2.1.

Renal Ischemia

All animal procedures were approved by the University of California, Davis Animal Use and Care Committee. Adult male Wistar rats weighing $300\text{to}500\mathrm{g}$ were anesthetized using $100\text{‐}\mathrm{mg}∕\mathrm{kg}$ sodium pentobarbital $(50\mathrm{mg}∕\mathrm{mL})$ injected intraperitoneally. Laparotomy was performed through a midline abdominal incision. Both kidneys and their respective vascular pedicles were then dissected free of surrounding fascia and fatty tissue, and mobilized anteriorly into the laparotomy wound so they could be imaged by an external camera. Care was taken to avoid twisting the vascular pedicle and causing premature ischemia in either kidney.

Unilateral warm renal ischemia was induced by left renovascular pedicle clamping and then studied by analyzing tissue autofluorescence under excitation with laser light of different wavelengths. Reperfusion was studied in vivo after unclamping of the renovascular pedicle. In all groups, the right kidneys were left uninjured and used as internal controls. At the conclusion of the imaging studies, all rats were sacrificed. Ischemia times in the experimental groups were chosen based on the previous demonstration by Calman and Bell that rat kidneys clamped for less than $50\min$ recovered, while those clamped for greater than $70\min$ did not.⁵

Renal ischemia time in group 1 ( $n=4$ rats) was $18\min$ , followed by unclamping and a reperfusion phase of $18\min$ . Images were acquired under a laser excitation wavelength of $335\mathrm{nm}$ . Renal ischemia in group 2 ( $n=4$ rats) was $85\min$ , followed by unclamping and a $35\text{‐}\min$ reperfusion phase. Images were acquired under a laser excitation wavelength of $335\mathrm{nm}$ . Renal ischemia time in group 3 ( $n=3$ rats) was $85\min$ , followed by unclamping and $35\min$ of reperfusion. Images were acquired under both 335- and $260\text{‐}\mathrm{nm}$ excitation during the warm ischemia and reperfusion phase. In group 4( $n=2$ rats), we monitored the autofluorescence intensity under $260\text{‐}\mathrm{nm}$ excitation during $20\min$ of renal ischemia, followed by a $20\text{‐}\min$ reperfusion phase. In group 5 ( $n=4$ rats), we perform cross-polarized light scattering imaging measurements at $650\mathrm{nm}$ to monitor the change of hemoglobin absorption during $20\min$ of ischemia time, followed by $30\text{‐}\min$ of reperfusion. In group 6 ( $n=2$ rats), we studied kidney autofluorescence spectra using 335- and $260\text{‐}\mathrm{nm}$ excitation at the time of clamping $\left(0\min \right)$ and at $75\min$ of warm ischemia.

2.2.

Optical Spectroscopy Methods

In each experimental rat, both kidneys were exposed and a black, nonfluorescing cloth was used to cover the remainder of the rat’s body and viscera to avoid autofluorescence from other tissues. The anterior surfaces of both kidneys were simultaneously excited.

The kidneys in groups 1 and 2 were imaged using their native autofluorescence under exposure to $335\text{‐}\mathrm{nm}$ laser illumination. A set of lenses was used to enlarge the laser beam and its central portion was used to provide nearly uniform illumination of the kidneys. The laser illumination wavelength was obtained by second harmonic conversion of the output of an optical parametric oscillator (Opotek, Incorporated, Carlsbad, California) tuned at $670\mathrm{nm}$ . Excitation with $335\mathrm{nm}$ is known to penetrate tissue to a depth on the order of a few hundred $\mu m$ , depending on tissue type.⁶ Resulting autofluorescent images from each pair of kidneys were captured by a liquid nitrogen cooled charge-coupled device (CCD) array detector ( $512\times 512\text{pixels}$ , Roper Scientific, Trenton, New Jersey). A camera lens was used to project the images to the CCD detector through a $395\text{‐}\mathrm{nm}$ long wavelength pass filter placed in front of the camera. This combination of filter and detector captures images using the autofluorescence from the kidneys in the $\sim 400\text{‐}\text{to}1000\text{‐}\mathrm{nm}$ spectral region. This range was selected to optimize signal-to-noise ratio and enable short image acquisition times. Both excitation and emission spectral bands for image acquisition were chosen based on our previous experience using $335\text{‐}\mathrm{nm}$ excitation with multiple narrow band and long pass filters.⁴ Starting from the time of left renal vascular pedicle clamping, images were taken every $15\mathrm{s}$ for the duration of the ischemia and reperfusion phases. The intensity of the captured images was normalized to the intensity map of the photoexcitation beam, which was recorded by acquiring the fluorescence image of the beam imprint on a fluorescing paper. Normalization was performed by pixel-by-pixel division of autofluorescence images by the images of the beam profile on the fluorescing paper. Images were analyzed using commercially available computer software (Roper Scientific) to calculate the ratio of autofluorescence intensity of the ischemic to control kidneys. The software calculates the average pixel intensity over a rectangular area of the image. We used this feature to obtain the average intensity of the kidneys by covering as many pixels as possible of the digitized image of each kidney.

