Open Access
Add to BibSonomy
Abstract
Imaging cells in their 3D environment with molecular specificity is important to cell biology study. Widely used microscopy techniques, such as confocal microscopy, have limited imaging depth when probing cells in optically scattering media. Optical coherence tomography (OCT) can provide millimeter-level depth for the imaging of highly scattered media but lacks the contrast to distinguish cells from the extracellular matrix or to distinguish between different types of cells. Photothermal OCT (PT-OCT) is a promising technique to obtain molecular contrast at the imaging scale of OCT. Here, we report PT-OCT imaging of live, nanoparticle-labeled cells in 3D. In particular, we demonstrate detection and mapping of a single cell in 3D without causing cell death, and show the feasibility of 3D cell mapping through optically scattering media. This work presents live cell detection and mapping at an imaging scale that complements the major microscopy techniques, which is potentially useful to study cells in their 3D native or culture environment.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Cells naturally live in 3D environment, where they interact with other cells and the extracellular components in all directions to regulate the development and physiology of organism and the function of individual cells [1]. Cumulative evidence has indicated the essence of cell-cell and cell-extracellular matrix (ECM) interactions in the phenotypic expression and function of cells [2,3], yet many cell studies are limited to 2D planar culture, suboptimal for understanding the in vivo process and mechanism. In recognition of this limitation, 3D cell culture techniques are established to better emulate in vivo systems [4,5], and intravital imaging techniques are also developed to directly acquire the in vivo information [6,7]. However, it remains challenging to perform high-resolution imaging of targeted cells in the 3D culture media or tissue environment that are optically scattering [8,9], which is a significant hurdle for studying cellular activities in 3D environment.
Traditional 2D optical imaging methods do not provide quantitative assessment of cells in 3D environment due to the lack of depth resolvability. Optical sectioning by tightly focusing, rejecting out-of-focus light and scanning the focused beam over depth enabled 3D cell imaging with a micron- or submicron-level resolution and has been driving discoveries in cell biology. Relevant imaging modalities, such as confocal laser scanning microscopy, are widely used for studies of cells in relatively transparent samples [10]. In highly scattering media, however, it is largely difficult to image over hundreds of microns in depth using confocal microscopy [11]. Optical coherence tomography (OCT) employs coherence gating for axial sectioning and has a typical imaging depth at the millimeter level with a micro-scale resolution [12]. This imaging scale is complementary to confocal techniques, suitable to probe cells deeper inside the sample and important for studying a large number of cells within their 3D environment. Historically, OCT has been primarily used for imaging tissues [13], with notable success in high-resolution characterizations of both tissue structure and function. While traditional OCT structure imaging can be used for 3D cell tracking [14], additional contrast is required to distinguish cells from scattering environment and to differentiate cell types. One emerging method is dynamic OCT, where active transport in cells introduces rapid variations of OCT signal over time that can be quantified to highlight cells from tissue environment [15]. However, the feasibility of getting imaging contrast between different types of cells or performing targeted cell imaging is unclear and remains to be further explored. Taking advantage of molecular signatures of different cells, OCT-based molecular imaging could potentially provide robust imaging contrast to detect and map targeted cells in 3D [16,17].
Photothermal OCT (PT-OCT) is a promising approach for molecular imaging [18,19]. In general, PT-OCT employs a pump light to modulate the temperature surrounding specific light absorbers through photothermal effect and probes the resulted optical pathlength change using the OCT phase signal [20,21]. With absorption as the basic contrast mechanism, PT-OCT relies on exogenous absorbers for imaging or targets endogenous absorbers. Nanoparticles have been used as the primary exogenous agents due to their tunable, high extinctions from a wide range of wavelengths. Although most studies have focused on probing nanoparticles as the final target in tissue [22–25], the possibility of using nanoparticles to label specific biomolecules provides the potential for molecular imaging. For example, Gordon et al. recently demonstrated the use of PT-OCT to image functionalized gold nanorods that target ICAM2, a tissue protein, in the mouse retina and showed the molecular response to anti-VEGF treatment [26]. In addition to nanoparticles, indocyanine green, an FDA-approved dye, was also used as an exogenous agent for PT-OCT, which produced contrast for imaging the inner limiting membrane of the eye [27]. In probing endogenous sources of absorption, PT-OCT imaging of hemoglobin and melanin has been demonstrated [25,28–30], indicating the physiologically-relevant level of PT contrast.
Despite the continuous advancement of PT-OCT in both methodologies and applications [22,24,28,29,31–35], the focuses were largely on tissue imaging. Few studies were reported for cell imaging [18,36]. In an early, pioneering work of PT-OCT, Skala et al. reported depth-resolved 2D imaging of live cells targeted by gold nanospheres [18], and through optical lock-in detection, Pache et al. presented high-sensitivity, high-resolution microscopic imaging of live cells in 2D culture with gold nanoparticles [36]. These are foundational for cell imaging with PT-OCT; however, imaging of cells through highly scattering media, a unique feature of OCT, has yet to be demonstrated, and further study and characterization are needed to advance PT-OCT for applications in cell biology experiments.
