Functional photoacoustic calcium imaging using chlorophosphonazo III in a 3D tumor cell culture

: This study demonstrates that chlorophosphonazo III (CPZ III) can be used as a contrast agent for photoacoustic calcium imaging. CPZ III can pass across the plasma membrane for labeling intracellular Ca 2 + without cytotoxicity. In optical-resolution photoacoustic microscopy (OR-PAM), the photoacoustic (PA) signal intensity was strongly correlated with the presence of CPZ III and Ca 2 + at various concentrations. The sensitivity of PA signal reception was enhanced by using an 8 MHz single-element focused ultrasound detector due to their matched frequency characteristics. Differences in the PA signal intensity were successfully found between the core and margin areas of tumorspheres in three-dimensional cell cultures. These findings indicate that CPZ III can serve as a novel PA contrast agent for functional Ca 2 + imaging using OR-PAM.


Introduction
Tumorigenesis involves the formation of metabolic gradients mediated by nutrient and oxygen supplementation from the margin to core regions, thus causing zonation of the tumor [1]. The proliferating cells are mainly distributed in the peripheral margin region due to the high levels of nutrients and oxygen, while quiescent and necrotic cells are found in the core region due to the lack of nutrients and oxygen. Dysregulated Ca 2+ homeostasis of tumor cells is strongly associated with tumorigenesis due to the abnormal expression of Ca 2+ transporters or ion channels, and so thus targeting Ca 2+ signaling has been hypothesized as an emerging strategy for cancer therapy [2]. An increased Ca 2+ level in the core region of tumorspheres for defending oxidative stress via TRPA1-Ca 2+ -channel-mediated Ca 2+ influx has also been reported, and the reduced expression of TRPA1 can damage this protective mechanism [3]. This indicates the potential for the therapeutic targeting of Ca 2+ signaling by using a tumorsphere model as the drug screening platform. The approaches in the literature involve forming tumorspheres for Ca 2+ imaging on Matrigel-coated glass in order to overcome the low penetration of optical calcium imaging systems. Although avascular tumorsphere model mimics the micrometastasis observed in patients [4], the cell-matrix interactions in such a tumorsphere model are close to a two-dimensional cell culture and thus do not realistically mimic the real physiological three-dimensional (3D) tumor microenvironment [5]. The present study used scaffold-based multicellular tumorspheres to mimic the 3D tumor microenvironment, which resulted in the cellular behavior being much more realistic.
The 3D Ca 2+ dynamics in in vivo animal models can be imaged using various optical imaging methods, including wide-field two-photon microscopy [6], light-field deconvolution microscopy [7], or light-sheet fluorescence microscopy [8], but most previous experiments have employed model organisms that are relatively transparent in order to improve the imaging depth. Nevertheless, the imaging depth is still smaller than 200 µm. In contrast, photoacoustic (PA) imaging uses an ultrasound (US) system to detect the acoustic signals resulting from the rapid thermoelastic expansion of chromophores induced by excitation from a pulsed laser source. This approach can produce both structural and functional images of chromophore-labeled cells at significant penetration depths, and thus has potential for deployment in investigating 3D volumetric Ca 2+ dynamics.
The local PA signal intensity after photonic extinction is given by where P 0 is the intensity of the locally generated initial PA pressure wave, Γ is the Grüneisen parameter, which denotes the PA efficiency of the absorber and is defined as Γ = βν s 2 /C p , and A is the absorbed optical energy density, which is defined as A = µ a F. Γ is related to heat capacity C p , the thermal expansion coefficient of optical absorber β, and speed of sound ν s in the tissue. Here, µ a is the local absorption coefficient and F is the local laser fluence. Accordingly, the PA intensity is proportional to the absorption coefficient of the absorbers and the local laser fluence, which are the two factors that are often adjusted and varying in in vitro cell studies.
