Abstract
Fluorescence lifetime is an intrinsic parameter of the fluorescent probe, independent of the probe concentration but sensitive to changes in the surrounding microenvironment. Therefore, fluorescence lifetime imaging could potentially be applied to in vivo diagnostic assessment of changes in the tissue microenvironment caused by disease, such as ischemia. The aim of this study was to evaluate the utility of noninvasive fluorescence lifetime imaging in distinguishing between normal and ischemic kidney tissue in vivo. Mice were subjected to 60-minute unilateral kidney ischemia followed by 6-hour reperfusion. Animals were then injected with the near-infrared fluorescence probe Cy5.5 or saline and imaged using a time-domain small-animal optical imaging system. Both fluorescence intensity and lifetime were acquired. The fluorescence intensity of Cy5.5 was clearly reduced in the ischemic compared with the contralateral kidney, and the fluorescence lifetime of Cy5.5 was not detected in the ischemic kidney, suggesting reduced kidney clearance. Interestingly, the two-component lifetime analysis of endogenous fluorescence at 700 nm distinguished renal ischemia in vivo without the need for Cy5.5 injection for contrast enhancement. The average fluorescence lifetime of endogenous tissue fluorophores was a sensitive indicator of kidney ischemia ex vivo. The study suggests that fluorescence lifetime analysis of endogenous tissue fluorophores could be used to discriminate ischemic or necrotic tissues by noninvasive in vivo or ex vivo organ imaging.
Although still in the early stage of development for clinical applications, fluorescent techniques based on the detection of either endogenous or exogenous fluorophores could provide informative, noninvasive means for medical diagnosis. The detection of warm ischemia in hypothermically preserved kidneys by optical spectroscopy technique may have potential clinical value in predicting kidney viability before, during, and immediately after transplantation. 3 Moreover, optical spectroscopy imaging using 335 nm laser excitation (autofluorescence) has been used to quantify the degree of warm ischemic injury in vivo. 4 The key limitations of optical imaging techniques have been poor tissue penetration and light scattering. These limitations could be alleviated by use of near-infrared light, which allows deeper tissue penetration.5–7 More recently, the time-domain characteristics of the fluorescence, rather than its intensity, have been exploited for microscopy and imaging applications.8–10 The fluorescence lifetime of a probe, the time that a probe spends in the excited state before returning to the ground state, is an intrinsic parameter of the probe, insensitive to the probe concentration but responsive to the local tissue microenvironment. Microscopic techniques measuring fluorescence lifetime, collectively referred to as fluorescence lifetime imaging microscopy (FLIM), can separate the probe excited-state lifetimes into different decay components. Thus, where conventional microscopy detects fluorophores with similar fluorescence intensity distribution throughout a cell, FLIM may detect regional differences in the fluorescence lifetimes, which would indicate different local microenvironments. 10 Fluorescence lifetime characteristics could thus provide useful information about changes in the immediate vicinity of a fluorescent probe. 11 So far, fluorescence lifetime imaging has been applied predominantly in artificial biologic systems or cell cultures. Only recently, in vivo time-domain imaging systems have been developed that can provide information on both the fluorescence intensity and lifetime of the investigated fluorescent probe.
The aim of this study was to evaluate the ability of time-domain near-infrared fluorescence lifetime imaging in vivo to discriminate ischemic tissues in the animal model of acute renal failure. This report provides evidence that fluorescence lifetimes of both exogenous and endogenous fluorophores are sensitive imaging parameters that could be used for noninvasive in vivo and ex vivo localization and monitoring of tissue ischemia and/or necrosis.
