2.1.

Study Object

Experiments were performed on white outbred rats (a total of 20 animals) kept in the standard vivarium conditions. The studies were performed in accordance with the ethical principles established by the European Convention for the Protection of Vertebrate Animals used for experimental and other scientific purposes with the permission of the Local Ethical Committee of the Nizhny Novgorod State Medical Academy (Protocol no. 9, October 15, 2014). Before procedure, intraperitoneal anesthesia was performed with a Zoletil solution in a dose of 50 mg/kg, then the animals were fixed on a special plate. Irradiation was carried out in vivo using the ${\mathrm{Co}}^{60}$ unit “Terabalt” (UJP, Prague, Czech Republic) with a mean beam energy of 1.25 MeV, by a local field, by single doses equivalent to 2, 10, and 40 Gy, respectively. The exposure time was calculated in accordance with the size of the irradiation field, the depth, and the single dose. In 24 h, 1 week, and 1 month after exposure, the animals were withdrawn from the experiment using ether anesthesia.

Bladder and rectum were dissected and ex-vivo CP OCT study was performed immediately on the entire surface of the organs. The samples were constantly wetted with 0.9% NaCl solution to avoid drying. The preparations of the bladder and rectum of two unirradiated animals were used as controls. A part of the organ was separated for microscopic examination before the CP OCT procedure. The samples for the nonlinear microscopy and light microscopy were fixed for 24 h in a 10% buffered solution of neutral formalin. Then, the samples were embedded into paraffin (“Histomix-extra,” “BioVitrum,” Russia), followed by the mounting of paraffin blocks. The slices of 0.3- and $10\text{-}\mu \mathrm{m}$-thick were produced from the resulting blocks using the Leica 450 RM rotary microtome (Leica Microsystems, Germany). The slices of $0.3\text{-}\mu \mathrm{m}$-thick were stained with picrofuchsin according to van Gieson and were used to verify the results obtained by the nonlinear microscopy. The preparations were studied with a Leica DMLS microscope (Leica Microsystems, Germany). The samples of $10\text{-}\mu \mathrm{m}$-thick obtained from the same blocks were dewaxed and investigated using the method of the nonlinear microscopy without additional staining.

2.2.

Nonlinear Microscopy: Imaging and Processing

Studies of unstained slices of bladder and rectum samples were performed on an inverted laser scanning microscope LSM Axiovert 510 Meta (Carl Zeiss, Germany). A short-pulse femtosecond laser MaiTai HP (Spectra Physics) with a pulse repetition rate of 80 MHz and a duration of about 100 femtoseconds was used. Deparaffined slices placed on a slide were covered with a cover glass of $170\text{-}\mu \mathrm{m}$ thick. The images of the structure were constructed with the oil-immersion lens EC Plan-Neofluar with 40-fold magnification and a numerical aperture of 1.3, which made it possible to obtain a field of $318\times 318$ microns with a resolution of $1024\times 1024\text{pixels}$. The following spectral characteristics of the visualization system were used: excitation at a wavelength of 800 nm, detection in the wavelength range of 362 to 415 nm, containing a peak of the nonlinear response of collagen at the second harmonic (400 nm).¹³ Registration was carried out with the fully open pinhole.

To quantify the features of the collagen fibers, the regions of interest (ROI) on the submucosal layer were selected (Fig. 1). The areas corresponding to the quality criteria (the absence of blood vessels, epithelium, and muscle fibers in the area of interest and the absence of dark fields corresponding to areas that were out of focus) were outlined. In total, 109 images of the bladder and 145 images of the rectum were obtained. In total, 265 ROI on the bladder SHG-images and 343 ROI on the rectum SHG-images were analyzed.

Fig. 1

Examples of ROI selection. (a) Intact bladder SHG-image, (b) intact rectum SHG-image, and (c) contours.

