Consistent with previous reports 411, mean arterial pressure (MAP) was significantly lower in animals after ischemia (I) and reperfusion (R) (3.0 h I/R and 6.0 h I/R) compared to sham animals, but remained normotensive (> 80 mmHg) throughout the study. MAP did not differ between the I/R groups. Central venous pressure was not different (data not shown).

The measurement of arterial blood gases carried out after the microscopy procedure showed normal oxygenation, a moderate acidosis, and adequate pCO₂ for all groups (Table 1).

Hepatic sinusoidal HbsO₂ of the different groups are shown in Figure 1. Animals treated with 3.0 h I/R have significant lower hepatic HbsO₂ values (56.2 (13.1)) when compared with sham (68.4 (14.1); p < 0.01). No statistically significant differences were observed between 3.0 h I/R and 6.0 h I/R treated animals. However, an obvious shift of hepatic HbsO₂ towards a lower oxygenation was observed when compared with 3.0 h I/R treated animals. Animals treated with 6.0 h I/R and a continuous infusion of endothelin-1 (ET-1) showed significant reduced HbsO₂ values (44.8 (14.7)) when compared with 3.0 h I/R treated animals (56.2 (13.2); p < 0.006). More than half of the measured data from these animals revealed HbsO₂ values lower than 50%. There was no apparent difference in the local tissue hemoglobin (Hb) content detected (data not shown).

Animals subjected to 3.0 h I/R revealed significantly higher NAD(P)H autofluorescence (141.6 (12.8)); therefore, a significant decline in hepatic tissue oxygenation was observed when compared with sham (100.0 (6.7)) (Figure 2). Three hours I/R treated animals failed to show a significant difference in NAD(P)H autofluorescence when compared with the 6.0 h I/R treated animals. Animals treated with 6.0 h I/R and a continuous infusion of ET-1 demonstrated significantly higher NAD(P)H autofluorescence (161.1 (13.8)) when compared to the 3.0 h I/R treated animals (141.6 (12.8)). There was a highly significant correlation found between NAD(P)H autofluorescence and hepatic HbsO₂ detected in the same animal (p < 0.005; r² = 0.94), as depicted in Figure 3.

Serum alanine aminotransferase (ALT) and serum aspartate aminotransferase (AST) levels are summarized in Table 2. When compared with sham animals, mice treated with 3.0 h I/R exhibited significantly higher levels of ALT and AST. No significant changes between 3.0 h I/R and 6.0 h I/R animals were detectable. When compared with 3.0 h I/R, mice treated with 6.0 h I/R and a continuous infusion of ET-1 showed significant higher ALT and AST levels. The results of labelling lethally injured hepatocytes with propidium iodide (PI) are shown in Figure 4. The 3.0 h I/R treated animals exhibited a significantly increase in lethally injured hepatocytes (120.4 (44.0)) compared with sham (25.7 (17.9)), whereas the 6.0 h I/R group had a significant higher number of dead hepatocytes (260.1 (52.7)) than the 3.0 h I/R treated animals. The treatment of 6.0 h I/R animals with a continuous ET-1 infusion further elevated the degree of lethally injured hepatocytes (361.8 (56.0)) when compared to the 6.0 h I/R treated animals. Regression analysis between lethally injured hepatocytes and hepatic HbsO₂ revealed a significant correlation (p < 0.001; r² = 0.86), as shown in Figure 5.

Male C57/BL6 mice (eight to ten weeks old, weighing 23.7 (11.1) g) were used for all experiments. The experimental protocols were in compliance with the guidelines of the Committee on the Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council as well as those of Germany. Animals were maintained under controlled conditions (22°C, 55% humidity and 12-hour day/night cycle) with free access to tap water and a standard laboratory chow.

Mice (n = 7, for each group) were randomly assigned to either a Sham or a hindlimb ischemia/reperfusion (I/R) group. Animals of the I/R groups were treated with 60 minutes bilateral hindlimb ischemia induced by tightening a tourniquet above the greater trochanter of each leg while under anaesthesia. Sham animals were not subjected to ischemia, but remained anaesthetized for the same period of time. Tourniquets were removed just prior to recovery from anaesthesia. The animals were awake during the 3 hours (3.0 h I/R) or the 6 hours (6.0 h I/R) reperfusion periods, and re-anaesthetized for the intravital microscopy procedure.

To further induce liver microcirculatory disturbances and contribute towards a reduction in liver oxygen supply 6.0 h I/R, mice were further randomized to a group treated with a continuous infusion of ET-1 (70 pmol/min., i.v.) starting 15 minutes prior to microscopy. This dose of ET-1 was chosen because it produced alteration in the oxygen distribution, along with derangements in the hepatic tissue perfusion 19.