In groups 3 and 4, the same imaging system described above for groups 1 and 2 was used. Group 3 kidneys were imaged with 260- and $335\text{‐}\mathrm{nm}$ excitation, at 0, 18, and $85\min$ of ischemia. The $260\text{‐}\mathrm{nm}$ laser illumination wavelength was obtained from the same optical parametric oscillator (OPO) laser system. The goal of this set of experiments was to compare the optical spectroscopic properties of ischemic kidneys when imaged under different excitation wavelengths. We used the same spectral band imaging filter ( $395\text{‐}\mathrm{nm}$ -long wavelength pass) for both excitation wavelengths to facilitate their near-simultaneous use during the execution of the experiment. Our previous experience with probing different excitation wavelengths demonstrated that excitation of hypothermically preserved rat kidneys ex vivo with a laser tuned at $260\mathrm{nm}$ resulted in no changes (within experimental error) in intensity ratio between a normal kidney and a kidney that had undergone $60\min$ of warm ischemia prior to hypothermic preservation.⁷ In the current study, we examined whether ischemic kidneys, during in vivo measurements under $260\text{‐}\mathrm{nm}$ excitation, would display similar spectroscopic properties. To examine the same effect at shorter ischemia times, group 4 experiments utilized $260\text{‐}\mathrm{nm}$ excitation to monitor the autofluorescence from normal and injured kidneys during $20\min$ of renal ischemia time, followed by $20\min$ of reperfusion.

To monitor any changes in the absorption by hemoglobin during $20\min$ of ischemia and $20\min$ of reperfusion, normal and injured rat kidneys in group 5 were studied using the same experimental setup, with the exception that images were obtained with cross-polarized light scattering at $650\mathrm{nm}$ . The illumination was delivered to the kidneys using a fiber bundle that was coupled to a white light source. Both a linear polarized and a $650\pm 20\text{‐}\mathrm{nm}$ ( $40\text{‐}\mathrm{nm}$ bandwidth) narrow band interference filter were positioned at the output of the illumination fiber to deliver polarized and nearly monochromatic illumination. A second polarizer with its polarization orientation orthogonal to that of the illumination was positioned in front of the imaging lens to capture cross-polarized images of the kidneys.⁴ Images were captured every $5\mathrm{s}$ , starting from the time of left renal vascular pedicle clamping, for the duration of the ischemia and reperfusion phases. Images were analyzed using the same methods described earlier.

During ischemic injury, the absorption spectrum of hemoglobin changes due to transition from oxy- to deoxyhemoglobin. A reverse change is expected during reperfusion when fresh, oxygenated blood recirculates through the injured kidney. Cross-polarized imaging was used to: 1. eliminate the specular reflection component (which remains mostly polarized and bears no information from below the surface of the imaged organ); and 2. obtain images of kidneys generated by photons that have penetrated the tissue, thus offering optimal information regarding changes in the absorption within the substance of the kidney.⁸ We chose $650\mathrm{nm}$ for imaging because there is a large difference in the absorption coefficient between oxy- and deoxyhemoglobin in this part of the spectrum, allowing for visualization of the effect in an imaging mode. Increased deoxyhemoglobin concentration leads to more absorption of the $650\text{‐}\mathrm{nm}$ light and, if occurring during these experiments, would be expected to result in decreased intensity of the kidney images.

