OSA's Digital Library

Optics Express

Optics Express

  • Editor: C. Martijn de Sterke
  • Vol. 16, Iss. 7 — Mar. 31, 2008
  • pp: 4930–4944
« Show journal navigation

Quantification of in vivo autofluorescence dynamics during renal ischemia and reperfusion under 355 nm excitation

Rajesh N. Raman, Christopher D. Pivetti, Dennis L. Matthews, Christoph Troppmann, and Stavros G. Demos  »View Author Affiliations


Optics Express, Vol. 16, Issue 7, pp. 4930-4944 (2008)
http://dx.doi.org/10.1364/OE.16.004930


View Full Text Article

Acrobat PDF (248 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

We explore a method to quantitatively assess the ability of in vivo autofluorescence as a means to quantify the progression of longer periods of renal warm ischemia and reperfusion in a rat model. The method employs in vivo monitoring of tissue autofluorescence arising mainly from NADH as a means to probe the organ’s function and response to reperfusion. Clinically relevant conditions are employed that include exposure of the kidney to ischemia on the order of tens of minutes to hours. The temporal profile during the reperfusion phase of the autofluorescence intensity averaged over an area as large as possible was modeled as the product of two independent exponential functions. Time constants were extracted from fits to the experimental data and their average values were found to increase with injury time.

© 2008 Optical Society of America

1. Introduction

The advantages of optical spectroscopy methods for tissue characterization are well recognized and include the ability to detect biochemical changes with the short acquisition times necessary for real-time in vivo analysis. Specific optical measurements of changes in cell metabolism were first made by Chance [6

6. B. Chance, “Spectrophotometry of intracellular respiratory pigments,” Science 120, 767–775 (1954). [CrossRef] [PubMed]

] by monitoring the autofluorescence of NADH, an intrinsic fluorescing biomolecule whose concentration is sensitive to hypoxic conditions. Since then, studies have primarily examined short-term ischemia (over several minutes) that is detected locally, where fiber optic microfluorometers typically illuminate a spot less than 2 mm in diameter [7–10

7. S. Kobayashi, K. Nishiki, K. Kaede, and E. Ogata, “Optical consequences of blood substitution on tissue oxidation-reduction state microfluorometry,” J. Appl. Phys. 31, 93–96 (1971).

]. In an exploratory study, we previously found that by monitoring prolonged ischemia over the full kidney surface in an imaging arrangement and using the contralateral (normal) kidney as a control (for normalization), the normalized autofluorescence intensity from the injured kidney under 335 nm excitation was sensitive to ischemia and reperfusion [11

11. J. T. Fitzgerald, A. P. Michalopoulou, C. D. Pivetti, R. N. Raman, C. Troppmann, and S. G. Demos, “Real-time assessment of in vivo renal ischemia using laser autofluorescence imaging,” J. Biomed. Opt. 10, 044018 (2005). [CrossRef]

].

2. Methods

2.1 Rat preparation and renal ischemia

All animal procedures were approved by the University of California, Davis, Animal Use and Care Administrative Advisory Committee. Adult male Wistar-Furth rats weighing 300-400 grams were placed and maintained under general anesthesia using an inspired isoflurane concentration between 1.25 and 2%, which was delivered together with oxygen at 1L/min. Isoflurane concentration was adjusted using a calibrated vaporizer system. Depth of anesthesia was controlled by adjusting isoflurane concentration in response to changes in respiratory rate. Following anesthesia induction, the animal’s chest and abdomen were shaved and prepped with alternating rounds of iodine scrub and 70% ethanol. Next, a midline laparotomy was performed, the small bowel exteriorized, and the left renal pedicle exposed. A non-fluorescent black cloth was then placed over the surrounding tissue to prevent non-kidney tissue autofluorescence from interfering with our measurements. In order to keep the abdominal cavity moist, the intestine was wrapped in saline-soaked gauze, and saline solution was dripped onto the kidneys at 5 minute intervals. A bulldog clamp was used to occlude arterial and venous flow resulting in unilateral ischemic injury to the left kidney, henceforth referred to as the injured kidney. The contralateral right kidney served as an internal control (normal kidney) in this experiment. Rats were studied in three different ischemic injury time groups as follows: 20 minutes (n=12 rats), 50 minutes (n=12), and 150 minutes (n=15). At the end of the injury phase, the clamp was released and the injured kidney was allowed to reperfuse for at least 40, 60, or 60 minutes following injury times of 20, 50, or 150 minutes, respectively.

Using the same procedure, two additional rats were used to record the spectral characteristics of the autofluorescence under 355 nm excitation prior to clamping the renal artery and vein, through the injury phase, and during reperfusion. The first rat underwent 150 min of injury and 90 min of reperfusion, while the second rat underwent 20 min of injury and 40 min of reperfusion. Upon completion of the experiment, the rats were euthanized under general anesthesia.

2.2 Experimental arrangement

The autofluorescence images were recorded through a 420-640 nm band-pass filter positioned in front of a liquid nitrogen-cooled CCD camera (Roper Scientific, Trenton, NJ). A first image was acquired prior to clamping to establish the baseline. Then image acquisition continued immediately after the clamp was applied and blood flow was obstructed through the renal artery and vein. Each image was normalized to the beam spatial profile as recorded on a homogeneous fluorescent piece of paper at the start of the experiment. Each image was also normalized to laser power as recorded by a small fluorescent piece of plastic placed alongside the kidneys. The plastic was chosen as a reference material due to its comparable fluorescence intensity to that of the tissue, and because its intensity does not change with prolonged exposure to the laser wavelengths used in this experiment. The average autofluorescence intensity (henceforth referred to as the “signal”) over as large an imaged area as possible (typically ~1.5 cm2) was recorded for each kidney surface.

