Jian X. Zhang

Patterns of vasoregulatory gene expression in the liver response to ischemia/reperfusion and endotoxemia.

Oxidative stress and inflammatory reactions associated with stresses that may lead to shock promote hepatic microcirculatory dysfunction, which may lead to hepatic injury. Because altered liver microcirculation may result from an imbalance in the expression of stress-induced vasoactive mediators, our study was conducted to investigate changes in the expression of genes encoding endothelin-1 (ET-1), its receptors, ET(A) and ET(B), heme-oxygenase 1 (HO-1), and inducible nitric oxide synthase (iNOS), using two different rat models of liver stress: ischemia/reperfusion of the liver and lipopolysaccharide (LPS)-induced endotoxemia. In ischemia/reperfusion experiments, rats were subjected to 1 h hepatic ischemia, followed by 6 h of reperfusion. Endotoxemia was induced by i.p. injection of LPS (1 mg/mL/kg body weight); rats were studied after 6 h. mRNA levels were estimated using semiquantitative reverse transcriptase-polymerase chain reaction (RT-PCR) on total RNA samples prepared from experimental and sham control rat livers. In the ischemic reperfused livers the levels of mRNA for ET-1, ET(B), HO-1, and iNOS were significantly elevated. The fold increase versus sham was 2.5+/-1.1 (ET-1), 2.1+/-1.3 (ET(B)), 2.1+/-.8 (HO-1), and 6.4+/-3.9 (iNOS). In contrast, the expression of ET(A) receptor gene was reduced after ischemia/reperfusion (to 73+/-1% of sham). In the separate experiments we analyzed the same mRNAs levels after 1 h of ischemia (no reperfusion), and did not detect any changes. During endotoxemia we observed a marked increase in iNOS mRNA level (>24-fold), as well as a marked elevation of the other four mRNAs. The fold increase versus sham was 6.1+/-1.7, ET-1); 1.5+/-.3 (ET(A)); 1.6+/-.4 (ET(B)); and 2.4+/-.34 (HO-1). These results show that liver stress, induced by ischemia/reperfusion or LPS injection have characteristic patterns of vasoregulatory genes expression indicating that, although both stresses result in an increase in specific vascular reactivity, different pathways are involved in inducing the hepatic vascular stress response.

Survival transplantation of preserved non???heart-beating donor rat livers: preservation by hypothermic machine perfusion1

Background. Non–heart-beating donor (NHBD) livers are an untapped source with the potential to provide relief to the current donor shortage problem. Hypothermic machine perfusion (MP) has the potential to reclaim and preserve these marginal donor organs. Methods. This study compared 5-day survival in a rat NHBD liver transplantation model with simple cold storage (SCS) and MP-preserved tissues that had experienced 30 min of warm ischemia followed by a 5-hr preservation period with the University of Wisconsin solution. Total release of lactate dehydrogenase (LDH) and alanine aminotransferase (ALT) were determined at major time points. Bilirubin levels and histology were examined after 5-day survival. Results. Six of seven control livers and five of six MP livers survived, whereas SCS tissues had survival in zero of seven. The results showed that MP livers had reduced release of LDH and ALT after 5 hr of storage, 5.07±1.42 and 2.02±0.69 U (mean±SE), respectively, compared with SCS, 15.54±0.81 and 3.41.3±0.73 U, respectively. Bilirubin values after 5-day survival of MP livers (1.17±0.49 mg/dL) were comparable to controls (0.91±0.36 mg/dL). Histology confirms that SCS displayed increased necrosis and MP tissue showed regions of near normal hepatic structure. Conclusions. These results suggest that MP for 5 hr improves survival and reduces cellular damage of liver tissue that has experienced 30 min of warm ischemia when compared with SCS tissues. Further studies need to be conducted, but this study suggests that MP preservation has the potential to reclaim and preserve NHBD liver tissues.