In addition to the properties of the optical absorbers and the specifications of the employed laser, the PA imaging performance at a certain depth of a targeted tissue is also strongly influenced by the sensitivity of the US transducers used for detecting PA signals. The US pressure waves generated by a pulsed laser usually cover a broad frequency range, whereas the restricted spectral bandwidth of a US transducer restrains the received spectral range of PA signals [9]. Although US transducers with a higher center frequency offer a better image spatial resolution, the PA signal attenuation in tissue increases at higher frequencies [10]. Furthermore, there is a trade-off between the detection sensitivity and the detection bandwidth, and so selecting a US transducer with a bandwidth that matches the generated PA signals will optimize the reception sensitivity for PA signals [11].
Nowadays, most of the PA Ca 2+ contrast agents were designed based on the genetically encoded fluorescent Ca 2+ indicators (GECIs) that emitted PA signals due to Ca 2+ binding to calmodulin that has been genetically modified to fuse to fluorescence proteins [12][13][14][15]. However, the optical absorbance wavelength of such PA Ca 2+ contrast agents is usually lower than 550 nm, which decreased the penetration ability in 3D scaffolds or tissue. Moreover, introducing GECIs into tissue or cells sometimes had challenges, especially in the nervous system, due to the low transfection or viral infection rate of neurons. The major advantage of chemical Ca 2+ contrast agents over GECIs is the ease of introducing and rapidly applying these contrast agents for cellular experiments due to their cell-permeant ability. However, very few reports showing the feasibility of chemical Ca 2+ contrast agents for PA Ca 2+ imaging [16][17][18][19]. Among these studies, CaSPA_550 has been reported as a chemical PA Ca 2+ contrast agents with a good cell permeability and selectivity of Ca 2+ over Mg 2+ , Zn 2+ , and Cu 2+ [19]. Nevertheless, the prominent optical absorbance of CaSPA_550 is also at 550 nm.
Chlorophosphonazo III (CPZ III) is an analog of arsenazo III (AZ III), with both reagents being Ca 2+ chelators that can be employed for measuring Ca 2+ using a colorimetric method. Nevertheless, AZ III is an organic arsenic compound that contains traces of arsenic, and so its disposal is hampered by concerns about environmental contamination; however, this is not an issue when using CPZ III. The generation of free radicals from AZ III in cells and the high concentrations of AZ III employed during PA imaging may also reduce the cell viability [16,20]. CPZ III stably forms a 1:1 complex with Ca 2+ over a long period [21,22]. Although CPZ III has been suggested as a potential PA contrast agent for imaging Ca 2+ dynamics [19], detailed PA profiles of CPZ III in live cells have not been experimentally determined previously. Two absorbance peaks of CPZ III in the presence of Ca 2+ ions occur at around 610 and 660 nm, which are located in the near-infrared (NIR) window (600-900 nm). This makes CPZ III attractive for 3D PA Ca 2+ imaging using PA microscopy due to its low degree of optical scattering in 3D scaffolds or tissue. Except for AZ III, two other NIR PA Ca 2+ contrast agents have been reported [17,18]. Although one of the studies showed that the PA Ca 2+ contrast agent "L" has higher selectivity of Ca 2+ than CPZ III [17], both these two studies were lacking the experimental evidence to clarify the cytotoxicity of the contrast agents or the feasibility of live cell PA Ca 2+ imaging. We therefore selected CPZ III as the potential PA contrast agent for observing changes in intratumoral Ca 2+ in a 3D tumor cell culture. This study not only explored the use of a novel PA contrast agent for Ca 2+ imaging, but it also sheds light on the 3D Ca 2+ imaging technique field, and the results can be used to guide the development of Ca 2+ -targeting cancer therapeutic strategies.

CPZ III dye preparation and absorption spectra measurements
CPZ III (CAS 1914-99-4, Santa Cruz Biotechnology, Dallas, TX, USA) was dissolved in DMSO (Sigma-Aldrich, Saint Louis, MO, USA) to make a stock solution. CPZ-III-Ca 2+ complex solutions with various molar ratios of Ca 2+ to CPZ III for spectra measurements were prepared. Absorption spectra of 1-ml samples of CPZ-III-Ca 2+ complex solutions were measured using a UV-visible spectrometer (V650, Jasco, Easton, MD, USA).