We report PT-OCT for 3D detection and mapping of live cells labeled with gold nanorods (AuNRs). The method creates binary PT contrast for targeted cell detection. We demonstrated the contrast for imaging AuNR-labeled cells through experimental comparison. We performed validation experiments to show that the single cell can be detected with 3D PT-OCT imaging without causing cell death. Fluorescence imaging and transmission electron microscopy (TEM) were conducted to support the observed heterogeneity of the PT signal and its location in individual cells. Finally, we present 3D imaging of labeled cells through optically scattering media. This work is a step forward for targeted cell imaging with PT-OCT towards studying cells in their native or 3D culture environments.
2. Materials and methods
2.1 Nanorods and cells
Commercially available AuNRs (NanoXact, citrate surface, nanoComposix) were used in this study. The average lengths and thickness are 40 nm and 15 nm, respectively (Fig. 1(A)). PEG-SH (Sigma-Aldrich) was mixed with AuNRs and stirred for 12 hours to covalently modify the surface of the AuNRs with PEG via Au-thiol linkage. The resulting PEG-coated AuNRs were collected by centrifugation at 16,000 rpm for 15 minutes, washed twice with distilled water, and stored at 4 °C to prevent aggregation. Dynamic light scattering confirmed monodispersity of functionalized AuNRs (Fig. 1(B)), as the hydrodynamic diameter was measured to be ∼13.5 nm. The extinction of AuNRs reached its maximum at ∼660 nm (Fig. 1(C)).
Fig. 1. Characterization of AuNRs. (A) A TEM image of AuNRs. (B) Normalized distribution of the AuNRs hydrodynamic dimension showing a hydrodynamic diameter of around 13.5 nm with a good uniformity. (C) Normalized extinction spectrum of AuNRs showing the peak extinction at ∼660 nm. (D) A TEM image of AuNRs within a cell through endocytosis.
For all experiments, mouse breast cancer 4T1 cells were used. Cells were cultured at 37 °C in cell culture flasks with 4 mL RPMI 1640 media containing 10% fetal bovine serum and were kept in a moist atmosphere with 5% CO2. Subculture and experimental use of cells were carried out when the cells grew to 80% confluence. Labeling cells with AuNRs was achieved through endocytosis. Briefly, cells were incubated with AuNRs and subsequently washed to remove the non-endocytosed AuNRs. Cellular uptake of AuNRs was confirmed by TEM imaging (Fig. 1(D)).
2.2 PT-OCT system
The PT-OCT system was an integration of a continuous wave solid state 660 nm laser (OBIS, Coherent) and a lab-built spectral domain OCT system. The OCT system had a supercontinuum laser (SuperK EXTREME EXR-9 OCT, NKT Photonics). It provided an axial resolution of ∼7 µm in air and a transverse resolution of ∼9 µm. A line-scan camera (2048R, Sensors Unlimited) was used for the spectrometer (Cobra 1300, Wasatch Photonics) that mapped the wavelength range of 1297 ± 123 nm to 2048 pixels. The OCT system provided an A-scan rate of up to ∼147 kHz, and was measured to have a phase stability of ∼0.17 radians over 100 ms from cells in 3D culture. The 660 nm light was coupled into a single-mode fiber and was incorporated into the sample arm of OCT system through a longpass dichroic mirror with 1000 nm cut-on wavelength (Fig. 2(A)). A focus-adjustable collimator was utilized to deliver the 660 nm pump light to be focused on the same plane as the OCT imaging beam. Angular and transverse adjustments of the collimator were made to align the 660 nm beam to be co-focused with the OCT beam. The alignment was guided and confirmed by a laser beam profiler (Fig. 2(B)). The 660 nm laser functioned with an analog modulation, and the modulated light was triggered by the same TTL signal used to trigger the OCT system frame grabber to start acquiring one M-mode data. Thus, the 660 nm photothermal modulation was synchronized with the OCT M-mode imaging. After each M-mode imaging, a small amount of time interval was given to make sure all data in the computer RAM was transferred out so that prolonged data acquisition can be performed. For automatic 3D M-mode imaging, a step movement of the X galvanometer mirror was set at the start of each M-mode imaging, and a step movement of the Y galvanometer mirror was set after a defined number of X-mirror steps. The synchronization of the PT-OCT system is illustrated in Fig. 2(C).
Fig. 2. System and data acquisition with PT-OCT system. (A) Schematic of the integration of OCT and pump beams in the sample arm. (B) Characterization of beam positions showing the alignment of OCT and pump beams (the same X-Y between two graphs). (C) Illustration of the synchronization between modulation of pump beam and 3D OCT M-mode imaging.
2.3 PT-OCT imaging and data processing
The PT-OCT imaging was performed with an A-scan rate of 50 kHz and an M-mode duration of 100 ms. Modulation of the 660 nm light was through a sine wave at 200 Hz. For 2D M-mode imaging, the B-scan was set with 500 M-mode acquisitions. For 3D M-mode imaging, each B-scan had 200 M-mode acquisitions, and the volume had 200 B-scan locations. For any spatial scanning, both transverse axes were oversampled to maintain the transverse resolution. The OCT imaging beam had a power of ∼9 mW on the sample. The power of the modulated 660 nm light had a maximal amplitude of ∼44 mW on the sample, and the fluence of 660 nm light at one scan location was ∼2.4 kJ/cm2, which, based on a previous study [37], is not expected to affect the survival of live cells in culture. A study of the PT signal over different amplitudes of modulation power was performed.
