2.1.

ALA-PDD using Fluorescence Photoswitching

The concept of fluorescence photoswitching (Fig. 1) is as follows: PDD for a deeply located tumor is performed by exciting the photosensitizer with green light. During the fluorescence detection, there is continual photobleaching of the photosensitizer, accompanied by the formation of the photosensitizer photoproducts. The photoproducts formed accumulate as the photosensitizer photobleaches. After notable photobleaching of the photosensitizer, the wavelength of the excitation light is changed to also excite the accumulated photoproducts. The fluorescence emitted from the combination of the photosensitizer and its photoproducts yield fluorescence intensity higher than that of the photobleached photosensitizer alone. Thus, the tumor detection efficacy is improved with the increased fluorescence emission intensity.

Fig. 1

Concept of fluorescence photoswitching. The irradiation of PpIX during PDD leads to its photobleaching and the formation and accumulation of Ppp. PDD with 505 nm is primarily used to excite PpIX, whereas 450 nm is primarily used in Ppp excitation. The photobleaching of PpIX leads to a reduced fluorescence emission when excited, and the excitation of the accumulated Ppp increases fluorescence emission, leading to fluorescence photoswitching. The fluorescence emission from PpIX is shown in red, and the fluorescence emission from Ppp is shown in yellow.

The photosensitizer used in 5-aminolevulinic acid-based photodynamic diagnosis (ALA-PDD) is PpIX. Thus, the expected dominant photoproduct is Ppp. Similar to PpIX, Ppp can be excited at various wavelengths. It has been reported that the highest relative fluorescence intensity of Ppp is observed for excitation with 450 nm.²² The absorption spectra of PpIX irradiated with 505 nm light (Fig. S1 in the Supplementary Material) showed a peak at 667 nm, which is absent in the unirradiated PpIX. Similar to other studies, the absorption spectra also showed the broadening of the Soret band, which increased in the absorption intensity at around 450 nm.^22,23 A ratio of the irradiated absorption spectra to the unirradiated absorption spectra showed peaks at 520 nm in addition to the 450, 560, and 615 nm observed by Bagdonas et al.²² Excitation with 450 nm offers less tissue scattering and deeper penetration than 405 nm, which is commonly used in PDD.¹¹ Furthermore, Ppp is less susceptible to photodegradation than PpIX;^22,24,25 thus, there may be an extended fluorescence emission time with fluorescence photoswitching. Based on the wavelength of the excitation light, the photobleaching of Ppp differs. The Ppp photobleaching rate is slower than that of PpIX when excited at 410 nm, but a rate similar to that of PpIX, when excited at 440 nm, was reported.²⁶ Another report showed that the irradiation of $8.5\mathrm{J}/{\mathrm{cm}}^2$ at $140\mathrm{mW}/{\mathrm{cm}}^2$ was needed to halve the concentration of PpIX at 635 nm, whereas $27\mathrm{J}/{\mathrm{cm}}^2$ was needed for a similar effect in Ppp at 670 nm.²⁴

2.2.

Sample Preparation for Spectroscopic Analysis

PpIX (P8293, Sigma-Aldrich) was diluted to $10\mu \mathrm{M}$ with dimethyl sulfoxide (DMSO, D4540, Sigma-Aldrich), and $60\mu \mathrm{L}$ of the diluted PpIX was placed per well in a clear-bottomed black 96-well plate (353219, Falcon). The samples were exposed to a green light-emitting diode (LED) with a nominal wavelength of 505 nm and a bandwidth (full-width at half-maximum; FWHM) of 30 nm (M505L3, Thorlabs). The wells were irradiated following the irradiation method described in Sec. 2.3 with an increasing energy density of 0, 10, 20, 30, 40, and $50\mathrm{J}/{\mathrm{cm}}^2$ for power densities of 20 and $50\mathrm{mW}/{\mathrm{cm}}^2$. Each of the irradiated samples was further irradiated with an energy density of $10\mathrm{J}/{\mathrm{cm}}^2$ using an LED with a nominal wavelength of 455 nm and a bandwidth (FWHM) of 18 nm (M455L3, Thorlabs), with the same power density used in the green-light irradiation. The irradiation conditions for the solution analysis were varied to determine the effect of irradiance on the photoproduct formation rate.