Optical-resolution photoacoustic microscopy
A schematic of the optical-resolution photoacoustic microscopy (OR-PAM) system established for performing PA calcium imaging is shown in Fig. 1. The laser pulses were generated by a 523-nm Nd:YLF laser (IS8II-E, EdgeWave, Würselen, Germany) with an adjustable pulse repetition frequency from 10 to 5000 Hz, an energy of 10 mJ per pulse, and a pulse duration of 6 ns. To generate a laser pulse with an optical wavelength appropriate for PA Ca 2+ imaging when using CPZ III as the PA contrast agent, a dye pulse laser (Cobra, Sirah Lasertechnik, Grevenbroich, Germany) with a tuning wavelength in the range of 602-660 nm and a peak output at 627 nm was utilized, and the 523-nm Nd:YLF laser was directed into the dye pulse laser containing an organic source (Exciton, Lockbourne, OH, USA). The product manual of the laser indicates that its energy conversion efficiency at 627 nm is 27% and may be changed depending on the pump power level. The pulse-to-pulse laser energy variation was determined when 10% laser was passed through a beamsplitter (Thorlabs, Newton, NJ, USA) and redirected to a pyroelectric energy meter (PE10-C, Ophir Optronics Solutions, Jerusalem, Israel), which allowed the received PA signals to be normalized to the variations in the laser energy. The microscope objective used for laser focusing had a 4-mm working distance and a numerical aperture of 0.45 (CFI Plan Apo Lambda 10×, Nikon, Tokyo, Japan). The emitted PA signals were detected using single-element focused US transducers. The objective was confocally and coaxially aligned with the US transducers. Acquired PA signals were amplified with a pulser-receiver (5073PR, Olympus, Waltham, MA, USA) and digitized with a 14-bit analog-to-digital conversion card at a sampling rate of 200 MSa/s (CSE1422, GaGe, Lockport, IL, USA) with triggering by a photodiode. Rapid raster scanning was accomplished by using an x-y galvanometer optical scanner (6231HB, Cambridge Technology, Bedford, MA, USA). In our previous study, imaging experiments on copper wires demonstrated that the lateral resolution of the OR-PAM was 1.6 µm, and the axial resolution was 5 µm [23].
3D volumetric imaging was achieved by controlling a three-axis motorized translation stage (MT3-Z8, Thorlabs). The external triggers for the laser, the commands to the galvanometer scanner, and the control of the 3D scanning routes were all generated by a DAQ card (USB-6341, National Instruments, Austin, TX, USA) housed in a PC to control scanning patterns through a customized LabVIEW program (National Instruments). Each depth-resolved A-line signal was created by a single laser pulse. Bright-field images of the samples were generated by using a white-light source and captured through a CMOS camera (CCN-B013-U, Mightex Systems, Toronto, Canada). To achieve live cell imaging, the 3D cell culture samples were maintained at 37°C on the sample stage by a silicone rubber heater in a customized clear acrylic chamber containing 5% CO 2 during imaging.

Cell culture, CPZ III staining, and cell viability assay
Neuro2A cells (a mouse neuroblastoma cell line) were applied as the targeted cancer cell model in this study. The cells were cultured with Dulbecco's modified Eagle's medium (Thermo Fisher Scientific, Waltham, MA, USA) containing 10% fetal bovine serum (Thermo Fisher Scientific) and 1% penicillin/streptomycin (Thermo Fisher Scientific) in an incubator at 37°C containing 5% CO 2 . After cells were passaged and cultured for 24 hours on glass coverslips, CPZ III was added at 0, 50, 75, 100, or 150 µM to the culture medium for 30 minutes. The coverslips with the CPZ-III-stained cells were then washed using phosphate-buffered saline to remove the excess CPZ III before being investigated using fluorescence microscopy with a Cy5 filter set (620-660 nm excitation and 700-775 nm emission; Leica Microsystems, Wetzlar, Germany). Cell viability was evaluated using a trypan blue exclusion test. CPZ-III-stained cells were trypsinized off the culture dishes using 0.25% trypsin-EDTA (Thermo Fisher Scientific) followed by trypan blue staining (Thermo Fisher Scientific). The number of stained dead cells was excluded from the total number of cells (unstained and stained) when determining cell viability.