The major steps of data processing is presented in Fig. 3(A). From the OCT M-mode data, both OCT amplitude and phase were extracted for structure and optical pathlength information, respectively. From the B-scan of M-mode data, a structure B-scan image was first generated by selecting one time point from the M-mode. This structure image was used to produce a binary mask eliminating the background. From each non-background pixel, the temporal phase profile was used for spectral analysis after detrending, and the amplitude spectrum was obtained by a fast Fourier transform (FFT). The spectral amplitude was scaled by a factor of 2500 due to the FFT function in MATALB. For the pixel from the AuNR-labeled region of cells, the high absorption of light led to localized thermal expansion that altered the optical pathlength, and as a result, a modulated OCT phase signal at the same frequency (200 Hz) of the 660 nm light modulation was detected (Fig. 3(B)). For the pixel from a non-labeled cell, such phase modulation and frequency peak were absent (Fig. 3(C)).
Fig. 3. Data processing for PT-OCT imaging. (A) Major steps for generating PT-OCT image. Representative temporal phase profiles (top) and amplitude frequency spectra (bottom) from (B) AuNR-labeled cell and (C) cell without AuNRs. (D) Illustration of peak, first valleys, background and PT amplitude from a frequency spectrum. (E) Histogram of PT amplitudes from cells without AuNRs showing the threshold for determining PT signal for PT image.
Similar to previous work from Tucker-Schwartz et al. [23], we defined the PT amplitude as the amplitude at 200 Hz subtracting the average amplitude of background, and the background covered the frequency range of 150-250 Hz except for 200 Hz (Fig. 3(D)). However, our PT image was produced in a different way. Aiming to detect and map targeted cells, we employed the strategy of using binary PT signal to form PT image. Specifically, with amplitude spectrum and calculated PT amplitude, we utilized three criteria to determine a true positive or a false positive PT signal, including 1) whether the amplitude at 200 Hz is larger than the maximum of the background, 2) whether the amplitudes of the first valleys on both sides of 200 Hz are smaller than half of the amplitude at 200 Hz, and 3) whether the PT amplitude is larger than a statistically defined threshold. The first two criteria characterized the peak at the modulation frequency (Fig. 3(D)). The third criterium categorized whether the PT amplitude was within the noise level which was represented by the PT amplitudes from unlabeled cells. To establish the threshold, we collected PT amplitude values from over 150 locations of unlabeled cells with our maximal power amplitude of 660 nm light modulation. Outliers were removed, which were determined as over 1.5 times of the interquartile range above the third quartile or below the first quartile. The data had a normal distribution (R-squared: ∼0.96), and we defined the threshold as µ+3.09σ, where µ is the mean and σ is the standard deviation of data (Fig. 3(E)). This threshold was consistently used across all samples and all experiments in this study. Specifically, the PT amplitude below or equal to the threshold was determined as the noise (Fig. S1). A pixel that satisfied all three criteria was assigned the PT signal value of 1, and a pixel that did not satisfy all three criteria was assigned the PT signal value of 0. Mapping of the PT signal formed the PT image. In this work, we color-coded the PT image with magenta, which highlighted the location of AuNR-labeled cells.
2.4 3D cell culture samples
All PT-OCT experiments were performed with 3D cell culture. Hydrogel with 0.5% agarose (UltraPure, Invitrogen) and PRIM 1640 cell culture media was made as the 3D matrix for cells. Cell suspension was added to and mixed with the agarose gel at the late stage of solidification. 0.5% agarose hydrogel was chosen because it did not have apparent scattering signals in OCT image, which was optimal for controlled experiment to demonstrate cell detection and mapping with PT contrast. During PT-OCT imaging, 3D cell culture was placed in an onstage incubator (ibidi) maintained at 37 °C with 5% CO2. Imaging was performed through a heated, transparent lid. 3D cultures with two different cell concentrations were used: a low concentration (1000 cells/mL) for testing of single-cell detection, and a high concentration (2 million cells/mL) for all other experiments. For single cell detection, polycaprolactone microparticles were added as dispersed landmarks to identify the imaging location across three imaging systems. The size of the microparticles is large enough for directly visualization by the naked eye, allowing for convenience in quickly and easily identifying the region for imaging. We also want to note that, in the hydrogel samples for single cell detection, we only included cells and large microparticles in the hydrogel. For experiments testing cell imaging through scattering media, an additional layer of agarose gel that contains optically scattering substance (polystyrene beads or unlabeled cells) was placed on top of the 3D cell culture.