An additional experiment was performed to test if NADH photobleaching during the time duration of the experiment is not affecting our experimental observation. Specifically, a kidney that was kept refrigerated ex vivo for 48 hours (to ensure it was metabolically inactive) was imaged under the same excitation and image acquisition conditions. In order to prevent the tissue from drying during the experiment, which can change the tissue scattering and absorption properties independent of changes in its metabolic state, the tissue was first immersed in isotonic saline solution at room temperature for 1 hour to reach thermal equilibrium. Thereafter the immersed kidney received a 0.22 mJ/cm2 dose of 355 nm excitation while its autofluorescence image was recorded over 5 second exposures, as described above, for a duration of 2 hours.

Furthermore, the dynamics of the spectral characteristics of the autofluorescence was investigated in the two additional rats under experimental conditions similar to those used in the imaging experiments. Spectra were acquired using a spectroscopy system suitable for in vivo measurements over a wide spectral range. Specifically, the rat kidneys were irradiated with 355 nm in vivo at a dosage of 1.96 mJ/cm2. The autofluorescence was filtered through a 385 nm long-pass filter and focused onto the entrance slit of an imaging spectrometer (Triax 320, Jobin Yvon Horiba, Edison, NJ) containing a grating blazed at 450 nm. The spectrally analyzed autofluorescence was recorded onto a liquid nitrogen-cooled camera (Roper Scientific, Trenton, NJ).

2.3 Statistical methods

One-way analysis of variance (ANOVA) was employed on the time constants derived from modeling the signal as a double exponential function to determine any overall association of the temporal dynamics of the acquired signal with ischemia time. A p-value ≤0.05 was considered significant. A post-hoc Tukey test was then used to check for pairwise differences between adjacent and non-adjacent injury times.

3. Results

3.1 Experimental results

There are two general optical behaviors, irrespective of injury time, that we observed in the temporal dependence of the autofluorescence intensity during injury. The first behavior is exemplified in Fig. 1(A) showing the behavior of a rat kidney that underwent 150 minutes of ischemia followed by 60 minutes of reperfusion (asterisks). This behavior was observed in 18 of the 39 rats used in this study. A typical example of the second behavior that was observed in the remaining 21 rats is depicted in Fig. 1(B). The characteristic difference between these two optical behaviors is that during the injury phase, the signal in the first behavior declines during injury and falls below the baseline (pre-injury) value while in the second behavior the signal stays above baseline. Upon release of the clamp, the signal is rapidly decreasing in behavior 2 until point (c) in Fig.1. Table 1 lists by injury time group and behavior type the magnitude of this change in intensity ΔI along with the number of rats in parentheses from each group. ΔI is negligible for type 1 kidneys but in type 2 was significantly larger for the case of each injury time (p<0.005, two-sample t-test), supporting the observation of two distinct behaviors. White light photos taken at the end of 150 mins. of injury of the corresponding kidneys in Figs. 1(A) and 1(B) are shown in (C) and (D). Kidneys exhibiting behavior type 1 appeared dark-red in color, while those exhibiting type 2 appeared gray-brown.

Fig. 1. Typical signal profiles from individual rats. 150 mins. injury (asterisks) and 20 mins. injury (solid circles) exhibiting behavior 1 (A) and behavior 2 (B). The signal intensity is normalized to its pre-injury value (intensity=1 at t=0). The letters (a-d) represent characteristic features in the time-dependent signal. White light photos of injured kidneys (and corresponding normals) at the end of 150 mins. of injury which exhibited the signal profiles in (A) and (B) are shown in (C) and (D), respectively.

The letters (a–d) in Figs. 1(A) and 1(B) correspond to characteristic features in the signal profile. The signal intensity is normalized to its pre-injury value (intensity is 1 at t=0). Immediately after the clamp was placed, the signal began to rise from baseline, reaching a peak within the first minute (a). This early increase was observed in both general behaviors. Point (b) indicates the subsequent rapid decrease in signal intensity, after which the signal either continued to decline at a slower rate (behavior 1) or began again to increase (behavior 2). Moreover, the second behavior is also accompanied by a quick drop in the signal intensity after unclamping and as a result, the rest of the reperfusion phase in both behaviors is similar.

Table 1. Mean±SD of ΔI parameter separated by behavior type for the different injury time groups.

table-icon
View This Table
| View All Tables

During reperfusion the signal increased in two phases: it initially increased from (c) to (d) to reach a transient maximum, followed by a decrease leading to an inflection and a subsequent slower increase to final asymptotic value. In 9 of the cases exhibiting behavior 2, the intensity drop and subsequent inflection occurred immediately upon unclamping, while in the remaining 12 cases the intensity drop occurred more slowly over the first 5–10 minutes (later for longer injuries) before beginning its two-phase optical signal increase (Fig. 1(B)). The insets in Figs. 1(A) and 1(B) depict (with the data plotted as solid circles) the signal behavior during the reperfusion phase from two different animals following 20 minutes of injury for comparison. In general, a more pronounced inflection was observed in the reperfusion phases following longer injuries. Three of the 15 rats that underwent 150 minutes of injury did not show any signal return towards baseline (and therefore no inflection) during reperfusion.

Figure 2 displays the intensity of the autofluorescence of the ex vivo kidney tissue under identical excitation and image acquisition conditions to those used in the in vivo experiments. The objective of this set of experiments was to test if a signal decrease similar to that observed in the in vivo experiments can be detected. However, no signal decrease was observed over the 2-hour monitoring period.

Fig. 2. The autofluorescence intensity temporal profile of ex vivo kidney over 2 hours.

In vivo autofluorescence spectra from a kidney under 355 nm excitation normalized to peak intensity are shown in Fig. 3(A) from one of the additional rats that underwent 150 minutes of injury followed by 90 minutes of reperfusion. Specifically, this figure shows the spectrum prior to clamping (injury), at the end of the injury phase, and at the end of the reperfusion phase. A time lapse movie of the normalized spectrum through the entire process is shown in Fig 3(B). The emission peak is observed near 480 nm before injury and at about 460 nm following 150 minutes of injury. These spectra resemble the emission spectra of NADH with the addition of spectral features that may be attributed to re-absorption of the emission by blood. To elucidate the origin of the additional features in the recorded spectrum, the ratio of the spectrum at the end of the injury phase to that at intermediate time points was estimated. Figure 3(C) depicts this spectral ratio profile for several time points. For both rats in which spectral measurements were recorded, the spectral ratio profile for the spectrum recorded prior to injury (t=0 minutes) contains peaks matching those of oxygenated hemoglobin (oxy-Hb). As injury progresses, the location of the peak intensity shifts towards shorter wavelengths while the features at 540 and 580 nm evolve toward a single broad peak at 560 nm. A time lapse of this spectral ratio profile is shown in Fig. 3(D).