Phantom design and fabrication
To keep the CPZ-III-Ca 2+ complex solutions within a confined area as well as motionless during imaging, a plastic tube with an inner diameter of 0.6 mm and an outer diameter of 1 mm was embedded in 10% gelatin gel (Sigma-Aldrich). The tube was placed close to the bottom of the glass dish so that the laser could be focused on the region of interest. This configuration was used to detect the PA signals of the CPZ-III-Ca 2+ complexes as described in Sections 3.2 and 3.3. A schematic of the phantom setup is presented in Fig. 2.

PA imaging transducers
Four focused US transducers with different center frequencies and focusing properties were utilized in the PA imaging experiments. The detailed specifications of the transducers are presented in Table 1.  To generate Neuro2A tumorspheres, cells were trypsinized off the culture dishes and groups of 50 cells were used to form single tumorspheres in ultra-low-attachment round-bottomed plates (Corning, Corning, NY, USA). The cell suspensions were then cultured for 3-7 days in tumorsphere medium [24] to allow them to grow to at least 50 µm in diameter. Before PA imaging, the formed tumorspheres were removed from the plate and stained with CPZ III at 37°C for around 60 minutes, depending on the size of each tumorsphere. The CPZ-III-stained tumorspheres were then mixed with 20 µl of 5 mg/ml Matrigel (Corning) and the mixtures were loaded into wells with a diameter of 3 mm and height of 2 mm that had been created in the 0.6% agar mold on the glass culture dish. The gelation of Matrigel was performed at 37°C for 30 minutes, after which the height of the Matrigel was 1 mm. A schematic of the 3D cell culture setup is presented in Fig. 3.  Table 1 were used for detecting PA signals in the phantom experiments. CPZ III solutions containing Ca 2+ at various concentrations were freshly prepared by mixing CaCl 2 (Sigma-Aldrich) with CPZ III solutions in the reaction buffer. The solutions were loaded into the tube embedded in the fabricated phantom (Fig. 2) before imaging. Two-hundred A-lines were acquired at 1 kHz, with the energy normalized using MATLAB software (MathWorks, Natick, MA, USA). The peak amplitude values of the A-line data were collected, averaged, and normalized to the average values for CPZ III solutions containing no calcium. The final analyzed data were obtained from five independent experiments. For PA Ca 2+ imaging in the Neuro2A 3D cell culture, the self-made single-element 8 MHz focused transducer was used to receive PA signals in accordance with the validation results described in Section 3.2. Before imaging, the reaction buffer was added to the culture dish to cover the gel and the surface of the US transducer. The intratumoral baseline Ca 2+ level was imaged for the first 30 seconds, and then thapsigargin was added to the reaction buffer so that the final concentration reached 2 µM. After imaging for another 90 seconds, the CaCl 2 solution was added to the buffer to produce a final concentration of 1 mM, and then imaging was performed for the following 60 seconds. Five segmented images with a size of 100 µm × 100 µm (x-and y-axis, respectively) along the z-axis with a step size of 5 µm were acquired to obtain the Ca 2+ dynamics in the middle of the tumorsphere. The PA signal intensities collected from the first and last images were averaged to represent the Ca 2+ dynamics in the margin region of the tumorsphere, and the same process was applied to the middle-three images to represent the core region.

Characterization of CPZ III
The absorption spectra of CPZ III alone and the CPZ-III-Ca 2+ complexes in the reaction buffer are depicted in Fig. 4. The spectra crossed each other at 588 nm, corresponding to the isosbestic point. CPZ-III-Ca 2+ complexes showed two absorption peaks, at 610 and 660 nm, which were covered by the tuning wavelength range of the selected dye laser (i.e., 602-660 nm) used in OR-PAM. In the presence of free Ca 2+ , the absorbance at 610 nm increased linearly with the molar ratio of Ca 2+ to CPZ III. However, linearity was not evident at 660 nm, where the absorbance did not increase when the molar ratio of Ca 2+ to CPZ III exceeded 0.67. The absorbance at 627 nm, corresponding to the peak energy output over the tuning range of the dye laser selected for the OR-PAM system, also increased linearly with the molar ratio of Ca 2+ to CPZ III. Moreover, the difference in absorbance between 0.05 and 0.1 mM Ca 2+ was 2.5-fold larger at 627 nm than at 610 nm, and 1.9-fold larger between 0.1 and 2 mM Ca 2+ . Such a larger absorbance contrast suggests that the change in the PA signal for different concentrations of CPZ-III-Ca 2+ complexes is enhanced when using the selected dye laser. Hence, although the peak absorption was at 610 nm, selecting a dye laser with a peak output at 627 nm has great potential in differentiating changes in the PA signal intensity among CPZ-III-Ca 2+ complexes containing Ca 2+ at various concentrations.