2.3.

Irradiation

The samples in the 96-well plate, having a bottom area of $0.32{\mathrm{cm}}^2$ per well, were irradiated from the bottom, with the LED light irradiated through a collimation adapter (COP1-A, Thorlabs). The light was passed through an iris diaphragm (IH-50R-N, Sigmakoki, Japan) with a diameter set to 0.64 cm before being projected to the bottom of the well to keep the light beam diameter fixed for each well. The irradiation powers were set using a tuneable LED driver (LEDD1B, Thorlabs) and a variable neutral density filter (NDHN-50, Sigmakoki, Japan). The LED power was measured using a power sensor with a filter (PD-300 UV, Ophir Optronics, Israel) attached to a display (Nova II, Ophir Optronics). The energy density was set as a function of time, with each sample irradiation time controlled by changing the position of the well plate using a two-axis motorized linear stage (SGSP26-150(XY), Sigmakoki, Japan) and a two-axis stage controller (SHOT-702, Sigmakoki). The well plate temperature was maintained at 37°C during light exposure using a microwarm plate (KM-1, As One, Japan) to mimic the human body temperature.

2.4.

Fluorescence Spectroscopy Analysis

The fluorescence emission from the PpIX solution in Sec. 2.2 was measured after irradiation using a fluorescence microplate reader (Spectra Max Gemini EM, Molecular Devices) at excitation wavelengths of 450 and 505 nm.

2.5.

Optical Setup for Fluorescence Imaging

Figure 2 shows the setup used in the ex vivo and in vivo fluorescence imaging of the fluorophore. The fluorescence emission images from the irradiated fluorophore were taken using a CMOS monochrome camera (Chameleon 3, CM3-U3-31S4M, FLIR; with a camera lens ${\mathrm{L}}_1$ $f=16\mathrm{mm}$, $F/1.4$, NMV-16M1, Navitar), fitted with a 600 nm long-pass filter, ${\mathrm{F}}_1$ (FELH0600, Thorlabs) to cut off the excitation light. The exposure time for obtaining the fluorescence images was adjusted between 180 and $5000\mu \mathrm{s}$ based on the ex vivo or the in vivo experiments. The images were obtained for 455 and 505 nm excitations using the 455 and 505 nm LEDs described in Sec. 2.2, with the samples placed on a surface 26 mm from the longpass filter, ${\mathrm{F}}_1$. The LEDs used in the excitation were irradiated through a lens, ${\mathrm{L}}_2$ (ACL2520U-DG15-A, Thorlabs), and bandpass filters with an FWHM of 25 nm, ${\mathrm{F}}_2$ (86-654, Edmund Optics) and ${\mathrm{F}}_3$ (86-653, Edmund Optics), for 505 and 455 nm LED, respectively. The LEDs with the bandpass filters, ${\mathrm{F}}_2$ and ${\mathrm{F}}_3$, resulted in center wavelengths of 500 and 450 nm, respectively. The LEDs were angled 55 deg from the camera and placed approximately equal distances from the sample plate and the camera.

Fig. 2

Optical setup for the fluorescence imaging.

2.6.

Sample Preparation for Fluorescence Imaging

PpIX (P8293, Sigma-Aldrich) was diluted to a concentration of 1 mM with DMSO (D4540, Sigma-Aldrich). The solution was further diluted to $10\mu \mathrm{M}$ with clear epoxy resin (87136, Tamiya, Japan) and was heated in a water bath at 80°C for $\sim 1\min$ to deaerate the mixture. The solution was then poured into a petri dish to a height of 1 mm and was left to harden for more than 48 h. PpIX pellets with a diameter of 3 mm were cut out from the hardened epoxy resin sheet. Similar pellets were prepared with the PpIX solution replaced with DMSO. The pellets were irradiated with the green-light LED using the irradiation method described in Sec. 2.3. The power density for the irradiation was $20\mathrm{mW}/{\mathrm{cm}}^2$ with energy densities of 0, 2, 5, and $10\mathrm{J}/{\mathrm{cm}}^2$. The irradiance and the fluence were chosen to suit the experimental condition, as the initial concentration and the volume of the fluorophore can also affect fluorescence emission intensity and the fluorescence photoswitching of the fluorophore.