The parameter of the SHG signal intensity, which represents the average SHG signal intensity on all informative pixels in the image for the selected ROI, has been used as a quantitative criterion for assessing the structure of collagen depending on the dose and time after radiation exposure. The average intensity of the intact bladder/intact rectum SHG signal was obtained as a result of averaging the intensity values from a set of ROI. All the values obtained were normalized to the average value of the intensity of the SHG signal of intact samples taken as unity. The average value of the SHG signal intensity from all the analyzed ROI was calculated for each dose and time after irradiation, the result being presented as mean ± SD. Calculations were made in the ImageJ 1.39p software (NIH, USA). All tissue samples were fixed uniformly in a neutral formalin; the values of the signal intensity distribution of every sample were normalized to the average value of the intensity of the SHG signal of intact samples taken as unity. There are data that formalin fixation could affect a collagen structure.³³ Although the protocol of tissue processing was the same for all samples, we suggest that the resulting changes did not influence on the relative values of the signal intensity distribution on the nonlinear microscopy images. All data were obtained from the slices of the same thickness at the same parameters of the pump power, digital amplification, and the dynamic range of the recorded signal.

2.3.

Cross-Polarization Optical Coherence Tomography: Imaging and Processing

The cross-polarization OCT was used to assess the changes in the structure of biological tissues at the level of general organs (bladder and rectum) architectonics. The device “OCT 1300-U” (IAP RAS, Nizhny Novgorod, Russia) was used.³⁴

The wavelength of the superluminescent diode used as the radiation source was 1315 nm, the spectral width was 60 nm, which determined the spatial resolution at a depth of $20\mu \mathrm{m}$. The tomograph is equipped with a flexible fiber-optic probe with an end-window of an optical scanner of 2.7 mm in diameter, with a lateral resolution of $25\mu \mathrm{m}$. When scanning the ROI, the probe was brought into a contact with the object of investigation. Before OCT-scanning, the dissected organs were spread and fixed on the plate. OCT imaging was performed “step by step” all over the organ surface. From every organ, at least 15 images were obtained. The averaged data from all obtained images for every time and dose point have been presented in the paper.

The CP OCT system has a common-path optical layout and two signal acquisition channels: one for scattered light that maintained initial polarization (“copolarized”), and one that now exhibits orthogonal polarization (“cross polarized”).²³ The size of the CP OCT image in each polarization is 200 points in the transverse direction and 256 points in depth, which corresponds to $2.2\times 1.3{\mathrm{mm}}^2$. The brightest areas of the CP OCT image represented in the brown palette correspond to the maximum intensity signal, and the darkest ones correspond to the signal with minimum intensity.³⁴

A dimensionless parameter, the integral depolarization factor (IDF), which is the ratio of the received OCT signal powers, averaged over all informative pixels of the CP OCT image, in the orthogonal and original channels.³¹ The amplitude of the OCT signal in the orthogonal polarization depends on the degree of inverse cross scattering of the probe radiation by the tissues that are studied.³⁵ The main structural cross-scattering components are collagen fibers. The degree of cross-scattering depends on their structural and spatial organizations. Therefore, the IDF magnitude can be used to judge the structure of collagen fibers in the tissue, and, consequently, to assess the degree of radiation-induced damage of the organs’ extracellular matrix.

2.4.

Statistical Analysis

Statistical analysis was performed by ImageJ 1.39p (NIH) and Microsoft Excel 2007 software. The intact samples and the samples obtained at various times after irradiation (a day, a week, and a month) by the doses of 2, 10, and 40 Gy were compared using the one-tailed student’s t-test in Statistica 6.0 software. $P$ values $<0.05$ were considered to be statistically significant. Correlation analysis was performed between the mean values of the intensity of the SHG signal determined by a nonlinear microscopy method and the values of IDF were determined by CP OCT for bladder and rectum separately. A correlation cloud was created and linearly approximated by the Prism software (GraphPad Software Inc., La Jolla). The Spearman’s rank correlation coefficient and the significance level were calculated.

3.1.

Results of the Study of Radiation-Induced Changes in Bladder Collagen

On the images of intact rat bladder obtained using the nonlinear microscopy in an SHG detection regime, the main localization of collagen fibers were the urothelium basal membrane, the submucosa, the walls of the blood vessels, and the framework of the muscle layer [Fig. 1(a)].³⁶

No changes in the extracellular matrix were detected during the visual assessment of images of the SHG signal distribution of collagen of the bladder wall after exposure to a dose of 2 Gy irrespective of the time after irradiation (Fig. 2). The main manifestations of radiation damage to collagen on the images after irradiation at doses of 10 and 40 Gy were swelling of the submucosal layer, fragmentation of collagen fibers, destruction of connective-tissue structures, and more pronounced after exposure to a dose of 40 Gy (Fig. 2).³⁶

Fig. 2

Images of the bladder depending on the dose and time after irradiation. Left columns—SHG-images (nonlinear microscopy in SHG detection regime, image size $318\times 318\mu \mathrm{m}$, oil-immersion); right columns—van Gieson staining ($40\times$ objective magnification).