Animals received anaesthesia, by inhalation, for all procedures. As previously described 20, anaesthesia was performed using isoflurane (Forene, Abbott, Wiesbaden, Germany) in spontaneously breathing animals. The left carotid artery and the left jugular vein were cannulated under sterile conditions. The carotid artery cannula was used for the continuous measurement of systemic arterial blood pressure and heart rate, while central venous pressure was assessed via the jugular vein cannula. Throughout the experiment, normal saline was administered at a rate of 0.4 ml/hr to maintain normal mean arterial pressure. As formerly described 4, and for the realization of the intravital microscopy procedure in anaesthetized animals, a transverse subcostal incision was performed. Briefly, the ligament attachments from the liver to the diaphragm and to the abdominal wall were carefully released. For the evaluation of the hepatic microcirculation by intravital fluorescence microscopy, the animals were positioned on left lateral decubitus and the left liver lobe was exteriorized onto an adjustable stage. The liver surface was covered with a thin transparent film to avoid tissue drying and exposure to ambient oxygen. For equilibrium purposes, a pause of 10 minutes was allowed before data from microscopy and remission spectroscopy was collected. After microscopy, animals were killed by exsanguination, via the insertion of a cannula in the left femoral artery for the collection of arterial blood samples or via cardiac puncture.

Details of this technique have been described elsewhere 421. For observations of the liver microcirculation, we used a modified inverted Zeiss microscope (Axiovert 200, Carl Zeiss, Göttingen, Germany) equipped with different lenses (Achroplan × 10 NA 0.25 / × 20 NA 0.4 / × 40 NA 0.6). The image was captured using a 2/3" charge-coupled device video camera (CV-M 300, Jai Corp., Kanagawa, Japan) and digitally recorded (JVC HM-DR10000EU D-VHS recorder) for off-line analysis. As previously described 22, NAD(P)H autofluorescence, as a marker of the mitochondrial redox state, was assessed using the 10x objective lens. The liver was examined using a filter set consisting of a 365 nm excitation and a 397 nm emission bandpass filter. NAD(P)H autofluorescence was recorded over the complete left liver lobe, allowing at least 15 different fields of view. Non-viable hepatocyte nuclei were labelled in vivo with an i.v. bolus of the vital dye PI (0.05 mg/100 g). As previously stated 21, PI-labelled nuclei were used to identify lethally injured hepatocytes. The fluorescent labelling of these nuclei was viewed using the 20x objective lens and a filter set with a 510 to 560 nm excitation and an emission barrier filter greater than 590 nm. Quantification of redox state and cell death was performed off-line by frame-by-frame analysis of the videotaped images using Meta Imaging Series Software (Ver. 6.1; Universal Imaging Corp., Downington, PA, USA). NAD(P)H fluorescence was densitometrically assessed and expressed as "average intensity/liver acinus". Gain, black level and enhancement settings were identical in all experiments. PI-labelled hepatocytes were expressed as number of cells/10^-1 mm³.

Hepatic sinusoidal HbsO₂ was measured using the remission spectroscopy system Oxygen-to-See (O2C-ATS) supplied with the micro probe VM-3 (Lea Medizintechnik GmbH, Gießen, Germany). White light was continuously emitted via one channel of the micro probe light-guide and was continuously detected via another channel (channel diameter 70 μm). The backscattered light was analyzed in steps of 1 nm (500–650 nm). Each HbsO₂ value was defined by specific Hb spectra. The local tissue light absorbance depends on the total local tissue content of Hb. The local content of Hb was calculated from the local light absorbance and emission. The flexible VM-3 micro probe allowed the detection of oxygen saturation of the left liver lobe placed on the glass slide of the inverted microscope. A special clamping system fixed the micro probe close to the surface of the glass slide and permitted contact-free systematic scanning of the liver lobe (Figure 6). At least 35 different observation points per animal were randomly chosen and examined. Before each experiment, the white standard of the micro probe was calibrated according to the technical instructions of the manufacturer.

Blood was collected immediately after the microscopy procedure, via cardiac puncture. Blood samples were centrifuged at 6500 g, for 5 min, and the remaining serum analyzed, at 37°C, by means of standard enzymatic techniques.

Blood samples for blood gas analyses were collected in heparinized syringes, via the insertion of a cannula in the left femoral artery, at the end of the microscopy procedure. The samples were immediately analyzed using the automated blood gas analyzing system Radiometer ABL 700 (Radiometer Medical Aps., Bronshoj, Denmark).

Data in text and Tables is given as: Mean (SD). Statistical differences between groups and from baseline within each group were determined by ANOVA, followed by the Tukey post-hoc test. The Kolmogorov-Smirnov test was previously used to confirm the normal distribution of data. For checking the nature and extend of the relationship between two variables linear regression analysis was performed. All figures were generated with Sigma Plot (Ver. 8.0) and statistical analyses were performed using Sigma Stat software (Ver. 2.0; SPSS Inc.; München, Germany). Differences were considered significant for p < 0.05.