2.7.

Fluorescence Imaging: Pellets and Ex Vivo Analyses

The fluorescence images of the pellets irradiated outside of the mucosa were obtained with the setup in Fig. 2. The fluorescence images were obtained with the LED power density set to $1\mathrm{mW}/{\mathrm{cm}}^2$ and the camera exposure time of $180\mu \mathrm{s}$. Fluorescence images were also taken from a DMSO pellet for each excitation and were used as a background image.

Porcine stomach purchased from IVTeC Co., Ltd. (Tokyo, Japan) was cut, rinsed in saline to remove mucus, wrapped in aluminum foil, sealed, and frozen at $-25°\mathrm{C}$ until use. For the experiment, the porcine mucosa tissue was sliced at an arbitrary depth using a cryotome (CM1850, Leica, Germany), and the PpIX pellet was placed between the upper and the lower slices. The tissue with the pellet was irradiated with the setup in Fig. 2 using the 505 nm LED at a power density of $2\mathrm{mW}/{\mathrm{cm}}^2$. Fluorescence images were taken after 0, 0.5, 1, 1.5, 2, 2.5, 5, 6, 8, and $10\mathrm{J}/{\mathrm{cm}}^2$ of irradiation for excitations at 455 and 505 nm. The irradiance and the fluence used were chosen to suit the experimental condition, and the parameters used in capturing the images were the same for both excitation wavelengths to allow for a direct comparison.

After obtaining the fluorescence images, the tissue was sectioned to a thickness of $60\mu \mathrm{m}$ with the cryotome, and the section was analyzed with a slide scanner (NanoZoomer 2.0 RS, Hamamatsu Photonics, Japan) to determine the mucosal thickness at which the PpIX pellet was placed. The depth was determined by averaging the measured depths at 20 different points.

2.8.

Mouse Preparation

All procedures involving mice were performed in compliance with the institutional guidelines and regulations and reviewed by the animal experiment and welfare committee of Kochi Medical School.

A seven-week-old female BALB/c nu/nu mouse was housed in a plastic cage with stainless steel grid tops in an air-conditioned room with a 12 h light-dark cycle maintained at room temperature; and it was provided with water and food in the institute for animal experiments of Kochi Medical School.

The human prostate cancer cell line, PC-3, was obtained from Isaiah J. Fidler of the M. D. Anderson Cancer Center, Department of Cell Biology, University of Texas (Houston, Texas). The PC-3 cells were maintained in Dulbecco’s modified Eagle medium (D-MEM: 12100-046, GIBCO) supplemented with 10% fetal bovine serum (Hyclone, SH30910.03, Cytiva, Japan) and kept at 37°C under 5% ${\mathrm{CO}}_2$. The PC-3 cells ($2\times {10}^6$) suspended in $100\mu \mathrm{L}$ of D-MEM were injected into the dorsal region at the base of the tail of the mouse subcutaneously. The mouse was used in the experiment after 2 weeks.

A solution of aminolevulinic acid hydrochloride (ALA HCl; Al-05-1, Neopharma, Japan), diluted with distilled water, was injected intraperitoneally at a dose of $600\mathrm{mg}/\mathrm{kg}$ into the tumor-bearing mouse, and the PDD was performed 2 h after ALA HCl administration.

2.9.