Materials and Methods
Mouse Model of Renal Ischemia-Reperfusion Injury
All procedures using animals were approved by the Animal Care Committee following the guidelines established by the Canadian Council on Animal Care. A well-established mouse model of unilateral kidney ischemia was used. 12 Briefly, surgery was performed in CD1 mice (25 g) under isofluorane anesthesia. A midline abdominal incision was made, and heparin (50 IU) was administered intraperitoneally. The left renal vascular pedicle was occluded with a nontraumatic vascular clamp for 60 minutes, during which time the kidney was kept warm and moist using a thermal pad. The clamp was then removed, and the kidney was inspected for immediate color change, indicating successful reperfusion; the incision was then sutured. The body temperature was maintained normothermic during and after surgery using a thermal pad. The structural and functional consequences of brief periods of renal ischemia in this model have been documented. 12 After 6 hours of reperfusion, mice were imaged noninvasively as described below. The uninjured right kidney served as a control for each animal. At the end of the in vivo imaging protocol, mice were sacrificed using formalin perfusion and both kidneys were removed and imaged ex vivo.
In Vivo Time-Domain Optical Imaging
The small-animal time-domain eXplore Optix pre-clinical imager (GE Healthcare, Milwakee, WI) was used in this study. In the eXplore Optix imager using a measured time-resolved fluorescence signal, fluorophore depth and concentration can be calculated. 13 Prior to undergoing renal ischemia-reperfusion-induced injury, mice were imaged to obtain a background image. One group of animals (n = 6) subjected to transient left kidney ischemia-reperfusion as described above was injected with the near-infrared fluorescent probe Cy5.5 (50 nmol; fluorescence lifetime, τ[Cy5.5] = 1.0 ns according to the manufacturer's specifications) via the tail vein using a 0.5 mL insulin syringe with a 27-gauge fixed needle. The other group of animals (n = 6) subjected to transient left kidney ischemia-reperfusion was injected with the same volume of physiologic saline. Fifteen minutes after either Cy5.5 or saline injections, animals were imaged in eXplore Optix. We selected a 15-minute postinjection time for all in vivo imaging experiments because of the rapid clearance of Cy5.5 (molecular weight = 1 kDa) from the circulation. Typically, a majority of the Cy5.5 signal is located in the bladder 1 hour after injection. Fifteen minutes postinjection was found to produce the best signal to background ratio. Three naive animals (not subjected to surgery) and three animals subjected to “sham” surgery were also imaged prior to and 15 minutes after the intravenous injection of 50 nmol Cy5.5 or saline (see the experimental design flow chart, Figure 1A).
In all imaging experiments, a 670 nm pulsed laser diode with a repetition frequency of 80 MHz and a time resolution of 12 ps light pulse was used for excitation. The fluorescence emission at 700 nm was collected and detected through a fast photomultiplier tube and a highly sensitive time-correlated single-photon counting system. Each animal was positioned prone on a plate that was then placed on a heated base (36°C) in the imaging system. A two-dimensional midbody scanning region of interest (ROI) encompassing both kidneys was selected via a top-reviewing real-time digital camera. The optimal elevation of the animal was verified via a side-viewing digital camera. The animal was then automatically moved into the imaging chamber for laser scanning. The laser excitation beam controlled by galvomirrors was then moved over the selected ROI. Laser power and counting time per pixel were optimized at 60 μW and 0.5 seconds, respectively. These values remained constant during the entire experiment. The raster scan interval was 1.5 mm and was held constant during the acquisition of each frame; 1,024 points were scanned for each ROI. The data were recorded as temporal point-spread functions, and the images were reconstructed as fluorescence intensity, fluorescence lifetime, and concentration-depth maps.
In vivo contrast-enhanced fluorescence imaging of mice subjected to unilateral left kidney ischemia-reperfusion. Left kidney transient ischemia (60 minutes, followed by a 6-hour reperfusion) was induced as described in the Material and Methods section. After reperfusion, animals were injected with the near-infrared contrast agent, Cy5.5 (50 nmol), and dorsal images of the kidney regions were acquired after 15 minutes using the eXplore Optix time-domain pre-clinical imager (670 nm excitation and 700 nm emission filters). A, Experimental flow chart of the study design. B, Fluorescence intensity (1 and 3) and fluorescence lifetime (2 and 4) images of mice prior to (1 and 2) and after (3 and 4) left kidney ischemia-reperfusion. C, Topographic representation of the depth superimposed on the three dimensions of the animal profile, after left kidney ischemia-reperfusion. L = left side; R = right side. (D, Comparison of normalized fluorescence intensity-time decay curves of ischemic and nonischemic kidneys.