Fig. 3. (A) Normalized autofluorescence spectra of a rat kidney in vivo before ischemia (solid), following 150’ of ischemia (dash-dot), and following 90’ of reperfusion (dotted). (B) (147 kB) Movie of the normalized spectrum during 150’ injury and 90’ reperfusion. (C) The ratio of the spectrum after 150’ of injury to the spectrum at the given time point into injury. (D) (133 kB) Movie of this ratio during 150’ injury and 90’ reperfusion. [Media 1][Media 2]

3.2 Analysis of experimental results

The experimental results (Fig. 1) indicate that upon clamping, the signal intensity increased over the first minute to point (a). This result is consistent with an increase in NADH concentration upon oxygen deprivation [12

12. B. Chance, J. R. Williamson, D. Jamieson, and B. Schoener, “Properties and kinetics of reduced pyridine nucleotide fluorescence of the isolated and in vivo rat heart” Biochem. Z . 341, 357–377 (1965).

]. After the first minute, a maximum was reached followed by a decrease in the signal (until point (b)) that cannot be attributed to a further increase in the amount of NADH but rather some other process. This trend is more pronounced in the profiles from the 18 rats in which the first behavior was observed (Fig. 1(A)). We previously postulated on the nature of the process responsible for the autofluorescence intensity decrease during injury under 355 nm excitation that the emission yield of NADH was being modified, caused by either the increasing level of acidity in the cells or some other by-product of ischemia. For example a change in pH is known to alter fluorophore emission efficiency [13

13. H. J. Lin, P. Herman, and J. R. Lakowicz, “Fluorescence lifetime-resolved pH imaging of living cells,” Cytometry 52A, 77–89 (2003). [CrossRef]

], and a decrease in pH indeed occurs in the tissue during ischemia [10

10. A. Mayevsky, J. Sonn, M. Luger-Hamer, and R. Nakache, “Real-time assessment of organ vitality during the transplantation procedure,” Transplant. Rev. 17, 96–116 (2003). [CrossRef]

].

Since a two-phase signal increase in the profiles during reperfusion was consistently observed, we explored the possibility that the signal may be modeled by a double exponential function. During reperfusion the first exponential can approximate the decrease in NADH as it gets oxidized, while the second exponential may represent the additional factor(s) that contribute to the dynamics of the measured signal. During the reperfusion phase, the two exponential components take the following forms:

Component1:SN={SN0 tr<t<ΔτSN0ΔSN*(1Exp((tΔτ)τN))t>Δτ}
(1)
Component   2:SE=SE0+ΔSE*(1Exp(tτE))
(2)

SN and SE represent the contribution of each component to the measured signal. The intensity of the first component decreases exponentially (with relaxation time τN) by the amount ΔSN after a delay time Δτ, while the intensity of the second component increases with relaxation time τE by the amount ΔSE. Time t is measured relative to the release of the clamp tr, and SN0 and SE0 are the initial values of components 1 and 2, respectively, at the signal’s first local minimum of the reperfusion phase (Fig. 1, points (c)). The resultant model is formed by the product of these two components and satisfies the following conditions: 1) the product SN0 * SE0 is set equal to the initial measured signal value, 2) Δτ represents the delay time after the clamp is released until the signal reaches its first peak (Fig. 1, points ‘d’), and 3) the steady-state signal at the end of the experiment is measured and set equal to the product (SN0-ΔSN) * (SE0+ΔSE).

The dynamics of the signal is described by the relaxation time parameter for each component (τN and τE) and the delay time parameter Δτ for the first component. An example data fit and signal decomposition during reperfusion preceded by 50 minutes of injury is provided in Fig. 4. Best-fit values for τN, τE, and Δτ are labeled. Fits to the measured signal intensity vs. time were performed with custom code using MATLAB software (The MathWorks, Natick, MA) by adjusting the two relaxation times τN and τE as well as each component’s initial value SN0 and SE0. Best-fit values were obtained when the mean absolute error between resultant model and measured data was minimized.

Fig. 4. An example fit of the signal profile from a single rat following 50 min. injury. Relaxation (τN and τE) and delay (Δτ) time constants were extracted from the two fitting components.
Fig. 5. Representations of ischemic kidneys (from 39 rats) based on delay and relaxation time constants extracted from data fit.

3.3 Statistical analysis

The results of one-way ANOVA on the values of all three time constants were found to be statistically significant (p<0.05) with respect to injury time. F-values for Δτ, τN, and τE were 13.69 (p<0.0001), 5.55 (p<0.01) and 22.98 (p<0.0001), respectively. Tukey’s post-hoc test revealed that for each time constant, differences in the means between the 150-minute injury case and either the 50 minute injury or 20-minute injury case were statistically significant (right side of Table 2, asterisks), exceeding 5% significance level statistic q5%=3.47 for n=12. The difference between the (20 minute, 50 minute) pair for each time constant was not statistically significant.