REMODELING OF HEPATIC MICROVASCULAR RESPONSIVENESS AFTER ISCHEMIA/REPERFUSION

Although there is substantial evidence suggesting that the integrity of the microcirculation is an important determinant of tissue viability during reperfusion after ischemia in the liver, as well as other tissues, the mechanisms responsible for microvascular failure are not fully understood. It is now recognized that the microvascular response to reperfusion, similar to the whole organism response to shock, can consist of either a rapid exacerbation of injury after a severe ischemic episode or, alternatively, a more slowly developing alteration in responsiveness that occurs after a less severe insult. In the more slowly developing response, the alterations in vascular status are the result of up-regulation of stress-induced vascular mediators such as endothelin, nitric oxide synthase (NOS), and heme oxygenase, as well as changes in the reactivity of the effector cells to the mediators. The mechanisms for change in reactivity of vascular cells range from changes in receptor expression to overt phenotypic transformation, as can occur in the hepatic stellate cells in response to repeated injury. When maintained in balance, these counteracting constrictor and dilator influences can be protective; however, local imbalance can result in focal ischemia, thus propagating the injury. Thus, the remodeling of the hepatic microvascular responsiveness during reperfusion after ischemia may serve as a useful paradigm for consideration of the overall response of the organism to shock.

Endothelin-1 and heme oxygenase-1 as modulators of sinusoidal tone in the stress-exposed rat liver☆

Heme oxygenase (HO)‐1 is up‐regulated after ischemia/reperfusion and contributes to maintenance of hepatic perfusion and integrity. Blockade of HO‐1 leads to an increased portal pressor response in the stress‐exposed liver. We tested whether the increase in portal pressure reflects unmasking of a concomitant up‐regulation of the vasoconstrictor endothelin (ET)‐1. Hemorrhagic shock induced messenger RNAs encoding HO‐1 (16‐fold) and ET‐1 (9‐fold) with a similar time course in the liver. At maximum induction of both mediators, rats received either vehicle or the endothelin ETA/B antagonist bosentan (10 mg/kg intravenously). Subsequently, the HO pathway was blocked in all animals by tin‐protoporphyrin (SnPP)‐IX (50 μmol/kg intravenously). Portal and sinusoidal hemodynamics were measured using microflow probes and intravital microscopy, respectively. Blockade of the HO pathway led to a significant increase in portal resistance (sham/SnPP‐IX, 0.17 ± 0.046 mm Hg · min · mL−1; shock/vehicle/SnPP‐IX, 0.57 ± 0.148 mm Hg · min · mL−1; P < 0.05) and a decrease in sinusoids conducting flow (shock/vehicle/SnPP‐IX: baseline, 28.3 ± 0.85 sinusoids/mm; 10 minutes after SnPP‐IX, 23.1 ± 1.09 sinusoids/mm; P < 0.05). Intravital microscopy showed narrowing of failing sinusoids colocalizing with stellate cells after blockade of the HO pathway. Blockade of ETA/B receptors attenuated the increase in portal resistance (shock/bosentan/SnPP‐IX, 0.29 ± 0.051 mm Hg · min · mL−1) and prevented sinusoidal perfusion failure (shock/bosentan/SnPP‐IX: baseline, 28.2 ± 0.97 sinusoids/mm; 10 minutes after SnPP‐IX, 28.8 ± 1.18 sinusoids/mm) as well as sinusoidal narrowing. In conclusion, a functional interaction of the up‐regulated vasodilatory HO system and the vasoconstrictor ET‐1 on the sinusoidal level exists under stress conditions. Both mediator systems affect sinusoidal diameter via direct action on hepatic stellate cells in vivo. (HEPATOLOGY2002;36:1453–1465).

Potentiated hepatic microcirculatory response to endothelin-1 during polymicrobial sepsis.