2.5 Fluorescence and bright-field imaging
To evaluate the cell viability after PT-OCT imaging, live/dead staining (L3224, Thermo Fisher Scientific) was performed on the 3D hydrogel sample, which was then imaged using a BioTek Cytation C10 Confocal Imaging Reader. The validation of our live/dead staining and imaging protocol for cells in 3D hydrogel is shown in Fig. S2. The relative cell locations in 3D inside the hydrogel are not expected to change during the staining and imaging process. To verify and understand the heterogeneity of PT signal from cells, fluorescence imaging of AuNRs in cells was performed. Specifically, AuNRs labeled with FITC-PEG-SH (Nanocs) were incubated with cells cultured on 2D surfaces for cellular uptake. Cells were then fixed with 4% paraformaldehyde solution and stained with DAPI and rhodamine-conjugated phalloidin. Fluorescence images were taken using a Nikon 80i epifluorescence microscope. To observe and confirm the cell number in the single-cell detection experiment, bright-field imaging was performed on the 3D cell culture with a Zeiss Stemi 508 microscope. The cells were determined based on their size, the uniformity of the size, and their round shape. Different focuses were applied to assess the cell number at various depths.
2.6 TEM imaging
TEM imaging of AuNR-labeled cells was conducted to verify cellular uptake of AuNRs and to identify the intracellular destination of AuNRs. For sample preparation, cells were labeled with AuNRs using the same method as described for PT-OCT imaging. Cells were then trypsinized, washed, and centrifuged into cell pellets, which were fixed with 4% glutaraldehyde solution. Afterwards, the pellets were post-fixed with osmium tetroxide, rinsed with distilled water, and dehydrated in a series of graded ethanol. Finally, the pellets were embedded in epoxy and cut into thin sections for imaging.
3. Results
Fluorescence imaging and TEM validated the uptake of AuNRs by cells (Fig. 4). In particular, the cellular uptake of AuNRs presents a clear heterogeneity in the amount of AuNRs inside the cells (Fig. 4(A)), and the AuNRs are located in the cytoplasm (Fig. 4(B)). A representative TEM image of controlled AuNRs (not inside a cell) at the same magnification as Fig. 4(B) (right) is presented in Fig. S3 for comparison to identify AuNRs inside the cell.
Fig. 4. Location and amount of AuNRs in cells. (A) Fluorescence imaging of cells in 2D culture after AuNRs uptake, with the zoom-in region showing heterogeneous amounts of AuNRs in cells. Arrows and triangles: cells with relatively low and high amounts of AuNRs, respectively. DAPI labels the DNA in nucleus, and Phalloidin labels actin filaments in the cytoplasm. Scale bars: 50 µm. (B) A TEM image showing the AuNRs in the cell cytoplasm.
We demonstrated the PT-OCT contrast for imaging AuNR-labeled cells by comparing depth-resolved 2D images from cells with AuNRs (Fig. 5(A)) and without AuNRs (Fig. 5(B)). The structure image, PT image, and their overlay are presented. Cells with AuNRs generated clear PT signals that overlapped well with the cells in the structure image, while nearly no PT signals were seen from the cells without AuNRs, indicating the contrast of PT-OCT for depth-resolved detection and localization of labeled cells. We used cell PT signal density to assess the effect of different amplitudes of modulation power. The cell PT signal density was calculated as the total area of PT signal divided by the total area of cells, representing the percentage of the cell area that has a PT signal value of 1. With our maximal power of amplitude for photothermal modulation, the cell PT signal density from unlabeled cells was close to zero but was around 40% from AuNR-labeled cells (Fig. 5(C)), quantitatively showing the contrast. A study over different powers of pump light indicated an increasing contrast with an increase of modulation power amplitude (Fig. 5(D)). This agrees with the previous results from directly probing nanoparticles [23] and from model-based simulation analysis [21]. As an extension of 2D imaging, 3D imaging and reconstruction of cells with AuNRs shows the overlap of PT signals with the structure cell image (Fig. 6), demonstrating the feasibility of performing 3D mapping of labeled cells with PT-OCT.
Fig. 5. Validation of PT-OCT method for cell imaging. 2D depth-resolved structure, PT and overlay images of (A) AuNR-labeled cells and (B) cells without AuNRs. Scale bars: 100 µm. (C) Comparison of cell PT signal density between images of cells with and without AuNRs. Data: mean ± std with measurement points from different locations. (D) Cell PT signal density over different pump light power amplitude on the sample containing cells with and without AuNRs. Std is from measurements at different locations.
Fig. 6. 3D PT-OCT imaging of AuNR-labelled cells. Scale bars: 150 µm. Scale bars apply to all 3D directions.
We tested the feasibility of PT-OCT for detecting the single cell by 3D imaging of the low-concentration cell culture with microparticles as landmarks. With a microparticle selected and identified, the surrounding region was imaged by both 3D PT-OCT (Fig. 7(A)) and bright-field microscope at different focuses that confirmed the number of cells (Fig. 7(B)). As highlighted by Boxes 1-4 in Fig. 7(A), the PT signal can be detected from the single cell. The live/dead assessment of the PT-OCT imaging area of the same sample shows that the detection of single cell can be achieved without causing cell death (Fig. 7(C)). The live/dead imaging following 3D PT-OCT imaging was performed on three samples, and we also conducted live/dead imaging of three samples undergoing the same onstage incubation process but not imaged with PT-OCT. Comparison of the percentage of dead cells (Fig. 7(D)) reveals that PT-OCT imaging does not affect cell viability, indicating the feasibility for live cell detection, imaging and studies.