Data Analyses
The background consisting of a baseline image performed in each animal before the induction of ischemic injury was subtracted from each image. The pre- and postinjection images were manually coregistered to allow for background subtraction. The comparisons among selected parameters (fluorescence intensity and fluorescence lifetime) were made between the ischemic left kidney and the contralateral, uninjured kidney in each animal and between averaged values in experimental animals subjected to surgery and those in naive animals (ie, not subjected to surgery).
To estimate the fluorescence decay, eXplore Optix OptiView software (GE Healthcare) was used.
14
The software deconvolutes the measured fluorescent intensity-time decay curve using the Levenberg-Marquardt algorithm, which applies a nonlinear least-squares minimization algorithm to compute the coefficients of a multiexponential expansion of the fluorescence decay. The quality of curve fitting was judged by the reduced chi-square values and the randomness of the weighted residuals. When a monoexponential model was not adequate to describe the measured decay, a two-exponent fitting was used, and long (τ1) and short (τ2) fluorescence lifetime components, together with their weighted average value τav, were automatically calculated according to the following equations:
Normalized fluorescence intensity was represented in arbitrary units that reflect photon counts normalized with laser power and integration time and was calculated as photon counts ÷ [laser intensity (μW) X integration time (S)]. A three-dimensional reconstruction software from ART Advanced Research Technologies (Montreal, QC) was used for the topographic representations of the depth and visualization of Cy5.5 concentration and location within the animal profile. 15
Statistical analyses of differences between ischemic and contralateral ROIs were performed using t-tests. Significance was accepted for p < .05.
Histologic Assessment of Injury
Control and ischemic kidneys assessed by imaging in vivo and ex vivo were also evaluated for histologic changes. After imaging, animals were perfused with heparinized saline and 10% formalin. Vibrotome kidney sections (50 μm thickness) were stained with hematoxylin and eosin and viewed in phase contrast using an LCM PixCell IIe Microscope (Arcturus, CA). To visualize the Cy5.5 fluorescence signal, kidney sections from Cy5.5-injected animals were stained by brief (10 minute) exposure to fluorescein isiothiocyanate (FITC)-tomato lectin (Sigma, St. Louis, MO; 1:100) to provide tissue morphology background and then viewed in the near-infrared mode (660-680 nm excitation filter and 700 nm long-pass emission filter) using a Zeiss Axiovert 200 fluorescent microscope (Carl Zeiss, Maple Groove, MN). Image processing was performed using AxioVision LE (Carl Zeiss) Rel 4.4 software.
Results
Time-Domain Fluorescence Imaging of Kidneys in Mice Using a Near-Infrared Contrast Probe
Figure 1B shows representative fluorescence intensity (see Figure 1B, 1 and 3) and fluorescence lifetime (τ) (see Figure 1B, 2 and 4) images of a naive mouse (see Figure 1B, 1 and 2) and a mouse exposed to unilateral renal ischemia-reperfusion (see Figure 1B, 3 and 4), both injected with Cy5.5 and imaged in a prone position. The source of fluorescence foci identified as kidneys was confirmed visually by invasive in vivo imaging after surgical incision and skin retraction. The higher fluorescence focus on the left side above the kidney was identified as the stomach. Topographic three-dimensional representation of the ischemic animal (Figure 1C) demonstrates low Cy5.5 signal in the left ischemic kidney in contrast to high Cy5.5 fluorescence signal in the contralateral kidney. The average fluorescence intensity of the ROI covering the ischemic kidney was significantly lower than the fluorescence intensity of a similar-size ROI covering the contralateral kidney (Table 1). More than 80% reduction in fluorescence intensity was observed in the ischemic kidney compared with that in Cy5.5-injected naive or sham-operated nonischemic animals (see Table 1).