The autofluorescence spectra of kidneys from group 6 were recorded at $0\min$ and after $75\min$ of warm ischemia using a back-illuminated CCD detector ( $1340\times 400\mathrm{pixels}$ , Roper Scientific, Trenton, New Jersey) after the emission was spectrally analyzed using a single-gating spectrometer (TRIAX 320, Jobin Yvon Incorporated, Edison, New Jersey). The tunable output of the OPO laser was used for excitation with the laser wavelength alternately tuned at 335 or $260\mathrm{nm}$ to capture the spectra for various ischemia times at both these excitations. The autofluorescence was collected and focused into the slit of the spectrometer using reflective optical elements. To monitor spectral change independently of autofluorescence intensity, spectral profiles were normalized by dividing each by its peak intensity.

3.1.

Group 1: Imaging During $\mathit{18}\mathit{Min}$ of Ischemia and $\mathit{18}\mathit{Min}$ of Reperfusion ( $\mathit{335}\text{‐}\mathit{nm}$ Excitation)

Figure 1 shows a representative example of our experimental observations in group 1, depicting the injured-to-normal autofluorescence image intensity ratio over $18\min$ of warm ischemia, followed by an $18\text{‐}\min$ reperfusion phase. In this figure, the intensity ratio of one out of every three images is plotted over ischemia time. The intensity ratios rose slightly from 1.0 to 1.04 during approximately the first minute of ischemia, and then declined steadily to 0.78 at the time of unclamping. Intensity ratios for the other three $18\text{‐}\min$ injured kidneys at the time of unclamping were 0.76, 0.78, and 0.78. During the reperfusion phase, the intensity ratios rose with a slightly lesser slope than that observed during their decline with ischemia, finally reaching 0.98 after $18\min$ (Fig. 1). Intensity ratios for the other three $18\text{‐}\min$ injured kidneys at the end of the $18\text{‐}\min$ reperfusion phase were 0.97, 0.93, and 0.98.

Fig. 1

Injured-to-normal autofluorescence intensity ratios for kidneys subjected to $18\min$ of warm ischemia followed by $18\min$ of reperfusion. Images were obtained by using $335\text{‐}\mathrm{nm}$ laser excitation and a $395\text{‐}\mathrm{nm}$ long-pass filter.

3.2.

Group 2: Imaging During $\mathit{85}\mathit{Min}$ of Ischemia and $\mathit{35}\mathit{Min}$ of Reperfusion ( $\mathit{335}\text{‐}\mathit{nm}$ Excitation)

Figure 2 illustrates a representative example of the injured-to-normal autofluorescence image intensity ratio over $85\min$ of warm ischemia, followed by a $35\text{‐}\min$ reperfusion phase. In this figure, the intensity ratio of one out of every nine images is plotted. Intensity ratios rose during the first minute of ischemia, and then declined steadily. Between 50 and $60\min$ of ischemia time, the intensity ratios leveled off at 0.68, with no further decline observed for the remaining $25\text{to}35\min$ of ischemia. Intensity ratios for the other three $85\text{‐}\min$ injured kidneys at the time of unclamping were 0.62, 0.64, and 0.65. During the reperfusion phase, the intensity ratios did not change notably for any of the kidneys in group 2, remaining between 0.62 and 0.70.

Fig. 2

Injured-to-normal autofluorescence intensity ratios for kidneys subjected to $85\min$ of warm ischemia followed by $35\min$ of reperfusion. Images were obtained by using $335\text{‐}\mathrm{nm}$ laser excitation and a $395\text{‐}\mathrm{nm}$ long-pass filter.

3.3.

Group 3: Imaging During $\mathit{85}\mathit{Min}$ of Ischemia and $\mathit{35}\mathit{Min}$ of Reperfusion Using 335- and $\mathit{260}\text{‐}\mathit{nm}$ Excitation

Figure 3 shows representative autofluorescence images of kidneys in group 3 that were subjected to ischemic injury for 18 [Figs. 3(a) and 3(b)] and $85\min$ (Figs. 3(c) and 3(d)), with their respective controls. Each pair of kidneys was imaged using both 335- excitation [Figs. 3(a) and 3(c)] and $260\text{‐}\mathrm{nm}$ excitation [Figs. 3(b) and 3(d)] through a $395\text{‐}\mathrm{nm}$ long pass filter. Under $335\text{‐}\mathrm{nm}$ excitation, we observed notable differences in autofluorescence intensity between the injured and control kidneys at $18\min$ of ischemia, and these differences were more pronounced at $85\min$ of ischemia. However, under $260\text{‐}\mathrm{nm}$ excitation, we observed no notable differences at 18 and $85\min$ of ischemia. Furthermore, no notable changes were observed during the reperfusion phase following $85\min$ of warm ischemia with 335- or $260\text{‐}\mathrm{nm}$ excitation (data not shown).

