. Wang, X. Du, J. Liu, and Y. Liu, “Detection and compensation of dispersion mismatch for frequency-domain optical coherence tomography based on A-scan's spectrogram,” Opt. Expressx 28(13), 19229–19241 (2020). [CrossRef]
31. I. Abd El-Sadek, A. Miyazawa, L. Tzu-Wei Shen, S. Makita, S. Fukuda, T. Yamashita, Y. Oka, P. Mukherjee, S. Matsusaka, T. Oshika, H. Kano, and Y. Yasuno, “Optical coherence tomography-based tissue dynamics imaging for longitudinal and drug response evaluation of tumor spheroids,” Biomed. Opt. Express 11(11), 6231–6248 (2020). [CrossRef]
32. P. Mukherjee, A. Miyazawa, S. Fukuda, T. Yamashita, D. Lukmanto, K. Okada, I. A. El-Sadek, L. Zhu, S. Makita, T. Oshika, and Y. Yasuno, “Label-free functional and structural imaging of liver microvascular complex in mice by Jones matrix optical coherence tomography,” Sci Rep 11(1), 20054 (2021). [CrossRef]
33. A. C. Akcay, J. P. Rolland, and J. M. Eichenholz, “Spectral shaping to improve the point spread function in optical coherence tomography,” Opt. Lett. 28(20), 1921–1923 (2003). [CrossRef]
34. Y. Chen, J. Fingler, and S. E. Fraser, “Multi-shaping technique reduces sidelobe magnitude in optical coherence tomography,” Biomed. Opt. Express 8(11), 5267–5281 (2017). [CrossRef]
35. R. Tripathi, N. Nassif, J. S. Nelson, B. H. Park, and J. F. de Boer, “Spectral shaping for non-Gaussian source spectra in optical coherence tomography,” Opt. Lett. 27(6), 406–408 (2002). [CrossRef]
36. Y. Jia, O. Tan, J. Tokayer, B. Potsaid, Y. Wang, J. J. Liu, M. F. Kraus, H. Subhash, J. G. Fujimoto, J. Hornegger, and D. Huang, “Split-spectrum amplitude-decorrelation angiography with optical coherence tomography,” Opt. Express 20(4), 4710–4725 (2012). [CrossRef]
37. S. S. Gao, G. Liu, D. Huang, and Y. Jia, “Optimization of the split-spectrum amplitude-decorrelation angiography algorithm on a spectral optical coherence tomography system: erratum,” Opt. Lett. 41(3), 496 (2016). [CrossRef]
38. D. Sage, “OrientationJ : A series of ImageJ plugins for directional image analysis,” (2020).
39. 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(7), 676–682 (2012). [CrossRef]
40. B. Sun, “The mechanics of fibrillar collagen extracellular matrix,” Cell Rep. Phys. Sci. 2(8), 100515 (2021). [CrossRef]
41. M. Srinivasan, D. Sedmak, and S. Jewell, “Effect of fixatives and tissue processing on the content and integrity of nucleic acids,” Am. J. Pathol. 161(6), 1961–1971 (2002). [CrossRef]
42. A. J. Hobro and N. I. Smith, “An evaluation of fixation methods: Spatial and compositional cellular changes observed by Raman imaging,” Vib. Spectrosc. 91, 31–45 (2017). [CrossRef]
43. G. S. Peh, K. Adnan, B. L. George, H. P. Ang, X. Y. Seah, D. T. Tan, and J. S. Mehta, “The effects of Rho-associated kinase inhibitor Y-27632 on primary human corneal endothelial cells propagated using a dual media approach,” Sci. Rep. 5(1), 9167 (2015). [CrossRef]
44. T. Omelchenko, J. M. Vasiliev, I. M. Gelfand, H. H. Feder, and E. M. Bonder, “Mechanisms of polarization of the shape of fibroblasts and epitheliocytes: Separation of the roles of microtubules and Rho-dependent actin-myosin contractility,” Proc. Natl. Acad. Sci. U.S.A. 99(16), 10452–10457 (2002). [CrossRef]
45. C. Ramachandran, R. V. Patil, K. Combrink, N. A. Sharif, and S. P. Srinivas, “Rho-Rho kinase pathway in the actomyosin contraction and cell-matrix adhesion in immortalized human trabecular meshwork cells,” Mol Vis 17, 1877–1890 (2011).
46. A. T. Molines, J. Lemiere, M. Gazzola, I. E. Steinmark, C. H. Edrington, C. T. Hsu, P. Real-Calderon, K. Suhling, G. Goshima, L. J. Holt, M. Thery, G. J. Brouhard, and F. Chang, “Physical properties of the cytoplasm modulate the rates of microtubule polymerization and depolymerization,” Dev. Cell 57(4), 466–479 (2022). [CrossRef]
47. M. D. Betterton, “A new view of how cytoplasmic viscosity affects microtubule dynamics,” Dev. Cell 57(4), 419–420 (2022). [CrossRef]
48. J. Mo, M. de Groot, and J. F. de Boer, “Focus-extension by depth-encoded synthetic aperture in optical coherence tomography,” Opt. Express 21(8), 10048–10061 (2013). [CrossRef]
49. M. Wu, D. M. Small, N. Nishimura, and S. G. Adie, “Computed optical coherence microscopy of mouse brain ex vivo,” J. Biomed. Opt. 24(11), 1–18 (2019). [CrossRef]
50. Z. Yuan, D. Yang, Z. Yang, J. Zhao, and Y. Liang, “Digital refocusing based on deep learning in optical coherence tomography,” Biomed. Opt. Express 13(5), 3005–3020 (2022). [CrossRef]