Fig. 7. Validation of single cell detection for 3D PT-OCT imaging. (A) 3D PT-OCT image with boxes 1-4 highlighting single cells with PT signals. Scale bars: 100 µm in regular view and 30 µm in zoom-in views. Arrows: low-volume PT signal from single cells. (B) Bright-field imaging of the same location in the same sample at different focus planes, with the large polycaprolactone (PCL) particle as reference. Boxes 1’-4’ correspond to boxes 1-4 in (A). Scale bars: 100 µm. (C) Fluorescence imaging of live/dead staining of the same sample and the same location after PT-OCT imaging. Scale bar: 100 µm. (D) Comparable percentages of dead cells in 3D agarose gel in culture shown from samples after PT-OCT imaging and without PT-OCT imaging (control).
From the PT-OCT images of individual cells in Fig. 7, it can be seen that the volumes of PT signals are heterogeneous, with small signal volumes in two cells (arrows, Fig. 7(A)) while larger but different signal volumes in the other cells. This is due to the heterogeneity in cellular uptake of AuNRs, which is verified by fluorescence images of cells in 2D culture (Fig. 4(A)). From the zoom-in region, it is clear that cells exhibit varying levels of AuNRs uptake through endocytosis. Some cells show relatively few AuNRs, while others have a significantly higher uptake (Fig. 4(A)), resulting in distinct volumes of PT signals. It can be observed from Fig. 7(A) that the PT signals are predominantly located on the cell periphery, agreeing with a previous study showing that nanoparticles with a size equal to or larger than 10 nm tend to remain in the cytoplasm and do not enter the nucleus [38]. Using TEM imaging, we verified that the AuNRs used in this study (40 nm by 15 nm) are indeed located within the cell cytoplasm (Fig. 4(B)), which further supports our observations with PT-OCT.
Mapping of AuNR-labeled cells in 3D with PT-OCT through optically scattering media is demonstrated in Fig. 8. Specifically, cell imaging can be achieved through a layer of densely packed polystyrene (PS) beads in agarose gel with a thickness of ∼250 µm that represents a layer of highly scattering media (Fig. 8(A) and Visualization 1). Also, 3D imaging through a ∼250 µm-thick agarose layer containing cells without AuNRs highlights the PT-OCT contrast between labeled and unlabeled cells in one image (Fig. 8(B) and Visualization 2).
Fig. 8. 3D PT-OCT imaging of AuNR-labeled cells through optically scattering media, including (A) polystyrene (PS) beads and (B) cells without AuNRs (See Visualization 1 and Visualization 2, respectively). Scale bars: 100 µm. Scale bars apply to all 3D directions.
4. Discussion
The nanoparticle-labeled cell imaging with PT-OCT has two major features that complement existing cell imaging techniques. First, this method presents a mesoscale imaging of targeted cells, filling in the gap of imaging scales between the widely used confocal imaging [11] and bioluminescence imaging [39]. The tissue-level imaging scale provided by OCT is particularly important for studying interactions between cells and their surroundings in 3D. In comparison with photoacoustic tomography and microscopy [40], the simultaneously obtained scattering contrast from OCT provides a high-resolution structure context for cell analysis, which can be used to assess the tissue structural environment of cells. With the possibility of incorporating other functional OCT contrasts, such as angiography [41] and elastography [42], a potentially multi-contrast imaging tool based on OCT could benefit cell biology studies. Second, compared with fluorescence-based imaging, the PT contrast enabled by nanoparticles offers an excellent flexibility to accommodate specific cell imaging applications in a broader optical window. In this work, we used ∼660 nm absorption peak wavelength, which is away from the major light absorption band of blood, thus potentially suitable for in vivo applications. The OCT imaging wavelength over 1000 nm allows for a wide range of near-infrared absorption peak wavelengths to be used, which has already been available from AuNRs or other nanoparticles [43,44]. With relatively narrow peaks in the extinction spectra of nanoparticles, this opens the possibility for multiplexed cell imaging with a high specificity.
It is worth to note that, with PT-OCT, cell imaging and tissue imaging are different, due to the labeling of targets with optical absorbers, e.g., nanoparticles. It is possible to inject a desired amount of nanoparticles for imaging of specific tissue components, but for cell imaging, the labeling of cells requires the uptake of nanoparticles by cells, which has a relatively lower level of flexibility and control [45]. Thus, what we know about tissue imaging with PT-OCT might not be applicable for cell imaging, particularly about the PT signal from a low concentration of cells. Although extensive research has been performed for PT-OCT tissue imaging, very few studies were reported on PT-OCT cell imaging [18,36]. More PT-OCT studies of cell imaging are required to advance this specific PT-OCT imaging direction for potential applications in cell studies in 3D. The work presented here demonstrated, for the first time that the single cell labeled through regular nanoparticle uptake can be detected and mapped in 3D with PT-OCT without causing cell death, and that the labeled cells can be mapped in 3D through an optically scattering layer. These contribute to a further understanding of the feasibility of PT-OCT for cell imaging, which can facilitate the application of PT-OCT for studying cells in their 3D environment.