The average fluorescence lifetime, τav, of the left kidney did not change after ischemia-reperfusion, whereas τav of the contralateral nonischemic kidney was significantly lower compared with presurgery values (see Table 1). The fluorescence intensity-time decay curves were analyzed to differentiate the Cy5.5 fluorescence signal from that originating from endogenous fluorophores. As shown in Figure 1D (1), the fluorescence intensity time decay in the ischemic kidney after Cy5.5 injection was not different from that in the same kidney prior to the insult, indicating that Cy5.5 signal is not detectable in the ischemic kidney. This is likely due to severely compromised circulation in the ischemic kidney 6 hours after reperfusion. In contrast, Cy5.5 fluorescence decay was clearly detected in the nonischemic kidney (Figure 1D, 2); τav of the nonischemic kidney was similar to that observed in either kidney of sham or naive Cy5.5-injected animals (see Table 1). Similarly, the difference in the fluorescence time decay was detected between the ischemic and the contralateral kidney (Figure 1D, 3; see Table 1).
Both kidneys were also removed and imaged ex vivo. Representative fluorescence intensity images of kidneys dissected from a mouse exposed to unilateral renal ischemia-reperfusion and then injected with Cy5.5 for 15 minutes are shown in Figure 2, A and B. Both optical tomography (see Figure 2A) and topographic three-dimensional representation (see Figure 2B) show low fluorescence intensity in the ischemic kidney compared with the contralateral kidney. The concentration volume slicing shown in Figure 2A demonstrates low Cy5.5 fluorescence throughout the depth of the ischemic kidney and high Cy5.5 fluorescence intensity at a 3 to 8 mm depth in the planar sections of the contralateral kidney. The average Cy5.5 fluorescence intensity ex vivo was significantly lower in the left kidney compared with the contralateral kidney (see Table 1).
Cy5.5 Fluorescence Intensity and Lifetime Values in Kidneys of Control Animals or Animals Prior to or After Being Subjected to Unilateral Left Kidney Ischemia-Reperfusion
τav = average weighted fluorescence lifetime (ns); INT = fluorescence intensity in arbitrary units.
Values are means ± SD after background subtraction, n = 6 per group.
Kidney ischemia-reperfusion was induced as described in the Materials and Methods section. Both control and ischemic animals (at the end of reperfusion) were injected with 50 nmol Cy5.5 for 15 minutes and imaged live in an eXplore Optix time-domain preclinical imager. Kidneys were subsequently removed and imaged ex vivo.
p < .05 (t-test) between control animals and animals subjected to ischemia.
p < .05 (t-test) between the right and the left kidneys.
p < .05 (t-test) between ipsilateral kidneys before and after ischemia.
Interestingly, fluorescence lifetime was higher in the ischemic compared with the contralateral kidney both in vivo (see Figure 1A, 2 and 4) and ex vivo (see Figure 2C, 1 and Table 1). The intensity-time decay curve (see Figure 2C, 2) showed a slower fluorescence decay rate in the ischemic kidney.
Ex vivo contrast-enhanced fluorescence imaging of kidneys dissected from mice subjected to unilateral left kidney ischemia-reperfusion. After reperfusion, animals were injected with the near-infrared contrast agent Cy5.5 (50 nmol) and sacrificed, and their kidneys were imaged ex vivo using an eXplore Optix time-domain preclinical imager (670 nm excitation and 700 nm emission filter). A, Fluorescence intensity and depth-concentration optical tomography in ischemic (L) versus contralateral (R) kidneys. B, The topographic representation of the depth superimposed on the three-dimensional images of the ischemic (L) and contralateral (R) kidney. C, Fluorescence lifetime images (1) and fluorescence intensity-time decay curves (2) of ischemic (L) and contralateral (R) kidneys imaged ex vivo.