A clear dependence of the SHG signal intensity of the bladder wall collagen on the dose and the time after irradiation was observed after the quantitative assessment of SHG-images. After irradiation at a dose of 2 Gy, the values of the intensity of the SHG signal did not differ from the intact bladder samples irrespective of the term after irradiation. One day after irradiation, the decrease in the intensity of the SHG signal of collagen after irradiation in doses of 10 and 40 Gy, respectively, was observed (Fig. 3). Statistically significant difference ($p=0.05$) compared with the intact sample was revealed after irradiation at a dose of 40 Gy.

Fig. 3

Dynamics of the SHG signal of collagen depending on the dose and time after radiation exposure. Normalized mean intensity of the SHG signal in ROI of the bladder. ^*Statistically significant difference from the initial level.

A week after the exposure, a statistically significant decrease in the intensity of the SHG signal to $0.67\pm 0.12$ a.u. after irradiation at a dose of 10 Gy and to $0.52\pm 0.11$ a.u. after 40 Gy irradiation remained in comparison with the intact sample ($1\pm 0.13$ a.u.) (Fig. 3). A month after irradiation at a dose of 10 Gy, the intensity of the SHG signal was almost restored to the level of the intact sample—$0.91\pm 0.14$ a.u., whereas the signal intensity of the samples irradiated at a dose of 40 Gy did not even show a tendency to return to the initial level and amounted to $0.47\pm 0.15$ a.u. (Fig. 3).

The four layers can be differentiated on OCT images of intact bladder samples of rats in the initial polarization, corresponding to the layers on histological image: the upper layer with a moderate signal level corresponding to the urothelium; medium layer with a high signal level corresponding to the connective-tissue layer (proper mucous plate and submucosal layers); the next layer with a moderately intense signal corresponding to the muscle layer; and the lowest layer corresponding to the outer adventitious membrane [Figs. 4(a) and 4(c)]. In the orthogonal polarization, in these zones of the bladder, a distinct signal is clearly determined from collagen fibers in the proper mucous plate, the submucosa layer, and the adventitious membrane. The urothelium does not have a visible OCT signal as the transitional epithelium does not change the polarization of the probing radiation [Fig. 4(b)].

Fig. 4

OCT-image of intact bladder: (a) initial polarization, (b) orthogonal polarization, (c) standard histological image of bladder, HE staining. ($U$, urothelium; $S$, submucosa, and lamina propria; $M$, muscle; and $A$, adventitia)

There have been no visible changes in the structure identified after any dose expose, but the numerical image processing revealed changes (see Figs. 5 and 7). The IDF value of the intact bladder was $0.067\pm 0.01$ a.u. After irradiation at a dose of 2 Gy, the changes in the IDF were not statistically significant irrespective of the period after irradiation (Fig. 5). A day after irradiation of the bladder in doses of 10 and 40 Gy, respectively, the IDF values reduced to $0.053\pm 0.01$ and $0.05\pm 0.01$, respectively. A week after irradiation in doses of 10 and 40 Gy, the IDF values remained lower, which indicated the continuing processes of the connective-tissue matrix degradation. An increase in the values of the IDF almost to the initial level was observed 1 month after the exposure at a dose of 10 Gy. There was a pronounced decrease in the IDF to $0.029\pm 0.01$ 1 month after the radiation exposure at a dose of 40 Gy (Fig. 5). Statistically significant differences compared with the unirradiated control were detected in a day, a week, and a month after a radiation exposure in the doses of 10 and 40 Gy, respectively.

Fig. 5

Mean IDF depending on the dose and time after radiation exposure. ^*Statistically significant difference from the initial level.

The analysis of the data showed a high positive correlation (the Pearson correlation coefficient was 0.855 at a significance level of $p<0.005$) between the values of the SHG signal intensity and the IDF values of the bladder wall. The linear approximation showed a satisfactory value of the approximation reliability ($R^2>0.7$). This indicates that CP OCT with the bladder tissue IDF calculation can be used as a method for verifying the dynamics of changes in collagen structures after the ionizing radiation exposure.