Table 2:. Statistical analysis results. On the left, average values of extracted time constants with standard deviations (and number of rats in parentheses). On the right, pairwise comparison of injury time points (20, 50, 150 minutes) using Tukey test statistic q5%=3.47. Asterisks indicate significance at the 0.05 level.

table-icon
View This Table
| View All Tables

4. Discussion

Visual observation (of tissue color and general appearance) is currently most often the only method of assessing the degree of ischemic injury during a transplant procedure. However, this method is qualitative and therefore highly subjective. To address this issue, near-infrared spectroscopy has been used to measure tissue oxygen saturation [5

5. S. G. Simonson and C. A. Piantadosi, “Near-infrared spectroscopy: Clinical applications,” Crit. Care Clin. 12, 1019 (1996). [CrossRef] [PubMed]

,14

14. F. F. Jobsis, “Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters,” Science 198, 1264–1267 (1977). [CrossRef] [PubMed]

] as a means of inferring tissue functional status. Arguably, visual observation as well as an assessment under the red and near-infrared spectral regions is dominated by the presence of hemoglobin and its oxygenation state. Such observations depend on properties of blood in the tissue, such as the extent of near surface perfusion and state of oxygenation, and not on cellular activity. Indeed such techniques have been utilized to quantify perfusion in real-time in clinical kidney transplantation [15

15. A. Gorbach, D. Simonton, D.A. Hale, S.J. Swanson, and A.D. Kirk, “Objective, real-time, intraoperative assessment of renal perfusion using infrared imaging,” Am. J. Transplant. 3, 988–993 (2003). [CrossRef] [PubMed]

].

Microdialysis methods have been investigated in order to monitor changes in extracellular metabolic constituents of low molecular weight (e.g. H+, lactate, glucose, and glutamate) in vivo [1

1. P. O. Carlsson, A. Kiuru, A. Nordin, R. Olsson, J. Lin, P. Bergsten, L. Hillered, A. Andersson, and L. Jansson, “Microdialysis measurements demonstrate a shift to nonoxidative glucose metabolism in rat pancreatic islets transplanted beneath the renal capsule,” Surgery 132, 487–494 (2002). [CrossRef] [PubMed]

,2

2. A. Sola, L. Palacios, J. Lopez-Martim, A. Ivorra, N. Noguera, R. Gomez, R. Villa, J. Aguilo, and G. Hotter, “Multiparametric monitoring of ischemia-reperfusion in rat kidney: Effect of preconditioning,” Transplantation 75, 744–749 (2003). [CrossRef] [PubMed]

] during injury and reperfusion. Such measurements have been made in various organs including kidney [16

16. T. Eklund, J. Wahlberg, U. Ungerstedt, and L. Hillered, “Interstitial lactate, inosine, and hypoxanthine in rat kidney during normothermic ischaemia and recirculation,” Acta Physiol. Scand. 143, 279–286 (1991). [CrossRef] [PubMed]

], brain [17

17. J. C. Goodman, A. B. Valadka, S. P. Gopinath, M. Uzura, and C. S. Robertson, “Extracellular lactate and glucose alterations in the brain after head injury measured by microdialysis,” Crit. Care Med. 27, 1965–1973 (1999). [CrossRef] [PubMed]

], liver [18

18. A. Mehrabi, M Golling, C. Busch, B. Hashemi, R. Ahmadi, A. Volkl, M. M. Gebhard, E. Klar, and T. Kraus, “Experimental monitoring of hepatic metabolism by microdialysis glucose, lactate, and glutamate during surgical preparation of the liver hilus,” J. Surg. Res. 105, 128–135 (2002). [CrossRef] [PubMed]

], and intestine [3

3. T. Sommer and J. F. Larsen, “Detection of intestinal ischemia using a microdialysis technique in an animal model,” World J. Surg. 27, 416–420 (2003). [CrossRef] [PubMed]

], where concentrations of these molecules were found to reversibly change in response to ischemia. However, insertion of the probe can induce trauma in the sampled tissue volume, changing local tissue blood flow and metabolic state [3

3. T. Sommer and J. F. Larsen, “Detection of intestinal ischemia using a microdialysis technique in an animal model,” World J. Surg. 27, 416–420 (2003). [CrossRef] [PubMed]

]. In addition, dialysate measurement interpretation itself is not absolute but changes with tissue temperature, capillary exchange dynamics, and the presence of inflammatory response [18

18. A. Mehrabi, M Golling, C. Busch, B. Hashemi, R. Ahmadi, A. Volkl, M. M. Gebhard, E. Klar, and T. Kraus, “Experimental monitoring of hepatic metabolism by microdialysis glucose, lactate, and glutamate during surgical preparation of the liver hilus,” J. Surg. Res. 105, 128–135 (2002). [CrossRef] [PubMed]

].

A key advantage of the method used in this work is that it is noninvasive once the tissue is accessible. Unlike microdialysis and visual observation, NADH autofluorescence (originating primarily from the mitochondria) probes changes within the cells [19

19. B. Chance, P. Cohen, F. Jobsis, and B. Schoener, “Intracellular oxidation-reduction states in vivo,” Science 137, 499–508 (1962). [CrossRef] [PubMed]

]. NADH participates directly in cellular metabolism in the process of oxidative phosphorylation. In the presence of oxygen, the energy associated with the reduction of NADH to its oxidized state NAD+ is stored in ATP to drive various cellular functions. This process is reversible, and in the absence of oxygen (as in the case of ischemia) NADH can no longer be oxidized aerobically and NADH concentration increases in the cells. The reduced form of NADH fluoresces under UV excitation, while its oxidized form does not fluoresce. Point fluorometric measurements have shown a rapid increase in kidney tissue NADH fluorescence on the order of minutes under anoxic and ischemic conditions [7

7. S. Kobayashi, K. Nishiki, K. Kaede, and E. Ogata, “Optical consequences of blood substitution on tissue oxidation-reduction state microfluorometry,” J. Appl. Phys. 31, 93–96 (1971).

,10

10. A. Mayevsky, J. Sonn, M. Luger-Hamer, and R. Nakache, “Real-time assessment of organ vitality during the transplantation procedure,” Transplant. Rev. 17, 96–116 (2003). [CrossRef]

,19

19. B. Chance, P. Cohen, F. Jobsis, and B. Schoener, “Intracellular oxidation-reduction states in vivo,” Science 137, 499–508 (1962). [CrossRef] [PubMed]

].