In order to detect the calcium dynamics in the live-cell system, the cytotoxicity of CPZ III was measured in cells. Under the fluorescence microscopy with a Cy5 filter set, cells emitted fluorescence after CPZ III was loaded into cells for 30 minutes, with the fluorescence brightness increasing with the loaded concentration of CPZ III [ Fig. 5(A)], which corresponds to a higher optical absorbance. The fluorescence came from within the intracellular spaces, indicating the membrane permeability of CPZ III. For the same treatment time, the cell viability was not affected when cells were treated with CPZ III at 50-1000 µM [ Fig. 5(B)], demonstrating that CPZ III is not harmful to cells. However, the cell viability was significantly decreased when the CPZ III concentration over 750 µM. The cause of the decreased cell viability was due to the cytotoxic tolerance of Neuro2A cells, where the cell viability was significantly reduced when the concentration of DMSO (i.e., the solvent for the preparation of CPZ III reagent) in the culture medium was over 1.5% [ Fig. 5(C)]. Equation (1) indicates that the PA signal intensity is positively correlated with the optical absorbance, and thus we decided to use higher concentrations of CPZ III (i.e.,100 or 150 µM) as the starting concentration for detecting calcium dynamics. The absorption spectra of the CPZ-III-Ca 2+ complexes also suggested that the PA signals of Ca 2+ at concentrations ranging from 0.025 to 2 mM could be distinguished when the concentration of CPZ III was fixed at 150 µM. The Student's t test was applied for the determination of the significant difference between two sets of data. *, p<0.05; ***, p<0.001.

PA performance for different US transducer center frequencies
We used US transducers with center frequencies of 7.5, 8, 15, and 20 MHz (see Table 1) to evaluate the sensitivity of PA signal reception. The results showed that using an 8 MHz US detector with a bandwidth ranging from 6.4 to 13.5 MHz improved the PA signal intensity of CPZ-III-Ca 2+ complexes [ Fig. 6(A)], suggesting that the frequency content of the target detection signals was located within the bandwidth of the 8 MHz US detector. Moreover, it provided the best linear fit of the PA signal intensity as a function of the Ca 2+ concentration, with an R 2 coefficient of 0.97. The PA intensity of the CPZ-III-Ca 2+ complexes increased with the Ca 2+ concentration. The strong linear relationship between the detected PA intensity and the Ca 2+ concentration as well as the broadly detectable concentration range of Ca 2+ supports that intracellular calcium dynamics can be detected in a highly predictable and broadly applicable manner. Accordingly, we decided to employ the focused US transducer with the center frequency of 8 MHz as the PA signal detector in the subsequent experiments. The strong linear relationship between the detected PA signals and multiple CPZ-III-Ca 2+ complexes is consistent with the corresponding optical absorbance at 627 nm, indicating the presence of the absorption-dependent PA detection. The PA signal amplitude was linearly correlated with the absorption coefficient, and the frequency characteristics of absorbers also vary with their absorption coefficient [25][26][27]. According to Beer's law and the absorbance values at 627 nm shown in Fig. 4, the absorption coefficients of 150 µM CPZ III for Ca 2+ in the tubes with the diameter of 600 µm can be calculated as 21.7, 35, 48.3, 56.7, and 58.3 cm −1 for 0, 0.025, 0.05, 0.1, and 2 mM Ca 2+ , respectively. A previous study found that the peak frequency of PA signals for objects with an absorption coefficient of around 60 cm −1 was approximately 6 MHz [25], which is around the bandwidth of the 8-MHz US transducer selected for the present study (i.e., 6.4-13.5 MHz). In general, media exhibiting higher absorptivity produce acoustic signals with higher frequencies, and the PA image contrast can be enhanced by selecting a US detector whose bandwidth matches the spectral content of the emitted PA signals [25]. The spectrum of the detected PA signals of CPZ-III-Ca 2+ complexes showing that the center frequency of the PA signal was 7.7 MHz, and the -6dB bandwidth was ranged from 4 to 13.3 MHz [ Fig. 6(B)].