Fluorescence Imaging In Vivo

The skin of the mouse of Sec. 2.8 emitted high fluorescence, which interfered with the fluorescence emission from the lesion. Similar interference has been previously reported.^27,28 To specifically observe the fluorescence emitted from the lesion, the skin surrounding the lesion was cut open to expose the lesion, and PDD with 505 nm was performed using the setup in Fig. 2. During lesion exposure to irradiation, the images of the lesion excited at 455 and 505 nm were taken at intervals of $2.4\mathrm{J}/{\mathrm{cm}}^2$ for a total energy density of $9.6\mathrm{J}/{\mathrm{cm}}^2$. The power density of the LEDs used for the PDD and imaging was $\sim 4\mathrm{mW}/{\mathrm{cm}}^2$.

3.1.

Fluorescence Spectroscopic Analysis

The fluorescence emission of PpIX solution of Sec. 2.2 before light irradiation, excited with green and blue light, is shown in Fig. 3(a). The fluorescence emission intensity from PpIX excited with blue light and green light is defined as $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$, respectively. Prior to irradiation, the fluorescence intensity of $I_{\mathrm{G}}$ was higher than that of $I_{\mathrm{B}}$, with both spectra having their maximum peak at $\sim 633\mathrm{nm}$. After the irradiation [Fig. 3(b)], there was a photodegradation of the 633 nm peak, with an emergence of a peak at around 673 nm. The new peak indicates the formation of a photoproduct. The fluorescence intensity at 633 nm and the intensity at 673 nm were compared before and after light irradiation. The irradiation resulted in a higher $I_{\mathrm{B}}$ than $I_{\mathrm{G}}$ at the spectra peak of 673 nm. Furthermore, $I_{\mathrm{B}}$ at 673 nm was higher than both $I_{\mathrm{G}}$ and $I_{\mathrm{B}}$ at 633 nm. This indicates the occurrence of fluorescence photoswitching. At this point, the $I_{\mathrm{G}}$ at 633 nm reduced to 0.45 times its initial value. In contrast, $I_{\mathrm{B}}$ at 673 nm increased to 13.7 times the intensity prior to light irradiation, indicating the accumulation of the photoproduct.

Fig. 3

Fluorescence spectra of PpIX (a) unirradiated and (b) irradiated with a 505 nm LED, excited at 450 and 505 nm. The irradiation power density was $20\mathrm{mW}/{\mathrm{cm}}^2$, and the energy density was $50\mathrm{J}/{\mathrm{cm}}^2$ for the irradiation. The error bars represent standard deviations from three independent experiments.

The changes in the fluorescence intensity with irradiation were monitored at the wavelength peaks of 633 (PpIX) and 673 (photoproduct) nm as the irradiation energy density was increased (Fig. 4). The fluorescence intensity prior to irradiation is assumed to contain only the peaks of PpIX. After irradiation, the spectrum is modified to contain both the peaks of PpIX and the photoproducts. For the analysis of the formation of the photoproduct, the fluorescence intensity at 673 nm was corrected to remove that of PpIX using the method described by Bagdonas et al.²² For the two power densities investigated, the pattern of decay of the PpIX fluorescence intensity was similar. On the other hand, the rate of increase of the photoproduct was faster for the lower power density. The fluorescence photoswitching of the fluorescence intensities was observed after the energy densities of 30 and $39\mathrm{J}/{\mathrm{cm}}^2$ for 20 and $50\mathrm{mW}/{\mathrm{cm}}^2$, respectively.

Fig. 4

Fluorescence intensity of PpIX exposed to 505 nm, excited at 505 and 450 nm, monitored at 633 and 673 nm at irradiation power densities of (a) 20 and (b) $50\mathrm{mW}/{\mathrm{cm}}^2$. Fluorescence intensity of PpIX in (a) and (b) further irradiated with blue light with an energy density of $10\mathrm{mW}/{\mathrm{cm}}^2$, excited at 505 and 450 nm, monitored at 633 and 673 nm at irradiation power densities of (c) 20 and (d) $50\mathrm{mW}/{\mathrm{cm}}^2$. The fluorescence intensities of $I_{\mathrm{B}}$ at 673 nm were corrected to remove the fluorescence intensity of PpIX from that of the photoproduct. The point at which $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$ cross indicates the point of fluorescence photoswitching. The error bars represent standard deviations.