The utilization of a compact DPSS 355 nm laser source provides effective excitation of NADH because of the molecule’s broad absorption spectrum [20

20. R. R. Alfano and Y. Yang, “Stokes shift emission of human tissue and key biomolecules,” IEEE J. Quantum Electron. 9, 148–153 (2003). [CrossRef]

]. This source also has higher average power and can be coupled into a fiber bundle to achieve a more stable beam spatial profile illuminating the tissue compared to other sources previously used (such as an optical parametric oscillator [11

11. J. T. Fitzgerald, A. P. Michalopoulou, C. D. Pivetti, R. N. Raman, C. Troppmann, and S. G. Demos, “Real-time assessment of in vivo renal ischemia using laser autofluorescence imaging,” J. Biomed. Opt. 10, 044018 (2005). [CrossRef]

,21

21. A. P. Michalopoulou, J. T. Fitzgerald, C. Troppmann, and S. G. Demos, “Spectroscopic imaging for detection of ischemic injury in rat kidneys by use of changes in intrinsic optical properties,” Appl. Opt. 44, 2024–2032 (2005). [CrossRef] [PubMed]

]). As a result, using a DPSS laser as the excitation source reduced the experimental error and allowed to detect secondary features in the signal temporal profiles.

The in vivo autofluorescence spectra shown in Fig. 3(A) indicate gradual changes in the spectral characteristics during injury and reperfusion. In the emission spectrum at late injury times, the absence of either oxy-Hb or deoxy-Hb spectral features (particularly between 540 nm and 580 nm) yields a spectrum nearly identical to that of pure NADH, which we postulate to be the main fluorophore contributing to the recorded signal. This suggests that there is a reduction of blood concentration in the light-tissue interaction volume. The assumption that the blood volume in the interaction region is decreasing during injury can also explain the observed blue-shift in the emission peak during injury as shown in Fig. 3(A). The ratio of the spectrum at the end of injury to that before injury (Figs. 3(C)) reveals the spectral features of the interfering absorber. This resultant spectrum contains peaks nearly identical to the absorption peaks of oxy-Hb. Monitoring this ratio over time allows us to observe the decreasing influence of oxy-Hb in the interaction volume during injury. This volume is presumed to correspond to a layer on the order of 100 µm beneath the kidney capsule [22

22. Valery TuchinTissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis, (SPIE Press, Bellingham, WA, 2000).

]. These observations suggest that a reduction in total blood content in the interaction tissue volume may be the primary mechanism responsible for the observed changes in the spectral profile. This was the case for both rats used for the autofluorescence spectrum measurements.

After the kidney is clamped, the capillaries located in the kidney cortex are no longer perfused, and so the blood may be settling due to gravity. This is a well known effect in the medical and forensics fields. This can explain the reduction of the re-absorption by blood observed in the spectra with the progression of injury time. In addition, this can explain the abrupt decline in the signal with the onset of the reperfusion phase as seen in general behavior 2 (Fig. 1(B), (c)) as blood will again enter into the volume of tissue that we probe with this technique.

Changing blood content may also be playing a role in defining the two optical behavior types. Figures 1(C) and 1(D) contain white light photos showing that kidneys exhibiting different optical signal behaviors during injury also appear different in color visually at the end of injury. The dark-red color of the injured kidney in Fig. 1(C) suggests the presence of increased concentration of (perhaps coagulated) blood in a type 1 optical behavior kidney, while the gray tint of the type 2 optical behavior kidney suggests relatively reduced concentration of blood, such as perhaps from blood settling. However, it does not explain the overall reduction (in the first general behavior) or marginal decrease (in the second observed general behavior) in the total signal intensity during injury (Fig. 1), as blood settling would reduce the absorption of the emitted fluorescence and would contribute to a signal increase. Therefore, additional factors need to be taken into account to explain the observed temporal profile.

An exponential form was assumed for each component to model the asymptotic behavior of the resultant signal observed in the reperfusion phase because it can be described by a single relaxation time parameter. This asymptotic behavior is consistent with that observed by others attributed to the rapid accumulation of NADH during ischemia to reach a maximum value [8

8. A. Mayevsky and B. Chance, “Intracellular oxidation-reduction state measured in situ by a multichannel fiber-optic surface fluorometer,” Science 217, 537–540 (1982). [CrossRef] [PubMed]

,19

19. B. Chance, P. Cohen, F. Jobsis, and B. Schoener, “Intracellular oxidation-reduction states in vivo,” Science 137, 499–508 (1962). [CrossRef] [PubMed]

]. An example data fit during reperfusion preceded by 50 minutes of injury is provided in Fig. 4. In this way the reperfusion dynamics of the injured kidney of each experimental animal is represented by its relaxation time constants. The data shown in Fig. 5 indicate that these relaxation time constants become longer on average with injury time.

The delay time constant Δτ also increases with injury time (Fig. 5(A)). Once oxygenated blood flow is restored to the tissue, a delay time (Δτ) preceding the decrease in NADH concentration is expected because of the spatial separation between the blood compartment (capillary) and the NADH compartment (intracellular mitochondria, a double membrane-bound organelle within the membrane-bound cell). In addition, electron transport dynamics may be disrupted. These factors may increase the time for O2 to reach the mitochrondria and for NADH to become oxidized. The use of the delay time Δτ helps reproduce the frequently observed behavior of a two-phase optical recovery: a signal increase to an initial peak (Fig. 1, points (d)), then a slight decrease before continuing a slower final increase to its asymptotic value.

In all but three cases, the optical signals were observed to return towards baseline values during the reperfusion phase. In these three of the 15 kidneys that underwent 150 minute injury, the signal either stayed constant at their post-injury value or continued to decline. Visual observation at the conclusion of reperfusion revealed that these kidneys were dark in color, indicative of inadequate reperfusion. In all likelihood, in a clinical setting such kidneys would have been considered nonfunctional. Their signal dynamics are characterized by very long relaxation time constants (Fig. 5).