To discriminate the fluorescence lifetime contributed by Cy5.5 from that of endogenous fluorochromes, we further analyzed the fluorescence intensity time decay in kidneys ex vivo using a nonlinear least-squares Levenberg-Marquard algorithm. Whereas ischemic kidney fluorescence lifetime could be fitted using a single Levenberg-Marquard algorithm (τ = 1.88 ± 0.07 ns; see Table 1), the florescence intensity time decay in the contralateral kidney was best described by a double Levenberg-Marquard algorithm, identifying two lifetimes: τ1 = 2.92 ± 0.7 ns, relative amplitude A1% = 5 ± 8, indicative of endogenous fluorophores, and τ2 = 1.1 ± 0.1 ns, relative amplitude A2% = 95 ± 10, indicative of the Cy5.5 probe lifetime (τ[Cy5.5] = 1.0 ns).
Time-Domain Fluorescence Imaging of Kidneys in Mice without a Contrast Probe
Figure 3 shows representative time-domain fluorescence intensity (see Figure 3, 1 and 3) and fluorescence lifetime (τ) (see Figure 3, 2 and 4) images of a mouse prior to (see Figure 3, 1 and 2) and after exposure (see Figure 3, 3 and 4) to unilateral renal ischemia-reperfusion, injected with saline and imaged in a prone position. Table 2 shows that the average endogenous fluorescence intensity in the ROI of both ischemic and contralateral kidneys was similar to presurgery values. Additionally, the average weighted fluorescence lifetime of the ischemic kidney was similar to that of the contralateral kidney (see Table 2) and not significantly different from presurgery values (see Table 2). Therefore, neither endogenous fluorescence intensity nor τav was sufficiently different to distinguish the ischemic kidney.
In vivo endogenous fluorescence imaging of mice subjected to unilateral left kidney ischemia-reperfusion. Left kidney transient ischemia (60 minutes, followed by a 6-hour reperfusion) was induced as described in the Materials and Methods section. After reperfusion, animals were injected with physiologic saline and dorsal images of the kidney regions were acquired after 15 minutes using an eXplore Optix time-domain pre-clinical imager (670 nm excitation and 700 nm emission filters). Fluorescence intensity (1 and 3) and fluorescence lifetime (2 and 4) images of mice prior to (1 and 2) and after left kidney ischemia-reperfusion (3 and 4).
Similar analysis was performed in kidneys ex vivo. Representative fluorescence intensity (Figure 4A, 1 and 3) and lifetime (Figure 4A, 2 and 4) images of dissected kidneys from a control mouse (see Figure 4A, 1 and 2) and a mouse subjected to unilateral kidney ischemia-reperfusion (see Figure 4A, 3 and 4) show lower endogenous fluorescence intensity in the left ischemic kidney compared with the contralateral kidney (see Figure 4A, 3). In contrast, fluorescence lifetime in the ischemic kidney was significantly higher than in the contralateral kidney (see Table 2 and Figure 4A, 4). Therefore, both endogenous fluorescence intensity and average fluorescence lifetime identified the ischemic kidney ex vivo. Endogenous fluorescence intensity-time decay curves (Figure 4B, 1–3) demonstrated slower fluorescence decay in the ischemic kidney ex vivo compared with either the left kidney in the control animal (Figure 4B, 1) or the contralateral right kidney in the same animal (Figure 4B, 2); no difference in endogenous fluorescence decay was detected between the right kidneys of control animals and animals subjected to surgery (data not shown).