Further irradiation of each 505 nm irradiated PpIX with 455 nm was performed to determine how the photoproducts react when the excitation light is changed during PDD [Figs. 4(c) and 4(d)]. Irradiation with the blue light led to further PpIX photobleaching for both power densities. Comparing the PpIX photobleaching for each increment of $10\mathrm{J}/{\mathrm{cm}}^2$ between the green and blue light irradiation, the photobleaching was greater for the blue light than for the green light irradiation. Nonetheless, the PpIX photobleaching remained comparable between the two power densities even when the wavelength of the light source was changed. Considerable fluorescence intensity from the photoproduct was maintained after further blue-light irradiation for all energy densities, with each increment in energy density increasing the photoproduct’s fluorescence intensity. However, the fluorescence increase from the blue-light irradiated photoproduct was not as high as in the green-light irradiation.

3.2.

Fluorescence Imaging in Pellets

Figure 5(a) shows the images of 505 nm irradiated pellets. Unlike the spectroscopic analysis, in which the fluorescence intensity of PpIX and the photoproduct can be distinguished to some extent, the irradiation of the pellet excites both PpIX and the photoproduct. The increase in the fluorescence intensity of $I_{\mathrm{B}}$ can be observed with irradiation. A spectroscopic analysis of the irradiated pellets showed the emergence of the photoproduct peak at 673 nm. Hence, the increase in the fluorescence intensity with pellet irradiation should result from the formation of photoproducts. Fluorescence photoswitching was observed after $2\mathrm{J}/{\mathrm{cm}}^2$. This was the minimum irradiation time investigated.

Fig. 5

(a) Images of 505 nm irradiated PpIX pellets, and (b) a plot of fluorescence intensity of 505 nm irradiated PpIX pellets with increasing energy density. The error bars represent standard deviations from three independent experiments. $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$ represent the fluorescence emission intensity from the pellets excited at 455 and 505 nm, respectively.

The fluorescence intensity values from the pellets were plotted for easier comparison [Fig. 5(b)]. The saturations observed at the edges of the pellets were excluded from the values used in the comparison, and the background intensity was subtracted from the fluorescence images of the PpIX pellets. The unirradiated $I_{\mathrm{G}}$ and $I_{\mathrm{B}}$ were used to normalize each value. An increase in both $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$ can be seen after irradiation [Fig. 5(b)], caused by the excitation of both PpIX and the photoproduct at both excitation wavelengths. However, the rise in the intensity of $I_{\mathrm{B}}$ was higher than that of $I_{\mathrm{G}}$, with the intensity rising to double its initial intensity. For both excitations, the maximum fluorescence intensity occurs at $5\mathrm{J}/{\mathrm{cm}}^2$, after which the photodegradation of the fluorophore is observed. Photoproducts are also susceptible to photobleaching.^21,24 Thus, the decrease in the fluorescence intensity of the pellet observed around $10\mathrm{J}/{\mathrm{cm}}^2$ exposure could be caused by both PpIX and photoproduct photobleaching.

3.3.