Differences in the time constants between the 20 minute and 50 minute injury cases were found by statistical analysis to be small, while the time constants in the 150 minute injury case were significantly discernible from those of the shorter injury cases. It should be noted that while injury time is not synonymous with degree of injury, the extreme choice of ischemia duration (150 min) is expected to result in more significant injury than either 20 or 50 min of ischemia. There may have been a small difference in the amount of injury incurred in the 20 vs. the 50 minute injury cases; survival studies on various rat renal ischemic injury models have reported a relatively high survival rate (>75%) for injuries less than an hour but <20% for injuries 80 minutes or longer [24

24. P. Jablonski, B. O. Howden, D.A. Rae, C. S. Birrell, V. C. Marshall, and J. Tange, “An experimental model for assessment of renal recovery from warm ischemia,” Transplantation 35, 198–204 (1983). [CrossRef] [PubMed]

,25

25. K. Togashi, “Rapid restoration of injured kidney after immediate removal of opposite normal kidney,” Acta Pathol. Japon. 32, 749–757 (1982).

]. The large standard deviations in the time constants of a given time group are believed to be due in part to animal-to-animal variations in their response to ischemia and in part due to limitations of the model. The model can only reproduce a signal exhibiting a single inflection point based on its double exponential function construction.

An additional limitation of this study is the absence of monitoring of systemic hemodynamic status prior to, and after, anesthesia induction until clamping of the renovascular pedicle. Such measurements were not made in our experiments owing to the complex instrumentation that they would have required in rodents. Overall, however, our anesthesia protocol mimics the clinical scenario where hemodynamic variations are commonly observed. We cannot categorically exclude that hemodynamic monitoring might have demonstrated differences in hemodynamic parameters from animal to animal at the start of the experiment. If any such potential hemodynamic fluctuations were significant enough, they might have offered a possible explanation for the differences observed between behavior type 1 and type 2 kidneys.

The two optical behaviors observed in this study were clearly an unexpected result suggesting that in future studies, provisions to monitor for a correlation of the variations in hemodynamic parameters to the optical signal must be considered. Regardless of the cause leading to the expression of type 1 or type 2 behavior during ischemia, however, our method of analysis, where the more clinically important post-ischemia reperfusion phase is utilized to extract relevant information, is able to discriminate the reperfusion dynamics (as described by the time constants) of kidneys of prolonged injury from those of shorter injury times.

In our experiments we are concerned with several clinically relevant conditions. The first is heterogeneous tissue response to ischemia. In order to account for such variations, we imaged the entire exposed surface of the kidney, a key difference from what has been done in the past with microfluorometer-assisted measurements. Those measurements were typically made with optical fibers in contact with less than a 2 mm-diameter region of tissue and may depend on local tissue structure, such as the presence of vasculature, which can also give rise to motion artifacts during the measurement. We averaged the intensity over as large an area of the kidney as possible (~1.5 cm2). Second, the clinically relevant condition can involve tissue response beyond just the first couple of minutes of ischemia because in a clinical setting, tissue that has undergone a significant duration of ischemia (tens of minutes to hours) is frequently encountered. Furthermore, we focused our attention into investigating tissue response during the reperfusion phase because the injury phase usually takes place in the absence of medical care and access to monitoring instrumentation. It is during reperfusion when the tissue may exhibit its ability to physiologically recover from ischemia and may provide optical signatures permitting the early prediction of tissue recovery. To investigate these conditions, we have spectroscopically monitored injuries lasting up to 150 minutes, followed by reperfusion phases of at least an hour.

In conclusion, the experimental approach used in this work to assess renal tissue injury in an animal model incorporates previously uninvestigated features. Specifically, extended durations (up to hours) of ischemia and reperfusion are monitored, and as large a region of the kidney surface as possible is monitored in order to average over tissue heterogeneity. These are both conditions that are frequently encountered clinically. A potential application of this method would be to determine the characteristic time constants for a tissue recently transplanted or resuscitated and, using a plot like that shown in Fig. 5, assign it a quantitative value of the degree of ischemia that can assist in determining tissue viability, its ability to recover, and subsequent therapeutic intervention. The experimental results indicate that the adopted model can significantly discriminate between kidneys undergoing 150 minutes of ischemia from those that have undergone 20 or 50 minutes of ischemia via derived time constants describing reperfusion signal dynamics. Future experiments are necessary to determine the correlation of the autofluorescence dynamics to renal function by chronically monitoring rat survival that is dependent on the injured kidney alone. Regardless of whether the injured kidneys after 150 minutes of ischemia in the experiments described in this paper were still functional, these kidneys would still be expected to be more severely injured than those experiencing only 20 or 50 minutes of ischemia. The origin of the changes in signal intensity during ischemia and reperfusion has not yet been determined and may require the monitoring of additional parameters such as scattering of the excitation and the emission or intracellular pH.

Acknowledgments

This work was performed in part under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

This research is supported by funding from the Center for Biophotonics, an NSF Science and Technology Center, managed by the University of California, Davis, under Cooperative Agreement No. PHY 0120999.

References

1.

P. O. Carlsson, A. Kiuru, A. Nordin, R. Olsson, J. Lin, P. Bergsten, L. Hillered, A. Andersson, and L. Jansson, “Microdialysis measurements demonstrate a shift to nonoxidative glucose metabolism in rat pancreatic islets transplanted beneath the renal capsule,” Surgery 132, 487–494 (2002). [CrossRef] [PubMed]

2.

A. Sola, L. Palacios, J. Lopez-Martim, A. Ivorra, N. Noguera, R. Gomez, R. Villa, J. Aguilo, and G. Hotter, “Multiparametric monitoring of ischemia-reperfusion in rat kidney: Effect of preconditioning,” Transplantation 75, 744–749 (2003). [CrossRef] [PubMed]

3.

T. Sommer and J. F. Larsen, “Detection of intestinal ischemia using a microdialysis technique in an animal model,” World J. Surg. 27, 416–420 (2003). [CrossRef] [PubMed]

4.

D. G. Silverman, F. A. Cedrone, W. E. Hurford, T. G. Bering, and D. D. Larossa, “Monitoring tissue elimination of fluorescein with the perfusion fluorometer: a new method to assess capillary blood flow,” Surgery 90, 409 (1981). [PubMed]

5.

S. G. Simonson and C. A. Piantadosi, “Near-infrared spectroscopy: Clinical applications,” Crit. Care Clin. 12, 1019 (1996). [CrossRef] [PubMed]

6.

B. Chance, “Spectrophotometry of intracellular respiratory pigments,” Science 120, 767–775 (1954). [CrossRef] [PubMed]

7.