Further analysis of the endogenous fluorescence lifetime in kidneys in vivo and ex vivo was performed using the Levenberg-Marquard algorithm. In both live animals and ex vivo, the fluorescence lifetime of the ischemic kidney can be fitted using a double Levenberg-Marquard algorithm (Table 3), identifying a long (τ1) and a short (τ2) fluorescence lifetime. The relative amplitudes of both long and short fluorescence lifetimes, A1% and A2%, changed significantly with ischemia in vivo. Following ischemia, A1% increased and A2% decreased in the ischemic kidney, whereas both A1% and A2% remained constant in the contralateral kidney (see Table 3). Whereas the ratios of ischemic A1% to pre-ischemic A1% and ischemic A2% to pre-ischemic A2% were close to 1 for right kidney, they were 0.5 and 2.5, respectively, in the ischemic kidney. Therefore, the two-component lifetime analysis of endogenous fluorescence at 700 nm can clearly distinguish renal ischemia in vivo without the need for Cy5.5 injection for contrast enhancement. Analyses of the autofluorescence lifetime in kidneys ex vivo showed increases in both τ1 and τ2 in the ischemic kidney (see Table 3). However, to minimize the number of imaging parameters capable of identifying a pathologic condition, τav could be used instead since it can adequately identify the ischemic kidney ex vivo.
Histologic Changes
Histologic assessment of hematoxylin-eosin-stained kidney sections consistently revealed profound morphologic changes in kidneys subjected to ischemia-reperfusion (Figure 5, B1 and B2) compared with control kidneys (Figure 5, A1 and A2), including a loss of tubular architecture characterized by tubular dilatation, flattened tubular epithelium, and luminal debris. The tubular epithelial cells displayed condensed, fragmented, and intensely stained nuclei, suggestive of apoptosis and necrosis (see Figure 5, B1 and B2). Similar morphologic changes in this model of kidney ischemia-reperfusion injury have been described by others. 16 Kidney section examination using fluorescence microscopy demonstrated the presence of the Cy5.5 (red) signal in tubules of control kidneys (Figure 5, A3) and its absence in the ischemic kidney (Figure 5, B3), consistent with the in vivo imaging observations.
Fluorescence Intensity Values of Endogenous Fluorophores in Kidneys of Animals Imaged In Vivo and Ex Vivo Prior to and After Unilateral Left Kidney Ischemia-Reperfusion
τav = average weighted fluorescence lifetime (ns); INT = fluorescence intensity in arbitrary units.
Values are means ± SD, n = 6 per group.
Kidney ischemia-reperfusion was induced as described in the Materials and Methods section. At the end of reperfusion, animals were injected with saline for 15 minutes and imaged live in eXplore Optix time-domain preclinical imager. Kidneys were subsequently removed and imaged ex vivo.
p < .05 (t-test) between the left and the right kidneys.
Discussion
The capacity of the two optical imaging parameters fluorescence intensity and fluorescence lifetime to identify ischemic kidney in vivo and ex vivo was evaluated to determine whether they can provide complementary and/or incremental information. Both imaging parameters were evaluated in the presence or absence of an exogenous contrast-enhancing optical probe, Cy5.5, in the experimental model of unilateral kidney ischemia-reperfusion in mice. The analyses suggested that fluorescence lifetime imaging, especially in the absence of a contrast enhancer, is a more sensitive parameter than fluorescence intensity to discriminate between normal and ischemic tissue. The analysis of endogenous fluorescence lifetime may thus be particularly useful in assessing kidney status prior to or during transplantation.
The application of optical imaging techniques for medical diagnosis has been hindered by their poor tissue penetration and light scattering. Although the use of fluorophores with long emission in the near-infrared region achieves deeper tissue penetration and lower background in in vivo applications, it would be advantageous to develop an in vivo imaging technique that measures fluorescence lifetime rather than intensity because the method would be less dependent on transient changes in the local concentration of fluorophores but highly sensitive to changes in the tissue microenvironment. In this study, we used preclinical time-domain optical molecular imaging technology that allows the distinct separation of light scatter and light absorption in tissues and provides fluorescence lifetime information. The time-domain mode of acquisition, based on a time-correlated single-photon counting, enables more accurate recovery of probe concentration. Moreover, fluorescence lifetime-based imaging can discriminate exogenous near-infrared probes from autofluorescence based on their photophysical decay constants, even when their fluorescence intensities are comparable.