It further explains why the use of an 8-MHz US transducer with a bandwidth ranged from 6.4 to13.5 MHz can achieve a better imaging performance. Our results showed that the absorptivity increased with the molar ratio of Ca 2+ to CPZ III, and that it was below 50 cm −1 for the lower molar ratios investigated. This suggests that the imaging sensitivity and the signal contrast of these PA signals was reduced due to the spectral content being outside the US detector bandwidth.

PA signals of CPZ III
The intensities of PA signals received from the media loaded with CPZ III alone or CPZ III-Ca 2+ complexes are shown in Fig. 7. The peak PA amplitude was 10-fold higher for the CPZ-III-Ca 2+ complexes than for the CPZ-III-only media at the acquired confocal site of the focused US transducer and the microscope objective [ Fig. 7(A)]. We also investigated the PA response of CPZ-III-Ca 2+ complexes for 100 µM CPZ III with varying Ca 2+ concentrations, with the aim of determining the optimal dose of CPZ III. There were strong correlations in each CPZ III group with the received PA intensity, with Pearson's r values of 0.99 and 0.95 for 150 and 100 µM CPZ III, respectively [ Fig. 7(B)]. However, the linear fits revealed that the coefficient of determination was higher for 150 µM CPZ III (R 2 = 0.95) than for 100 µM CPZ III (R 2 = 0.75). This suggests that the intracellular calcium dynamics would be less predictable when using 100 µM CPZ III as the Ca 2+ indicator reagent. The weaker linear relationship between PA intensity and Ca 2+ concentration for 100 µM CPZ III is probably attributable to the reduced signal-to-noise ratio of PA signals due to the lower optical absorbance compared with 150 µM CPZ III. The PA intensity increased with the molar ratio of Ca 2+ to CPZ III, which quantitatively agrees with the spectra results in Fig. 4 and the results obtained by Yoshikami and Hagins, who performed experiments at a neutral pH value over the spectral range of 620-670 nm [28].

PA calcium imaging in 3D cell cultures
We observed the zonal Ca 2+ dynamics of tumorspheres located at depths of up to 1 mm by using thapsigargin, which is a widely used cellular endoplasmic reticulum (ER) stress inducer, followed by administering extracellular Ca 2+ to induce the intracellular Ca 2+ dynamics. Thapsigargin can rapidly discharge the ER Ca 2+ store via the selective inhibition of the activity of sarcoplasmic reticulum/ER Ca 2+ -ATPase (SERCA), followed by inducing a transient increase in the cytosolic Ca 2+ concentration and the acceleration of extracellular Ca 2+ entry via store-operated Ca 2+ (SOC) influx.
A 35-µm-wide peripheral region and a 15-µm-wide central region of the tumorsphere were defined as the margin area and the core area, respectively [ Fig. 8(A)]. A representative 3D reconstruction PA image was shown in Fig. 8(B). Distinctive zonal Ca 2+ dynamics of the tumorsphere were investigated using OR-PAM [ Fig. 8(C)]. The intratumoral PA Ca 2+ signals showed that treatment with thapsigargin resulted in elevation of the intracellular Ca 2+ signal intensity that was initiated at an imaging time point of 30 seconds in the margin area, followed by a 30-second delay in the core area, and that the Ca 2+ signal intensity peaked in both areas within 60 seconds [ Fig. 8(D)]. The peak values in the core and margin areas were elevated by 2.4-and 1.9-fold relative to the baseline levels, indicating that the amount of Ca 2+ released from the ER store into the cytosolic space was greater when cells were in a low oxygen condition (i.e., core area).