In PT-OCT imaging of AuNRs, lower PT modulation frequencies can produce higher PT amplitudes [21,23]. The 200 Hz modulation frequency of the pump light used in this study was selected for a strong PT amplitude while keeping the total time for 3D PT-OCT imaging at a reasonable level. We employed 20 cycles of modulation, which can be reduced to improve the imaging speed. A further improvement of imaging speed can be achieved by using pulse waves instead of sine waves for the pump light and by probing the transient PT response, which was recently demonstrated for fast PT-OCT imaging [33]. Also, employing optical lock-in detection that modulates the reference arm at the same frequency of the photothermal modulation could eliminate the need for extensive temporal sampling [31], thus improving the imaging speed.
The generation and use of the binary PT signal for the PT-OCT image is a feature of our method but is not significantly different from the previous PT-OCT signal processing method [23]. We utilized a similar spectral analyzing approach to obtain the PT amplitude, which was then binarized with three criteria to determine the true positive and the false positive PT signal. The true positive was assigned a value of one to highlight labeled cells. In the three criteria, the major one represents a threshold established through imaging and analysis of unlabeled cells. This binary contrast provides the convenience for identifying labeled cells in 3D, though the information of signal strength from PT-OCT is sacrificed.
Cellular uptake of AuNRs is an essential part for PT-OCT cell imaging. The heterogeneity of this process is presented in our study, which supports and explains our PT-OCT observations from individual cells. With recent studies, the heterogeneous uptake has started to be understood [45–47], which will contribute to the cell labeling strategy for PT-OCT imaging. In this work, we modified the surface of AuNRs with PEG to facilitate the uptake process by 4T1 cells, and the amount of AuNRs entering the cells was sufficient to produce PT signals from individual cells. There are multiple ways to improve the single cell detection to further enhance the PT signal from the single cell. In the material point of view, using nanoparticles with a higher absorption coefficient [48] and further promoting cellular uptake of nanoparticles [49] can boost the PT signal volume from single cells. From the detection aspect, a better phase stability of the OCT system could lower the threshold in determining the PT signal, leading to a reduced noise level and an improvement for single cell detection.
Rapid advancements at the intersection of nanotechnology and cell biology have established various approaches to target specific cells, and modify, control and regulate cells for a range of applications [50–53]. Not only can PT-OCT cell imaging benefit from such advancements, but it can also serve as an imaging tool to study the relevant processes and characterize the novel approaches. The demonstration in this study is on the imaging of pre-labeled cells, while our future work will focus on 3D detection and mapping of specific cells targeted by nanoparticles through molecular binding. Built on this work, we will also seek to demonstrate dynamic cell imaging and mapping of targeted cells in a living organism.
In summary, we reported PT-OCT for high-resolution live imaging of AuNR-labeled cells in 3D. With a binary PT-OCT imaging contrast, we demonstrated that the single cell in 3D can be detected without causing cell death, and we showed that 3D mapping of cells can be achieved through optically scattering media. There have been only few studies on PT-OCT cell imaging, and the results in this paper present new, important feasibility of PT-OCT for imaging labeled cells in 3D. This represents a step forward in driving the use of PT-OCT towards cell analysis, which complements existing imaging techniques for advanced studies of cells within their 3D native tissue or culture environments.
Funding
National Institutes of Health (R35GM142953).
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. C. S. Chen, J. Tan, and J. Tien, “Mechanotransduction at cell-matrix and cell-cell contacts,” Annu. Rev. Biomed. Eng. 6(1), 275–302 (2004). [CrossRef]
2. E. Cukierman, R. Pankov, and K. M. Yamada, “Cell interactions with three-dimensional matrices,” Curr. Opin. Cell Biol. 14(5), 633–640 (2002). [CrossRef]
3. E. Armingol, A. Officer, O. Harismendy, et al., “Deciphering cell–cell interactions and communication from gene expression,” Nat. Rev. Genet. 22(2), 71–88 (2021). [CrossRef]
4. M. Ravi, V. Paramesh, S. R. Kaviya, et al., “3D cell culture systems: advantages and applications,” J. Cell. Physiol. 230(1), 16–26 (2015). [CrossRef]
5. K. M. Yamada and E. Cukierman, “Modeling tissue morphogenesis and cancer in 3D,” Cell 130(4), 601–610 (2007). [CrossRef]
6. M. J. Pittet and R. Weissleder, “Intravital Imaging,” Cell 147(5), 983–991 (2011). [CrossRef]
7. Q. Huang, A. Garrett, S. Bose, et al., “The frontier of live tissue imaging across space and time,” Cell Stem Cell 28(4), 603–622 (2021). [CrossRef]
8. G. Hilzenrat, E. T. Gill, and S. L. McArthur, “Imaging approaches for monitoring three-dimensional cell and tissue culture systems,” J. Biophotonics 15(6), e202100380 (2022). [CrossRef]
9. P. Bon and L. Cognet, “On some current challenges in high-resolution optical bioimaging,” ACS Photonics 9(8), 2538–2546 (2022). [CrossRef]
10. S. Wang and I. V. Larina, “8 - High-resolution imaging techniques in tissue engineering,” in Monitoring and Evaluation of Biomaterials and their Performance In Vivo, R. J. Narayan, ed. (Woodhead Publishing, 2017), pp. 151–180.