Microscopic techniques measuring fluorescence lifetime (FLIM) can separate the probe excited-state lifetimes into different decay components. Because the fluorescence lifetime of a fluorophore is sensitive to environmental and physical processes that influence the excited state, these measurements can potentially provide very detailed information about molecular interactions in living cells, such as changes in protein-protein interactions over time. 10 In practice, however, most of the fluorescent probes that have been characterized in living cells exhibit multiexponential fluorescence decays, and this greatly complicates the interpretation of fluorescence lifetime measurements. Only recently, the feasibility of similar approaches has been tested in in vivo imaging. A recent study demonstrated that imaging the time-domain fluorescence intensity and lifetime of Cyp-RGD injected into A549 tumor-bearing mice could map spatial distribution of the lifetime related to the tumor environment. 8
Ex vivo endogenous fluorescence imaging of kidneys dissected from mice subjected to unilateral left kidney ischemia-reperfusion. After reperfusion, animals were injected with saline and sacrificed, and their kidneys were imaged ex vivo using an eXplore Optix time-domain preclinical imager (670 nm excitation and 700 nm emission filter). A, Endogenous fluorescence intensity (1 and 3) and fluorescence lifetime (2 and 4) images of kidneys ex vivo in control mice (1 and 2) and after left kidney ischemia-reperfusion (3 and 4). B, Comparison of normalized endogenous fluorescence intensity-time decay curves of ischemic and nonischemic kidneys.
Fluorescence Intensity Time Decay of Endogenous Fluorophores in Kidneys of Animals Subjected to Unilateral Left Kidney Ischemia-Reperfusion and Imaged in an eXplore Optix Time-Domain Preclinical Imager Either Live or After Kidney Dissection
A1% = relative amplitude of the first exponential fluorescence lifetime; A2% = relative amplitude of the second exponential fluorescence lifetime; τ1 = long-term fluorescence lifetime; τ2 = short-term fluorescence lifetime.
Data analysis was performed using a double Levenberg-Marquard algorithm to discriminate two fluorescence decay times, τ1 and τ2.
Values are means ± SD, n = 6 per group.
p < .05 (t-test) between ipsilateral kidneys before and after ischemia.
In this study, the fluorescence lifetime characteristics of the exogenous near-infrared probe, Cy5.5, and those of endogenous fluorophores were examined in an ischemic microenvironment in vivo using a unilateral kidney ischemia-reperfusion injury model, which allowed comparative assessments between ischemic and “normal” kidneys in the same animal. The ischemia-reperfusion model used is known to induce profound histologic and functional changes characterized by reduced clearance and subsequent renal failure.16,17 Distribution and fluorescence intensity of an exogenous probe, Cy5.5, were significantly reduced in ischemic kidneys compared with contralateral kidneys both in vivo and ex vivo, suggesting reduced clearance and/or secondary ischemia owing to endothelial injury and swelling. 18 The characteristic fluorescence lifetime of Cy5.5 was not detected in the ischemic kidney either in vivo or ex vivo, likely owing to the limited access and/or excretion of the probe by the ischemic kidney.
Hematoxylin-eosin (A1 and A2; B1 and B2) and fluorescent (A3; B3) staining of kidney sections of control mice (A) and mice subjected to 60-minute unilateral left renal artery occlusion followed by 6 hours of reperfusion (B). Hematoxylin-eosin staining of ischemic kidney sections (B1 and B2) shows evidence of tubular destruction with condensed and fragmented nuclei (arrows). Staining with the FITC-tomato lectin (green) similarly shows a loss of tubular architecture in ischemic (B3) compared with control (A3) kidneys. Cy5.5 signal (red; 660 to 680 nm excitation/700 nm long-pass emission) was detected in control (A3; arrows), but not ischemic (B3) kidney sections. Sections are representative of findings in six different animals.