3.2.

Results of the Study of Radiation-Induced Changes in Rectal Collagen

The basal membrane, a proper mucous plate, submucosa and walls of blood vessels, and the stromal skeleton muscle sheath are the main allocations of connective-tissue structures in intact rat rectum [Fig. 1(b)].

Visual changes in the rectum wall collagen arising from 1 day up to 1 month after exposure in the studied doses were similar to the changes in the bladder wall collagen structures (Fig. 6).

Fig. 6

Images of the rectum depending on the dose and time after irradiation. Left columns—SHG-images (nonlinear microscopy in SHG detection regime, image size $318\times 318\mu \mathrm{m}$, oil-immersion); right columns—van Gieson staining ($40\times$ objective magnification).

Quantitative analysis of the SHG images of rectum samples showed a decrease in the intensity of the collagen SHG signal to $0.86\pm 0.18$ a.u. a day after irradiation at a dose of 10 Gy and to $0.74\pm 0.04$ a.u. after 40-Gy irradiation (Fig. 7). The maximum decrease in the signal intensity was observed after exposure to a dose of 40 Gy (difference in comparison with the intact sample was statistically significant, $p=0.05$). A week after the exposure, the decrease in the SHG signal remained, having statistically significant differences compared with unirradiated control after exposure to a dose of 10 Gy ($0.58\pm 0.2$ a.u.) and 40 ($0.48\pm 0.03$ a.u.), respectively. A tendency of the SHG signal intensity returning to the original level was observed a month after exposure to a dose of 10 Gy, and signal intensity of the samples irradiated in a dose of 40 Gy remained significantly reduced: $0.45\pm 0.02$ a.u. Irradiation in a dose of 2 Gy did not significantly affect the SHG signal intensity irrespective of the period after exposure.

Fig. 7

Dynamics of SHG signal of collagen depending on the dose and time after radiation exposure. Normalized mean intensity of SHG signal in ROI of the rectum. ^*Statistically significant difference from the initial level.

Four layers are differentiated on OCT images of intact rectum samples in the initial polarization, corresponding to the layers on a histological image: the upper layer with a moderate signal level corresponding to the epithelium; the second with a high signal level corresponding to the submucosal layer; the third one with a moderate signal level—a muscular layer; the lower thin layer with an uneven border—a serous layer [Figs. 8(a) and 8(c)]. In the orthogonal polarization, in these zones of the rectum, a distinct signal is clearly determined from collagen fibers in the submucosa layer and the serous membrane. In this case, the epithelium does not demonstrate a visible signal [Fig. 8(b)].

Fig. 8

OCT-image of intact rectum: (a) initial polarization, (b) orthogonal polarization, (c) standard histological image of rectum, HE staining. ($E$, epithelium; $S$, submucosa; and $M$, muscle)

No visible changes in the structure were revealed on CP OCT images of rectum samples obtained at various times after irradiation in the studied dose range. After the quantitative analysis of CP OCT images of the intact rectum, the mean IDF value was $0.058\pm 0.01$. A day after irradiation of the rectum in doses of 2, 10, and 40 Gy, the IDF values reduced to $0.051\pm 0.01$, $0.044\pm 0.01$, and $0.046\pm 0.01$, respectively. In a week and a month after irradiation in doses of 2 and 10 Gy, the IDF values slightly increased. After a dose expose of 40 Gy, the IDF values decreased progressively and amounted to $0.38\pm 0.01$ in a week and $0.034\pm 0.01$ in a month, which indicated the persistent connective-tissue matrix degradation processes (Fig. 9). There were statistically significant changes compared with intact control in a day after 10-Gy exposure, and in a day, a week, and a month after irradiation in the dose of 40 Gy.

Fig. 9

Mean IDF depending on the dose and time after radiation exposure. ^*Statistically significant difference from the initial level.

The analysis of the data showed a high positive correlation between the values of the SHG signal intensity and the IDF values (the value of the linear Pearson correlation coefficient was $p=0.813$, with a significance level of $p<0.005$). The linear approximation showed a satisfactory value of the approximation reliability ($R^2>0.6$).