The addition of extracellular Ca 2+ induced a second increase in the intracellular Ca 2+ signal intensity at an imaging time point of 120 seconds in the margin area, followed by a 30-second delay in the core area [ Fig. 8(D)]. The Ca 2+ influx showed 1.3-and 1.8-fold increases within 30 seconds from the activation points for the core and margin areas, respectively. The delayed response in the core area may be due to thapsigargin or CaCl 2 taking around 30 seconds to diffuse from the margin area to the core area. The successful induction of Ca 2+ entry from the extracellular space indicates that the Ca 2+ release was accompanied by the activation of SOC influx in both areas. Moreover, it also demonstrated that after 3D PA imaging of the whole tumorsphere under an optical energy of 2.5 µJ per pulse, the cells were still viable to respond to the second stimulation (i.e., the added-back Ca 2+ ) for maintaining intracellular Ca 2+ homeostasis.
Under the normoxia condition, Ca 2+ pumping from SERCA is a major ATP charge of cells, where the mobilization of two Ca 2+ ions into the ER lumen costs one ATP molecule [29]. Once cells suffer from environmental stress (e.g., low oxygen), ATP-consuming cell functions (e.g., ion pumping, protein synthesis, or the cell cycle) will be suppressed [30]. When ATP is depleted under a low oxygen condition, activation of SERCA will be suppressed. Moreover, treatment with thapsigargin will also inhibit SERCA activity. These two ER stressors would together promote Ca 2+ release from the ER store into cytosol via inositol 1,4,5-trisphosphate receptors (IP 3 R) and ryanodine receptors (RyR) in the core area of tumorspheres that may contain local lower oxygen levels. The greater Ca 2+ influx might also be due to the increased gene expression of IP 3 R or RyR induced by low oxygen stress [31,32]. We also found that stress due to low oxygen induced dysregulation in the SOC influx rate, where after Ca 2+ was added back, the Ca 2+ signal intensity increased around 1.4-fold faster in the margin area than in the core area of tumorspheres.

Conclusions
This study is the first to demonstrate the feasibility of using CPZ III as a novel NIR contrast agent for functional Ca + imaging using OR-PAM. The nontoxicity, membrane permeability, and PA-signal-generation ability of CPZ III alone and CPZ-III-Ca 2+ complexes were determined. We also found that the frequency content of PA signals from CPZ-III-Ca 2+ complexes matched the bandwidth of an 8 MHz focused US detector. Using this US transducer allowed the intracellular calcium dynamics to be imaged in a highly predictable manner.
To image functions in physiological milieu is an emerging field with various applications for investigating pathological scenarios. Regarding the application of developing cancer therapies through Ca 2+ signaling during tumorigenesis, we established a scaffold-based 3D tumorsphere platform to mimic physiological tumor microenvironments. We have successfully demonstrated temporal and zonal changes in the intratumoral PA Ca 2+ signals in tumorspheres at depths of up to 1 mm following treatments with an SERCA inhibitor and extracellular Ca 2+ . Since the absorbance of CPZ III depends not only on the Ca 2+ concentration but also the environmental pH values, it has the potential to be a PA pH-sensitive contrast agent. The Ca 2+ dynamics could be measured separately in the margin area and the core area of a tumor based on the different local pH values by using the developed imaging system. This methodology will facilitate the development of Ca 2+ -targeted drugs that are sensitive to the tumor microenvironment.
The difference in PA signal intensity between 0 mM and 0.025 mM Ca 2+ shown in Fig. 7(B) was only 16% and 20% increase in 100 µM and 150 µM CPZ III groups, respectively. Due to the limitation of the system sensitivity, we could not detect differentiable PA signals from smaller concentrations of Ca 2+ .
The utilization of chemical PA Ca 2+ contrast agent on in vivo brain PA imaging has a limitation that the agents may have to pass through the blood-brain barrier to achieve a successful delivery to the brain [33]. And the obstacle for both GECIs or chemical PA Ca 2+ contrast agents is the optical diffusion limit may reduce the capability of deep tissue imaging [34]. Accordingly, the PA imaging technologies that work with the PA Ca 2+ contrast agent should be carefully selected in order to achieve the maximal PA imaging performance.