11. B. W. Graf and S. A. Boppart, “Imaging and t,” in Live Cell Imaging: Methods and Protocols, D. B. Papkovsky, ed. (Humana Press, Totowa, NJ, 2010), pp. 211–227.
12. A. F. Fercher, W. Drexler, C. K. Hitzenberger, et al., “Optical coherence tomography - principles and applications,” Rep. Prog. Phys. 66(2), 239–303 (2003). [CrossRef]
13. J. G. Fujimoto, C. Pitris, S. A. Boppart, et al., “Optical coherence tomography: an emerging technology for biomedical imaging and optical biopsy,” Neoplasia 2(1-2), 9–25 (2000). [CrossRef]
14. W. Tan, A. L. Oldenburg, J. J. Norman, et al., “Optical coherence tomography of cell dynamics in three-dimensional tissue models,” Opt. Express 14(16), 7159–7171 (2006). [CrossRef]
15. S. Azzollini, T. Monfort, O. Thouvenin, et al., “Dynamic optical coherence tomography for cell analysis [Invited],” Biomed. Opt. Express 14(7), 3362–3379 (2023). [CrossRef]
16. S. P. Mattison, W. Kim, J. Park, et al., “Molecular imaging in optical coherence tomography,” Current Molecular Imag. 3(2), 88–105 (2015). [CrossRef]
17. A. Wang, W. Qi, T. Gao, et al., “Molecular contrast optical coherence tomography and its applications in medicine,” in International Journal of Molecular Sciences, (2022). [CrossRef]
18. M. C. Skala, M. J. Crow, A. Wax, et al., “Photothermal optical coherence tomography of epidermal growth factor receptor in live cells using immunotargeted gold nanospheres,” Nano Lett. 8(10), 3461–3467 (2008). [CrossRef]
19. D. C. Adler, S.-W. Huang, R. Huber, et al., “Photothermal detection of gold nanoparticles using phase-sensitive optical coherence tomography,” Opt. Express 16(7), 4376–4393 (2008). [CrossRef]
20. M. Lapierre-Landry, J. M. Tucker-Schwartz, and M. C. Skala, “Depth-resolved analytical model and correction algorithm for photothermal optical coherence tomography,” Biomed. Opt. Express 7(7), 2607–2622 (2016). [CrossRef]
21. M. H. Salimi, M. Villiger, and N. Tabatabaei, “Three-dimensional opto-thermo-mechanical model for predicting photo-thermal optical coherence tomography responses in multilayer geometries,” Biomed. Opt. Express 13(6), 3416–3433 (2022). [CrossRef]
22. C. Zhou, T.-H. Tsai, D. C. Adler, et al., “Photothermal optical coherence tomography in ex vivo human breast tissues using gold nanoshells,” Opt. Lett. 35(5), 700–702 (2010). [CrossRef]
23. J. M. Tucker-Schwartz, T. A. Meyer, C. A. Patil, et al., “In vivo photothermal optical coherence tomography of gold nanorod contrast agents,” Biomed. Opt. Express 3(11), 2881–2895 (2012). [CrossRef]
24. Y. Jung, R. Reif, Y. Zeng, et al., “Three-dimensional high-resolution imaging of gold nanorods uptake in sentinel lymph nodes,” Nano Lett. 11(7), 2938–2943 (2011). [CrossRef]
25. M. Lapierre-Landry, A. Y. Gordon, J. S. Penn, et al., “In vivo photothermal optical coherence tomography of endogenous and exogenous contrast agents in the eye,” Sci. Rep. 7(1), 9228 (2017). [CrossRef]
26. A. Y. Gordon, M. Lapierre-Landry, M. C. Skala, et al., “Photothermal optical coherence tomography of anti-angiogenic treatment in the mouse retina using gold nanorods as contrast agents,” Trans. Vis. Sci. Tech. 8(3), 18 (2019). [CrossRef]
27. M. Lapierre-Landry, T. B. Connor, J. Carroll, et al., “Photothermal optical coherence tomography of indocyanine green in ex vivo eyes,” Opt. Lett. 43(11), 2470–2473 (2018). [CrossRef]
28. M. Lapierre-Landry, A. L. Huckenpahler, B. A. Link, et al., “Imaging melanin distribution in the zebrafish retina using photothermal optical coherence tomography,” Trans. Vis. Sci. Tech. 7(5), 4 (2018). [CrossRef]
29. Y. Biwei, V. K. Roman, B. M. Austin, et al., “Dual-wavelength photothermal optical coherence tomography for imaging microvasculature blood oxygen saturation,” J. Biomed. Opt. 18(5), 056005 (2013). [CrossRef]
30. S. Makita and Y. Yasuno, “In vivo photothermal optical coherence tomography for non-invasive imaging of endogenous absorption agents,” Biomed. Opt. Express 6(5), 1707–1725 (2015). [CrossRef]
31. J. M. Tucker-Schwartz, M. Lapierre-Landry, C. A. Patil, et al., “Photothermal optical lock-in optical coherence tomography for in vivo imaging,” Biomed. Opt. Express 6(6), 2268–2282 (2015). [CrossRef]