Whole organs can display complex endogenous fluorescence signatures that can have multiple sources. 19 Therefore, the comparison of endogenous fluorophore intensity and lifetime between ischemic and contralateral kidneys in live animals was less straightforward. The fluorescence intensity and τav were not significantly different between the ischemic and contralateral kidneys.
However, two-exponent fitting of the intensity time decay indicated that the relative amplitudes of both the first and second exponential lifetimes (A1% and A2%) are sensitive indicators of renal ischemia in vivo. The decay in the ischemic kidney consisted of a short lifetime with large amplitude and a long lifetime with relatively small amplitude, in contrast to the pre-ischemic profile characterized by a short lifetime with small amplitude and a long lifetime with large amplitude. Therefore, the two-component lifetime analysis of endogenous fluorescence at 700 nm can distinguish renal ischemia in vivo without the need for Cy5.5 injection for contrast enhancement. In vivo, tissue and environment heterogeneity within and between organs can lead to differences in endogenous fluorophore lifetimes. This modulation can be particularly useful for identifying ischemic tissues and differentiating them from normal tissues. For example, the differences can be caused by changes in extracellular pH, blood flow, tissue oxygen, and nutrient supply 20 or a combination of these factors. 21 Common cellular endogenous fluorophores such as nicotinamide adenine dinucleotide-reduced nicotinamide adenine dinucleotide or cytochrome c emit at shorter wavelengths and are therefore an unlikely source of fluorescence lifetime changes observed in our experiments. It is currently not clear which molecular changes or their combination in ischemic tissues in vivo cause specific shifts and/or multiexponential decay curves of fluorescence lifetimes of endogenous fluorophores.
In contrast to in vivo imaging, both fluorescence intensity and τav of endogenous fluoroprobes were sufficiently sensitive to indicate ischemic kidney ex vivo as sole parameters. This could be explained by the weaker signal and higher scattering of endogenous fluorophores through the skin layer in a live animal, which necessitated further component analysis of the fluorescence intensity-time decay curve to identify ischemia in vivo.
The translation of experimental molecular imaging approaches into diagnostic tools for clinical medicine is of considerable interest.22,23 The findings reported in this study indicate that monitoring changes in the tissue microenvironment through the fluorescence lifetime of endogenous fluorophores could be of particular value in assessing kidney status prior to or during transplantation procedures. Kidney transplantation remains the optimal treatment for patients with end-stage renal disease. However, owing to a general lack of organ donors, kidneys from marginal and non-heart-beating donors are increasingly being used, even though their viability may be compromised. There is currently no rapid yet accurate method for assessing donor organ viability that can be applied within the window of opportunity between harvesting and implantation. Noninvasive magnetic resonance spectroscopy is being increasingly applied to delineate biochemical changes reflecting energy reserve (adenosine triphosphate, nicotinamide adenine dinucleotide phosphate, phosphomonoester, and inorganic phosphate) and pH in organs subjected to cold storage, whereby magnetic resonance spectroscopy surface coils were placed external to cold storage containers. 24 Based on the preliminary studies reported here, we suggest that noninvasive optical imaging of the fluorescence lifetime characteristics of endogenous fluorophores could be developed for assessing the ischemic status of organs ex vivo prior to transplantation. Although requiring a significant further technological development for clinical application, such optical imaging techniques, with or without an optical contrast agent, could also be applied for monitoring organ graft dysfunction in the early post-transplantation period owing to rejection. This could provide an alternative or complementary method to the magnetic resonance imaging techniques currently used for assessing renal function in patients.25,26
In summary, we investigated the ability of time-domain optical imaging to detect experimental acute renal ischemia noninvasively. Analyses of both fluorescence intensity and fluorescence lifetime imaging parameters enabled a reliable demarcation of the ischemic kidney in vivo and ex vivo.
Footnotes
Acknowledgments
We thank Mrs. Zahia Ichalalene from ART Advanced Research Technologies for generation of the fluorescence intensity-time decay curves using Matlab analysis and Mr. Tom Devecseri for his help with graphic and image processing.