Fig. 3

Autofluorescence in vivo images from the same pair of injured and control kidneys at $18\min$ (a and b) and $85\min$ (c and d) of warm ischemic injury. The top images (a and c) were obtained by using $335\text{‐}\mathrm{nm}$ laser excitation. The bottom images (b and d) were obtained using $260\text{‐}\mathrm{nm}$ laser excitation. All images were obtained using a $395\text{‐}\mathrm{nm}$ long-pass filter.

3.4.

Group 4: Imaging of the Kidneys During $\mathit{20}\mathit{Min}$ of Ischemia and $\mathit{20}\mathit{Min}$ of Reperfusion Under $\mathit{260}\text{‐}\mathit{nm}$ Excitation

The observation of almost no change in the autofluorescence intensity of kidneys undergoing ischemic injury and reperfusion under $260\text{‐}\mathrm{nm}$ excitation was further studied in the group 4 experiments. Autofluorescence images of the kidneys were recorded in real time under $260\text{‐}\mathrm{nm}$ excitation for $20\min$ of injury followed by $20\min$ of reperfusion. Figure 4 illustrates an example of the injured-to-normal autofluorescence image intensity ratio obtained during this process. Although there was a small decline during the injury phase followed by a recovery of the intensity during the reperfusion phase, the value of the ratio did not decline below 0.95.

Fig. 4

Injured-to-normal autofluorescence intensity ratios for kidneys subjected to $20\min$ of warm ischemia followed by $20\min$ of reperfusion. Images were obtained using $260\text{‐}\mathrm{nm}$ laser excitation and a $395\text{‐}\mathrm{nm}$ long-pass filter.

3.5.

Group 5: Imaging with Cross-Polarized Light Scattering at $\mathit{650}\text{‐}\mathit{nm}$ During Ischemia and Reperfusion

Figure 5 demonstrates an example of our experimental results using cross-polarized light scattering with $650\text{‐}\mathrm{nm}$ imaging during $20\min$ of ischemia and $20\min$ of reperfusion. The ratio of the intensity of the injured over that of the normal kidney declined to $\approx 0.75$ within the first $5\min$ of injury. The decline continued during the following $15\min$ of injury at a much slower pace to reach a minimum value of 0.67. Following reperfusion, the intensity ratio increased very rapidly and its value exceeded 1.0 to reach a maximum value of 1.1. This behavior was in sharp contrast to the measured change of the autofluorescence intensity ratio under $335\text{‐}\mathrm{nm}$ excitation (see Fig. 1). A similar behavior was observed in the intensity ratios for the kidneys of the other three rats of group 5—a fast decline of the ratio during the first $5\min$ of injury followed by a slower decline during the remaining $15\min$ of injury. The minimum values of the ratios for these three cases were 0.74, 0.76, and 0.83. Following unclamping, as fresh blood circulated through the kidneys, the recovery of the intensity ratios toward their initial values (1.0) was rapid and almost identical to the case shown in Fig. 5. In one case (the kidney that had minimum value of 0.74), the reperfusion phase led to an overshoot above 1 to a maximum value of 1.3. In the remaining two cases, the ratio reached a maximum value of 1.03 and 1.05.

Fig. 5

Injured-to-normal intensity ratios for kidneys subjected to $20\min$ of warm ischemia followed by $30\min$ of reperfusion. Images were obtained using crosspolarized light scattering imaging at $650\mathrm{nm}$ .

3.6.

Group 6: Emission Spectral Profiles at 335- and $\mathit{260}\text{‐}\mathit{nm}$ Excitation

Figure 4 shows emission spectra, normalized to peak autofluorescence intensity, of injured kidneys in group 6 excited with 335- and $260\text{‐}\mathrm{nm}$ laser light. The spectra represent an average from all animals using spectra from five different sites in each kidney. At $0\min$ of ischemia, kidneys excited with $260\mathrm{nm}$ demonstrated a peak in autofluorescence at around $330\mathrm{nm}$ , whereas those excited with $335\mathrm{nm}$ had peak autofluorescence at around $470\mathrm{nm}$ . Although image intensity changed with increasing ischemia time for kidneys excited with $335\mathrm{nm}$ (see results of groups 1, 2, and 3), the spectral profiles did not change with increasing ischemia time for either excitation wavelength.