Fluorescence Imaging Ex Vivo

The fluorescence emission images from the fluorophore within the mucosa exposed to green-light are shown in Fig. 6(a). Figure 6(b) shows a plot of the fluorescence intensity of the pellet against the irradiation energy density for each excitation. The mucosa depth at which the fluorophore was placed for this analysis was calculated to be $1.28\pm 0.27\mathrm{mm}$ from the surface. The tissue fluorescence for both 455 and 505 nm excitations was subtracted accordingly from that of the pellet to account for the background fluorescence effect. The fluorescence emitted from the fluorophore [Fig. 6(a); $0\mathrm{J}/{\mathrm{cm}}^2$] prior to irradiation shows that the intensity was slightly higher for $I_{\mathrm{B}}$ than for $I_{\mathrm{G}}$. $I_{\mathrm{B}}$ was 1.07 times higher than $I_{\mathrm{G}}$ after the background fluorescence was subtracted. The increase in $I_{\mathrm{B}}$ was observed after $0.5\mathrm{J}/{\mathrm{cm}}^2$ of irradiation. The photodegradation observed in $I_{\mathrm{G}}$ was as low as 0.74 times its initial fluorescence intensity, whereas $I_{\mathrm{B}}$ increased up to 2.8 times its initial fluorescence intensity. The maximum emission from $I_{\mathrm{B}}$ occurred after $8\mathrm{J}/{\mathrm{cm}}^2$. Thereafter, a reduction in the fluorescence intensity was observed. It is difficult to increase the sampling number of this investigation, as cutting the mucosa at the same depth to insert the pellet and having mucosa tissue with similar surface regularity is difficult. Nonetheless, a continual rise in the fluorescence intensity up to $7\mathrm{J}/{\mathrm{cm}}^2$ was observed in another investigation that was performed.

Fig. 6

(a) Fluorescence emission images from the fluorophore within the mucosa, with increasing irradiation energy density. (b) A plot of fluorescence intensity against irradiation energy density of fluorophore within the mucosa. The fluorophore was exposed to green light and excited at 455 and 505 nm. $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$ represent the fluorescence emission intensity from the fluorophore excited with blue light and green light, respectively.

The excitation light sources used in the fluorescence imaging (Fig. 2) have an excitation bandwidth (FWHM) of 25 nm. The additional bandwidth extends the wavelength for the molar excitation coefficient of the fluorophore for each excitation wavelength. Thus, the slightly higher fluorescence intensity of the unirradiated fluorophore seen in $I_{\mathrm{B}}$ [Figs. 5(a) and 6(a)] may be attributed to the extended excitation of PpIX. Because the fluorescence comparison is relative to the unirradiated fluorescence intensity for the individual excitation, the extended excitation of PpIX in both $I_{\mathrm{G}}$ and $I_{\mathrm{B}}$ should pose no issue.

3.4.

Fluorescence Imaging In Vivo

The fluorescence emission images of the ALA-administered lesion on a mouse with increased exposure to 505 nm excitation are shown in Fig. 7. The background fluorescence emission from the tissue was subtracted from the images. The fluorescence images were normalized using the fluorescence emission at $0\mathrm{J}/{\mathrm{cm}}^2$ exposure for each excitation wavelength and are displayed in a color map for easier visualization of the changes in the fluorescence intensity.

Fig. 7

In vivo fluorescence emission images of the ALA HCl administered-PC-3 tumor-bearing mouse. The fluorescence images were obtained at time intervals with increased exposure to 505 nm light. At each time interval, images were obtained for green ($I_{\mathrm{G}}$) and blue ($I_{\mathrm{B}}$) excitation wavelengths.

Similar to the ex vivo analysis results, the initial fluorescence emission in $I_{\mathrm{B}}$ is slightly higher than in $I_{\mathrm{G}}$ due to the fluorophore’s extended excitation wavelength to include part of the Q-band of PpIX. This should pose no issue in the result interpretation, considering that the increase in the fluorescence intensity (Fig. 7) is relative to the initial fluorescence emission ($0\mathrm{J}/{\mathrm{cm}}^2$ exposure) for each excitation. For both $I_{\mathrm{B}}$ and $I_{\mathrm{G}}$, the increase in the fluorescence intensity of the lesion is seen with increased irradiation energy density after $2.4\mathrm{J}/{\mathrm{cm}}^2$. The increase occurs up to $4.8\mathrm{J}/{\mathrm{cm}}^2$, after which the fluorescence intensity begins to decrease. It is worth noting that the continual generation of PpIX from the ALA administered can contribute slightly to the increase in the fluorescence intensity of the lesion.²⁹ However, the increase in the fluorescence intensity observed in Fig. 7 is likely caused by the formation of photoproducts. The fluorescence increase should be observed in the same part of the lesion for both blue- and green-light excitations if the increased fluorescence intensity was attributed to PpIX.