32. J. Wu, N. Wu, P. Tang, et al., “Pulse photothermal optical coherence tomography for multimodal hemodynamic imaging,” Opt. Lett. 46(22), 5635–5638 (2021). [CrossRef]
33. M. H. Salimi, M. Villiger, and N. Tabatabaei, “Transient-mode photothermal optical coherence tomography,” Opt. Lett. 46(22), 5703–5706 (2021). [CrossRef]
34. Y. Hu, A. Podoleanu, and G. Dobre, “Photothermal optical coherence tomography for investigation and imaging photothermal trapping of gold nano-rods in clear media and biological tissue,” J. Opt. 21(9), 095301 (2019). [CrossRef]
35. S. Mohammad Hossein, V. Martin, and T. Nima, “Effects of lipid composition on photothermal optical coherence tomography signals,” J. Biomed. Opt. 25(12), 120501 (2020). [CrossRef]
36. C. Pache, N. L. Bocchio, A. Bouwens, et al., “Fast three-dimensional imaging of gold nanoparticles in living cells with photothermal optical lock-in Optical Coherence Microscopy,” Opt. Express 20(19), 21385–21399 (2012). [CrossRef]
37. S. Wäldchen, J. Lehmann, T. Klein, et al., “Light-induced cell damage in live-cell super-resolution microscopy,” Sci. Rep. 5(1), 15348 (2015). [CrossRef]
38. S. Huo, S. Jin, X. Ma, et al., “Ultrasmall gold nanoparticles as carriers for nucleus-based gene therapy due to size-dependent nuclear entry,” ACS Nano 8(6), 5852–5862 (2014). [CrossRef]
39. C. E. Badr and B. A. Tannous, “Bioluminescence imaging: progress and applications,” Trends Biotechnol. 29(12), 624–633 (2011). [CrossRef]
40. J. Yao, A. A. Kaberniuk, L. Li, et al., “Multiscale photoacoustic tomography using reversibly switchable bacterial phytochrome as a near-infrared photochromic probe,” Nat. Methods 13(1), 67–73 (2016). [CrossRef]
41. C.-L. Chen and R. K. Wang, “Optical coherence tomography based angiography [Invited],” Biomed. Opt. Express 8(2), 1056–1082 (2017). [CrossRef]
42. S. Wang and K. V. Larin, “Optical coherence elastography for tissue characterization: a review,” J. Biophotonics 8(4), 279–302 (2015). [CrossRef]
43. X. Huang and M. A. El-Sayed, “Gold nanoparticles: Optical properties and implementations in cancer diagnosis and photothermal therapy,” J. Adv. Res. 1(1), 13–28 (2010). [CrossRef]
44. T. R. Jensen, M. D. Malinsky, C. L. Haynes, et al., “Nanosphere lithography: tunable localized surface plasmon resonance spectra of silver nanoparticles,” J. Phys. Chem. B 104(45), 10549–10556 (2000). [CrossRef]
45. S. Behzadi, V. Serpooshan, W. Tao, et al., “Cellular uptake of nanoparticles: journey inside the cell,” Chem. Soc. Rev. 46(14), 4218–4244 (2017). [CrossRef]
46. P. Rees, J. W. Wills, M. R. Brown, et al., “The origin of heterogeneous nanoparticle uptake by cells,” Nat. Commun. 10(1), 2341 (2019). [CrossRef]
47. A. Lesniak, A. Salvati, M. J. Santos-Martinez, et al., “Nanoparticle adhesion to the cell membrane and its effect on nanoparticle uptake efficiency,” J. Am. Chem. Soc. 135(4), 1438–1444 (2013). [CrossRef]
48. P. K. Jain, K. S. Lee, I. H. El-Sayed, et al., “Calculated absorption and scattering properties of gold nanoparticles of different size, shape, and composition: applications in biological imaging and biomedicine,” J. Phys. Chem. B 110(14), 7238–7248 (2006). [CrossRef]
49. L. Huang, X. Mao, J. Li, et al., “Nanoparticle spikes enhance cellular uptake via regulating Myosin IIA recruitment,” ACS Nano 17(10), 9155–9166 (2023). [CrossRef]
50. Y. Pan, X. Du, F. Zhao, et al., “Magnetic nanoparticles for the manipulation of proteins and cells,” Chem. Soc. Rev. 41(7), 2912–2942 (2012). [CrossRef]
51. N. K. Lee, S.-N. Kim, and C. G. Park, “Immune cell targeting nanoparticles: a review,” Biomater Res. 25(1), 44 (2021). [CrossRef]
52. M. F. Attia, N. Anton, J. Wallyn, et al., “An overview of active and passive targeting strategies to improve the nanocarriers efficiency to tumour sites,” J. Pharm. Pharmacol. 71(8), 1185–1198 (2019). [CrossRef]
53. D. C. Thang, Z. Wang, X. Lu, et al., “Precise cell behaviors manipulation through light-responsive nano-regulators: recent advance and perspective,” Theranostics 9(11), 3308–3340 (2019). [CrossRef]