S. Kobayashi, K. Nishiki, K. Kaede, and E. Ogata, “Optical consequences of blood substitution on tissue oxidation-reduction state microfluorometry,” J. Appl. Phys. 31, 93–96 (1971).

8.

A. Mayevsky and B. Chance, “Intracellular oxidation-reduction state measured in situ by a multichannel fiber-optic surface fluorometer,” Science 217, 537–540 (1982). [CrossRef] [PubMed]

9.

G. Renault, E. Raynal, M. Sinet, M. Muffat-Joly, J. Berthier, J. Cornillault, B. Godard, and J. Pocidalo, “In situ double-beam NADH laser fluorimetry: choice of a reference wavelength,” Am. J. Physiol. 246, H491–H499 (1984). [PubMed]

10.

A. Mayevsky, J. Sonn, M. Luger-Hamer, and R. Nakache, “Real-time assessment of organ vitality during the transplantation procedure,” Transplant. Rev. 17, 96–116 (2003). [CrossRef]

11.

J. T. Fitzgerald, A. P. Michalopoulou, C. D. Pivetti, R. N. Raman, C. Troppmann, and S. G. Demos, “Real-time assessment of in vivo renal ischemia using laser autofluorescence imaging,” J. Biomed. Opt. 10, 044018 (2005). [CrossRef]

12.

B. Chance, J. R. Williamson, D. Jamieson, and B. Schoener, “Properties and kinetics of reduced pyridine nucleotide fluorescence of the isolated and in vivo rat heart” Biochem. Z . 341, 357–377 (1965).

13.

H. J. Lin, P. Herman, and J. R. Lakowicz, “Fluorescence lifetime-resolved pH imaging of living cells,” Cytometry 52A, 77–89 (2003). [CrossRef]

14.

F. F. Jobsis, “Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters,” Science 198, 1264–1267 (1977). [CrossRef] [PubMed]

15.

A. Gorbach, D. Simonton, D.A. Hale, S.J. Swanson, and A.D. Kirk, “Objective, real-time, intraoperative assessment of renal perfusion using infrared imaging,” Am. J. Transplant. 3, 988–993 (2003). [CrossRef] [PubMed]

16.

T. Eklund, J. Wahlberg, U. Ungerstedt, and L. Hillered, “Interstitial lactate, inosine, and hypoxanthine in rat kidney during normothermic ischaemia and recirculation,” Acta Physiol. Scand. 143, 279–286 (1991). [CrossRef] [PubMed]

17.

J. C. Goodman, A. B. Valadka, S. P. Gopinath, M. Uzura, and C. S. Robertson, “Extracellular lactate and glucose alterations in the brain after head injury measured by microdialysis,” Crit. Care Med. 27, 1965–1973 (1999). [CrossRef] [PubMed]

18.

A. Mehrabi, M Golling, C. Busch, B. Hashemi, R. Ahmadi, A. Volkl, M. M. Gebhard, E. Klar, and T. Kraus, “Experimental monitoring of hepatic metabolism by microdialysis glucose, lactate, and glutamate during surgical preparation of the liver hilus,” J. Surg. Res. 105, 128–135 (2002). [CrossRef] [PubMed]

19.

B. Chance, P. Cohen, F. Jobsis, and B. Schoener, “Intracellular oxidation-reduction states in vivo,” Science 137, 499–508 (1962). [CrossRef] [PubMed]

20.

R. R. Alfano and Y. Yang, “Stokes shift emission of human tissue and key biomolecules,” IEEE J. Quantum Electron. 9, 148–153 (2003). [CrossRef]

21.

A. P. Michalopoulou, J. T. Fitzgerald, C. Troppmann, and S. G. Demos, “Spectroscopic imaging for detection of ischemic injury in rat kidneys by use of changes in intrinsic optical properties,” Appl. Opt. 44, 2024–2032 (2005). [CrossRef] [PubMed]

22.

Valery TuchinTissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis, (SPIE Press, Bellingham, WA, 2000).

23.

M. R. L. Stratford, C. S. Parkins, S. A. Everett, M. F. Dennis, M. Stubbs, and S. A. Hill, “Analysis of the acidic microenvironment in murine tumors by high-performance ion chromatography,” J. Chromatogr. A 706, 459–462 (1995). [CrossRef] [PubMed]

24.

P. Jablonski, B. O. Howden, D.A. Rae, C. S. Birrell, V. C. Marshall, and J. Tange, “An experimental model for assessment of renal recovery from warm ischemia,” Transplantation 35, 198–204 (1983). [CrossRef] [PubMed]

25.

K. Togashi, “Rapid restoration of injured kidney after immediate removal of opposite normal kidney,” Acta Pathol. Japon. 32, 749–757 (1982).

OCIS Codes
(170.6510) Medical optics and biotechnology : Spectroscopy, tissue diagnostics
(300.2530) Spectroscopy : Fluorescence, laser-induced
(170.2655) Medical optics and biotechnology : Functional monitoring and imaging
(170.6935) Medical optics and biotechnology : Tissue characterization

ToC Category:
Medical Optics and Biotechnology

History
Original Manuscript: December 17, 2007
Revised Manuscript: January 31, 2008
Manuscript Accepted: March 20, 2008
Published: March 26, 2008

Virtual Issues
Vol. 3, Iss. 4 Virtual Journal for Biomedical Optics

Citation
Rajesh N. Raman, Christopher D. Pivetti, Dennis L. Matthews, Christoph Troppmann, and Stavros G. Demos, "Quantification of in vivo autofluorescence dynamics during renal ischemia and reperfusion under 355 nm excitation," Opt. Express 16, 4930-4944 (2008)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-16-7-4930


