Hepatic excretory function in sepsis: implications from biophotonic analysis of transcellular xenobiotic transport in a rodent model

Abstract

Introduction: Hepatobiliary elimination of endo- and xenobiotics is affected by different variables including hepatic perfusion, hepatocellular energy state and functional integrity of transporter proteins, all of which are altered during sepsis. A particular impairment of hepatocellular transport at the canalicular pole resulting in an accumulation of potentially hepatotoxic compounds would have major implications for critical care pharmacology and diagnostics.

Methods: Hepatic transcellular transport, that is, uptake and hepatobiliary excretion, was studied in a rodent model of severe polymicrobial sepsis by two different biophotonic techniques to obtain insights into the handling of potentially toxic endo- and xenobiotics in sepsis. Direct and indirect in vivo imaging of the liver was performed by intravital multifluorescence microscopy and non-invasive whole-body near-infrared (NIRF) imaging after administration of two different, primarily hepatobiliary excreted xenobiotics, the organic anionic dyes indocyanine green (ICG) and DY635. Subsequent quantitative data analysis enabled assessment of hepatic uptake and fate of these model substrates under conditions of sepsis.

Results: Fifteen hours after sepsis induction, animals displayed clinical and laboratory signs of multiple organ dysfunction, including moderate liver injury, cholestasis and an impairment of sinusoidal perfusion. With respect to hepatocellular transport of both dyes, excretion into bile was significantly delayed for both dyes and resulted in net accumulation of potentially cytotoxic xenobiotics in the liver parenchyma (for example, specific dye fluorescence in liver at 30 minutes in sham versus sepsis: ICG: 75% versus 89%; DY635 20% versus 40% of maximum fluorescence; P<0.05). Transcutaneous assessment of ICG fluorescence by whole body NIRF imaging revealed a significant increase of ICG fluorescence from the 30th minute on in the bowel region of the abdomen in sham but not in septic animals, confirming a sepsis-associated failure of canalicular excretion.

Conclusions: Hepatocytes accumulate organic anions under conditions of sepsis-associated organ dysfunction. These results have potential implications for monitoring liver function, critical care pharmacology and the understanding of drug-induced liver injury in the critically ill.

Figure 1

Simplified illustration of sepsis-associated impairment of hepatobiliary excretion. Endo- and xenobiotics are taken up from the bloodstream via transport proteins at the basolateral pole. After transcellular transport with or without biotransformation, endo- and xenobiotics are biliary excreted via transport proteins at the canalicular pole. Sufficient hepatic excretory function depends on various factors which all can be impaired under septic conditions. Atorvastatin can serve as a typical example for profoundly impaired pharmacokinetics under septic conditions leading to supratherapeutic plasma levels after a single oral dose [23]. E, endobiotics; X, xenobiotics).

Figure 2

Effects of polymicrobial sepsis on sinusoidal microcirculation. Sinusoidal microcirculation (n = 4 animals/group, 5 regions of interest/rat) was assessed by intravital microscopy 15 hours after induction of sepsis. Bars represent the proportion of non-perfused sinusoids or velocity of sinusoidal prefusion, respectively. Open bars indicate sham-operated controls; filled bars indicate septic animals. *P < 0.05 for sham versus sepsis ('no': P = 0.009; 'brisk': P = 0.02). Septic animals had an overall shift to reduced sinusoidal perfusion.

Figure 3

Effects of polymicrobial sepsis on biliary excretion of two xenobiotics (DY635, ICG). Kinetics of (A) DY635 and (C) ICG with respect to biliary excretion in sham-operated as opposed to septic rats over an observation period of 30 minutes after intravenous administration (14 pmol/g; n = 6 animals/group; 15 hours after induction of sepsis). Cumulative biliary excretion of (B) DY635 and (D) ICG, respectively. *P < 0.05 for sham versus sepsis (DY635: P ≤ 0.001; ICG: P = 0.013). In septic animals biliary recovery of the two dyes was significantly reduced. ICG, indocyanine green.

Figure 4

Direct visualization of sinusoidal perfusion, hepatocellular uptake and biliary excretion of DY635 and ICG in polymicrobial sepsis assessed by intravital fluorescence microscopy. (A) Kinetics of uptake and excretion of equimolar amounts of both dyes (14 pmol/g) by the liver as measured by densitometric assessment of the dyes (n = 4 animals/group, 8 periportal and 8 pericentral regions of interest/rat; 15 hours after induction of sepsis). Individual grey density maxima were set to 100. Depicted data represent precentage values of individual maximum. *P < 0.05 for sham versus sepsis. #P < 0.05 for within-group comparison. (B) Representative images (original magnification x400) of the same field of view of a liver lobule at baseline and at 5, 15 and 30 minutes after administration of DY635 (upper panel) and ICG (lower panel), respectively. Bar indicates 20 μm. For ICG, 775- to 805-nm excitation and 845- to 855-nm emission band pass filters and for DY635 630- to 640-nm excitation and 650- to 690 emission band pass filters were used. Images were pseudocolored to reveal differences in pixel intensity. Pseudocolored white-red (green-blue) regions reflect areas with high (low) grey densities. For the first image, 365- to 395-nm excitation and 445- to 450-nm emission band pass filters were applied for visualization of the liver architecture (pseudocolor blue). Black areas represent the sinusoids whereas blue areas represent hepatocytes. Hepatocellular uptake of the dyes occurs rapidly after intravenous administration. Then dye concentrations decrease over time reflecting hepatobiliary excretion. In comparison to sham animals, hepatobiliary excretion is impaired in septic animals reflected by the slower decrease of fluorescence intensity. While we could not detect a different uptake of DY635 between septic and sham animals, uptake of ICG seems to be slower as peak fluorescence in septic animals was postponed from 5 to 15 minutes. ICG, indocyanine green.

Figure 5

Transcutaneous whole-body near-infrared fluorescence (NIRF) imaging of ICG enables non-invasive assessment of impaired hepatobiliary excretion in polymicrobial sepsis. Fifteen hours post induction of sepsis, ICG fluorescence was assessed transcutaneously by a whole body NIRF imaging device over a time period of 300 minutes (semi-quantitative analysis, n = 5 animals/group) at the indicated timepoints. *P < 0.05 for sham versus sepsis (90.min: P = 0.019; 150.min: P = 0.001; 300.min: P = 0.005). Kinetics of emitted fluorescence intensities of ICG detected in (A) the liver region or (B) the bowel region of the abdomen. While different fluorescence intensities between sham and septic animals could not be demonstrated in the liver region, fluorescence intensities in the bowel region significantly differed between sham and septic animals. This hints towards a substantial impairment of biliary ICG excretion in septic animals. ICG, indocyanine green.

Figure 6

Whole-body NIRF imaging at the opened situs clearly demonstrates an impairment specifically at the canalicular pole under conditions of sepsis. (A) Depicted are representative images taken 300 minutes after the experiment at the opened situs confirming an accumulation of ICG in the liver of septic animals, with an almost absent fluorescence signal in the bowel. The overlay is of fluorescence (false colors: blue (low fluorescence intensity) and red (high fluorescence intensity)) and white-light images. (B) Depicts the detected fluorescence intensities in liver and bowel of all animals at the open situs 300 minutes after administration of ICG. *P < 0.05 for sham versus sepsis ('liver': P = 0.008; 'bowel': P = 0.002).