We conducted this study to elucidate the role of endothelins (ET-1) in mediating the hepatic microcirculatory dysfunction observed in response to sepsis. Following 24 h of cecal ligation and puncture (CLP), we performed intravital microscopy both in vivo and on isolated perfused livers. Portal resistance increased in response to ET-1 in both sham and septic rats, with no significant difference between the two in either in vivo or in isolated livers. Sinusoidal volumetric flow (Qs) was evaluated using red blood cell velocity (VRBC) and sinusoidal diameter (Ds) to determine microvascular hemodynamic integrity. Qs decreased in response to ET-1 in livers from CLP rats compared with sham (P < 0.05, CLP vs. sham) in both in vivo and isolated livers. In vivo infusion of ET-1 resulted in greater constriction of sinusoids in the CLP group compared with sham (P < ,0.05), resulting in higher sinusoidal resistance. Microvascular hyper-responsiveness was accompanied by hepatocellular injury in CLP rats, but not in sham rats. RT-PCR was performed to measure mRNA levels of ET-1, its receptors ETA and ETB, inducible and constitutive nitric oxide (NO) synthase (iNOS and eNOS, respectively), and heme oxygenase 1 (HO-1). After CLP, both ET-1 and ETB mRNA increased, whereas ETA mRNA tended to decrease, although the change was not statistically significant. Livers from CLP rats showed no significant change in levels of eNOS mRNA, but showed a significant increase in iNOS expression (13.5-fold over sham). There was no change in the level of HO-1 mRNA between sham and CLP groups. Taken together, these results suggest that sepsis sensitizes the hepatic microcirculation to ET-1. More importantly, an impaired microcirculatory flow due to ET-1 in sepsis contributes to hepatic injury. Further, localized imbalances between endothelins and NO may mediate the altered microvascular response during sepsis.

Hepatic intercellular communication in shock and inflammation.

The liver is well recognized as a target for injury during low flow or inflammatory states. Functionally, the result is both metabolic and host defense dysfunction. Although the liver is clearly responsive to changes in systemic levels of various mediators, it is becoming apparent that substantial changes occur within the liver that are not directly dependent on extrahepatic factors. This is the result of complex interactions among the various cell types that exist in a highly organized arrangement within the functional subunit of the liver. The purpose of this review is to summarize the structural relationships which form the basis for this system of cell-cell communication and their functional implications both in the normal liver and during both low-flow and normal-flow inflammatory states.

Chronic ethanol consumption exacerbates liver injury following hemorrhagic shock: role of sinusoidal perfusion failure.

Although the deleterious effect of chronic ethanol consumption on subsequent stressful events has long been recognized, the pathophysiological mechanisms are incompletely understood. This study tested whether chronic ethanol consumption in doses that increase sinusoidal contractility increases susceptibility to hepatic microvascular failure and liver injury after hemorrhagic shock. Liver microcirculation was assessed by in vivo microscopy during hemorrhage and up to 24 h after onset of resuscitation and was compared with liver histology and serum enzyme levels. Mean sinusoidal blood flow was neither impaired by chronic ethanol feeding at baseline nor during hemorrhage and early resuscitation. However, failure of individual sinusoids to conduct flow was observed more frequently after fluid resuscitation in ethanol-fed animals (e.g. at 1 h after onset of volume therapy: 26% of sinusoids) than in controls (11%), reflecting substantial flow heterogeneity. Failing sinusoids had substantially smaller diameters than sinusoids conducting flow with a more profound and sustained response in ethanol-fed rats. At 24 h marked pericentral necrosis and increase in serum alanine aminotransferase levels were observed in six of nine surviving ethanol-fed animals but only in 1 of 10 pair fed controls and correlated with microvascular failure. These data suggest that early as well as late microvascular failure in this model of hemorrhagic shock and resuscitation is primarily mediated at the level of individual sinusoids. Chronic ethanol feeding exacerbates microvascular and hepatocellular injury after shock/resuscitation, probably involving increased sinusoidal contractile responsiveness.

Gadolinium chloride alters the acinar distribution of phagocytosis and balance between pro- and anti-inflammatory cytokines.

ABSTRACT Gadolinium chloride (GdCI3) is commonly used to deplete the liver of Kupffer cells (KC) and has been shown to decrease hepatic phagocytic activity and to abolish hepatic expression of certain KC-specific antigens. However, the exact fate of the KCs after GdCI3 treatment remains unclear. To determine if GdCI3 actually decreases the total number of KCs in the liver, we labeled phagocytically-active KC by administering fluorescent-labeled latex beads to rats treated with either normal saline or GdCI3. Total hepatic fluorescence and the distribution of fluorescence within liver acini were evaluated by intravital microscopy. Hepatic mRNA levels of KCR, a KC-specific gene product, and Pu-1, a ubiquitous monocyte gene product, were assessed by Northern blot analysis, and differences in the expression of pro-inflammatory (tumor necrosis factor (TNF)-±) and anti-inflammatory (interleukin (IL)-10) cytokines were assessed by reverse-transcriptase polymerase chain reaction (RT-PCR). Our results indicate that GdCI3 does not significantly reduce the number of phagocytically active cells in the liver, but alters the acinar distribution of these cells and may provoke a switch in the KC phenotype such that these cells no longer express KCR or IL-10. GdCI3 pretreatment inhibited stress-related induction of IL-10, but failed to down-regulate expression of TNF-±. This phenotypic change is likely to have important consequences because it permits relative overexpression of TNF-±.

Effect of activation on neutrophil-induced hepatic microvascular injury in isolated rat liver.

Polymorphonuclear neutrophils (PMNs) have been implicated in microvascular injury following ischemia and reperfusion (I/R) but the relative contribution of obstruction versus toxic mediators is not well defined. Therefore, the present study was performed to determine the contribution of exogenous or endogenous activation on PMN-induced microvascular and hepatocyte injury. Rat livers were isolated and perfused at constant pressure with Krebs buffer with red cells (Hct-10%) and monitored for perfused sinusoids (PS) and dead hepatocytes (propidium iodide-stained, DH) by intravital microscopy. PMNs isolated from the peritoneum after oyster glycogen injection were added to the perfusate either without or with activation by phorbol myristate acetate (PMA, 160 nM). Unactivated PMNs stuck in the liver but had no significant effect on either perfused sinusoids (11.1 ± .4/field, unactivated PMNs versus 11.9 ± .5/field, the time-matched control) or dead hepatocytes (1.2 ± .4/field, unactivated PMNs versus 1 ± .3/field, the time-matched control). Infusion of PMA-activated PMNs resulted in significant decrease in perfused sinusoids and increase in DH (9.5 ± .3/field for PS and 3.2 ± .6/field for DH, respectively). In contrast, when PMNs were “activated” by infusion into a liver previously made ischemic for 30 min, DH were significantly increased after 60 min (26.2 ± 4.5/field, I/R plus PMNs versus 12.4 ± 2/field, I/R only) but perfused sinusoids were not different from ischemia alone. These results demonstrate that oxidatively quiescent PMNs do not cause cellular or microvascular injury in spite of microvascular accumulation. Activated PMNs damage microcirculation or hepatocytes depending on the nature of the activation.

Endothelin-1 as a regulator of hepatic microcirculation: sublobular distribution of effects and impact on hepatocellular secretory function.

Endothelin-1 is a potent vasoconstrictor in the portal circulation causing sinusoidal constriction as well as presinusoidal resistance changes. Using in vivo videomicroscopy we studied the sublobular differences in dose response characteristics of sinusoid constriction at 10–13 min into continuous infusion of .5, 1.0, 3.0, and 5.0 pmol of endothelin-1 (ET-1)/100 g body weight/min in pentobarbital anesthetized rats. In addition, bile flow was monitored to estimate parenchymal secretory function. ET-1 evoked a profound constrictor effect in both sublobular regions studied. However, the maximal decrease of sinusoidal width in periportal inflow region (zone 1; 1 pmol/100 g/min: 4.8 ± .1 μ) was reached at slightly lower concentrations of the peptide than in pericentral outflow region (zone 3; 3 pmol/100 g/min: 6.2 ± .3μ) compared to respective baseline values (zone 1: 7.1 ± .1 μ; zone 3: 10.2 ± .1 μ), suggesting upstream binding and clearance of the peptide. The constrictor response in zone 1 was biphasic and at higher concentrations of the peptide (5 pmol/100 g/min: 5.5 ± .2 μ) sinusoidal widths increased again compared to the maximal response with 1 pmol. Secretory function as reflected by the bile flow was maintained or even slightly increased with the lower doses (.5 and 1 pmol) of ET-1 and during the first 10 min into infusion of the higher doses (3 and 5 pmol). Subsequently, an approximately 20–25% decrease in bile flow accompanied the infusion of higher doses of ET-1. These results indicate that the vascular effects occur at lower concentrations than the cholestatic effect and precede the cholestatic effect, suggesting a causal effect of microvascular perturbations on cholestasis that is associated with portal infusion of higher doses of ET-1.