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. P. O. Carlsson, A. Kiuru, A. Nordin, R. Olsson, J. Lin, P. Bergsten, L. Hillered, A. Andersson, and L. Jansson, "Microdialysis measurements demonstrate a shift to nonoxidative glucose metabolism in rat pancreatic islets transplanted beneath the renal capsule," Surgery 132, 487-494 (2002). [CrossRef] [PubMed]
  2. A. Sola, L. Palacios, J. Lopez-Martim, A. Ivorra, N. Noguera, R. Gomez, R. Villa, J. Aguilo, and G. Hotter, "Multiparametric monitoring of ischemia-reperfusion in rat kidney: Effect of preconditioning," Transplantation 75, 744-749 (2003). [CrossRef] [PubMed]
  3. T. Sommer and J. F. Larsen, "Detection of intestinal ischemia using a microdialysis technique in an animal model," World J. Surg. 27, 416-420 (2003). [CrossRef] [PubMed]
  4. D. G. Silverman, F. A. Cedrone, W. E. Hurford, T. G. Bering, and D. D. Larossa, "Monitoring tissue elimination of fluorescein with the perfusion fluorometer: a new method to assess capillary blood flow," Surgery 90,409 (1981). [PubMed]
  5. S. G. Simonson and C. A. Piantadosi, "Near-infrared spectroscopy: Clinical applications," Crit. Care Clin. 12,1019 (1996). [CrossRef] [PubMed]
  6. B. Chance, "Spectrophotometry of intracellular respiratory pigments," Science 120, 767-775 (1954). [CrossRef] [PubMed]
  7. S. Kobayashi, K. Nishiki, K. Kaede, and E. Ogata, "Optical consequences of blood substitution on tissue oxidation-reduction state microfluorometry," J. Appl. Phys. 31,93-96 (1971).
  8. A. Mayevsky and B. Chance, "Intracellular oxidation-reduction state measured in situ by a multichannel fiber-optic surface fluorometer," Science 217, 537-540 (1982). [CrossRef] [PubMed]
  9. G. Renault, E. Raynal, M. Sinet, M. Muffat-Joly, J. Berthier, J. Cornillault, B. Godard, and J. Pocidalo, "In situ double-beam NADH laser fluorimetry: choice of a reference wavelength," Am. J. Physiol. 246, H491-H499 (1984). [PubMed]
  10. A. Mayevsky, J. Sonn, M. Luger-Hamer, and R. Nakache, "Real-time assessment of organ vitality during the transplantation procedure," Transplant. Rev. 17, 96-116 (2003). [CrossRef]
  11. J. T. Fitzgerald, A. P. Michalopoulou, C. D. Pivetti, R. N. Raman, C. Troppmann, and S. G. Demos, "Real-time assessment of in vivo renal ischemia using laser autofluorescence imaging," J. Biomed. Opt. 10, 044018 (2005). [CrossRef]
  12. B. Chance, J. R. Williamson, D. Jamieson, and B. Schoener, "Properties and kinetics of reduced pyridine nucleotide fluorescence of the isolated and in vivo rat heart" Biochem. Z. 341, 357-377 (1965).
  13. H. J. Lin, P. Herman, and J. R. Lakowicz, "Fluorescence lifetime-resolved pH imaging of living cells," Cytometry 52A, 77-89 (2003). [CrossRef]
  14. F. F. Jobsis, "Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters," Science 198, 1264-1267 (1977). [CrossRef] [PubMed]
  15. A. Gorbach, D. Simonton, D. A. Hale, S. J. Swanson, and A. D. Kirk, "Objective, real-time, intraoperative assessment of renal perfusion using infrared imaging," Am. J. Transplant. 3, 988-993 (2003). [CrossRef] [PubMed]
  16. T. Eklund, J. Wahlberg, U. Ungerstedt, and L. Hillered, "Interstitial lactate, inosine, and hypoxanthine in rat kidney during normothermic ischaemia and recirculation," Acta Physiol. Scand. 143, 279-286 (1991). [CrossRef] [PubMed]
  17. J. C. Goodman, A. B. Valadka, S. P. Gopinath, M. Uzura, and C. S. Robertson, "Extracellular lactate and glucose alterations in the brain after head injury measured by microdialysis," Crit. Care Med. 27, 1965-1973 (1999). [CrossRef] [PubMed]
  18. A. Mehrabi, M Golling, C. Busch, B. Hashemi, R. Ahmadi, A. Volkl, M. M. Gebhard, E. Klar, and T. Kraus, "Experimental monitoring of hepatic metabolism by microdialysis glucose, lactate, and glutamate during surgical preparation of the liver hilus," J. Surg. Res. 105, 128-135 (2002). [CrossRef] [PubMed]
  19. B. Chance, P. Cohen, F. Jobsis, and B. Schoener, "Intracellular oxidation-reduction states in vivo," Science 137, 499-508 (1962). [CrossRef] [PubMed]
  20. R. R. Alfano and Y. Yang, "Stokes shift emission of human tissue and key biomolecules," IEEE J. Quantum Electron. 9, 148-153 (2003). [CrossRef]
  21. A. P. Michalopoulou, J. T. Fitzgerald, C. Troppmann, and S. G. Demos, "Spectroscopic imaging for detection of ischemic injury in rat kidneys by use of changes in intrinsic optical properties," Appl. Opt. 44, 2024-2032 (2005). [CrossRef] [PubMed]
  22. V. Tuchin, Tissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis, (SPIE Press, Bellingham, WA, 2000).
  23. M. R. L. Stratford, C. S. Parkins, S. A. Everett, M. F. Dennis, M. Stubbs, and S. A. Hill, "Analysis of the acidic microenvironment in murine tumors by high-performance ion chromatography," J. Chromatogr. A 706, 459-462 (1995). [CrossRef] [PubMed]
  24. P. Jablonski, B. O. Howden, D. A. Rae, C. S. Birrell, V. C. Marshall, and J. Tange, "An experimental model for assessment of renal recovery from warm ischemia," Transplantation 35, 198-204 (1983). [CrossRef] [PubMed]
  25. K. Togashi, "Rapid restoration of injured kidney after immediate removal of opposite normal kidney," Acta Pathol. Jpn. 32, 749-757 (1982).

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.

Multimedia

Multimedia FilesRecommended Software
» Media 1: AVI (147 KB)     
» Media 2: AVI (133 KB)     

« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited