Bernhard Urbaschek

The microcirculation during endotoxemia

The initial responses to endotoxemia are detectable in the microcirculation as a microvascular inflammatory response characterized by activation of the endothelium stimulating these cells from their normal anticoagulant state to a procoagulant state with increased adhesiveness for leukocytes and platelets. Concomitantly, arteriolar tone is lost and reactivity to a variety of agonists is modified. Tissue damage subsequently results not only from reduced perfusion of the exchange vessels, but also from injurious substances released from activated, sequestered leukocytes as well as activated endothelial cells, macrophages, and platelets. This is the result of endotoxins inducing activation and interaction of a number of effector cells, cascades, and acute-phase responses, such as the complement, coagulation, bradykinin/kinin, and hematopoietic systems accompanied by the release of a myriad of mediators. These include eicosanoids, cytokines, chemokines, adhesion molecules, reactive free radicals, platelet-activating factor, and nitric oxide. This paper briefly reviews the microvascular responses to endotoxemia and discusses some of the mechanisms involved.

Role of endotoxin in the hepatic microvascular inflammatory response to ethanol.

Kupffer cells (KC) and gut‐derived bacterial endotoxin have been implicated in the aetiology of alcoholic liver disease. Using in vivo microscopic methods, we have shown that ethanol ingestion in mice causes a dose dependent increase in leucocyte adhesion and endothelial cell swelling in hepatic sinusoids. Activation of KC is elicited at low doses while depression occurs at high doses and with chronic exposure. The responses are exacerbated in the presence of endotoxaemia or sepsis and are not seen in endotoxin‐resistant animals, implicating a role for endotoxin in the ethanol‐induced inflammatory response. In addition, the responses are abolished with anti‐TNFα suggesting that TNFα is a primary mediator of these events. Nitric oxide (NO) initially appears to play an important role in these events by stabilizing the TNFα ‐mediated hepatic microvascular inflammatory response to acute ethanol ingestion, thereby helping to protect the liver from ischaemia and leucocyte induced oxidative injury. Finally, an ongoing clinical study has confirmed a mild systemic endotoxaemia in patients hospitalized for alcoholic liver disease. All of these results support important roles for endotoxin, cytokines, nitric oxide and sinusoidal lining cells in the pathophysiology of liver injury resulting from ethanol alone or in combination with infection.

In vivo microscopic studies of the responses of the liver to endotoxin.

SummaryIn vivo microscopic methods concomitant with electron microscopic and histochemical procedures are being used to explore the sequelae of responses of Kupffer cells and the hepatic microvasculature to endotoxins. To gain further insight into the role of the liver in host defense and nonspecific resistance, the effects of endotoxin also are being studied in animals sensitized to endotoxin (BCG infection) or tolerant to endotoxin (pretreated with detoxified endotoxin, low doses of endotoxin, or in C3H/HeJ mice). The results to date, have demonstrated that endotoxin induces significant alterations in the hepatic microcirculation due to swelling of Kupffer and endothelial cells and the adhesion of leukocytes and platelets to the sinusoid wall. Lymphocytes frequently are associated with the Kupffer cells. Phagocytosis also is affected; following a brief period of stimulation, the rate of phagocytosis by Kupffer cells is depressed. In BCG infected animals all of these responses are exaggerated but can be minimized by pretreatment with detoxified endotoxin or minute concentrations of endotoxin 24 h prior to the challenge dose of endotoxin. The responses are not seen in the endotoxin low-responder, C3H/HeJ mouse which was found to have a deficiency in lysosomal enzymes and a paucity of functional Kupffer cells. The results provide some insight into the sequelae of cellular and microvascular events that occur in the liver during endotoxemia, endotoxin-related host defense mechanisms and non-specific resistance. In addition, support is provided for the central role of Kupffer cells in these events and that lysosomal enzymes participate in the toxic response elicited by endotoxin.

Effect of immunoglobulin G on the hepatic microvascular inflammatory response during sepsis.

The effects of intravenous immunoglobulin G (ivIG) on the hepatic microvascular inflammatory response to sepsis were studied in rats by in vivo microscopy. High doses of ivIG (300 mg/kg bw) (Sandoglobulin or rat IgG) significantly improved the 48 h survival of septic rats from 25-66% when ivIG was given before or immediately after cecal ligation and puncture. Circulating endotoxin also was significantly reduced. Eight hours after inducing sepsis, the average number of leukocytes adhering to the sinusoidal endothelium increased 15-fold and the average decrease in the number of perfused sinusoids was 22%. IvIG administration minimized these responses. In both septic and nonseptic animals, ivIG also reduced the phagocytic activity of Kupffer cells. The results suggest that high doses of ivIG not only reduce lethality but also limit hepatic microcirculatory dysfunction during sepsis by minimizing leukocyte-endothelial interactions that may be a result of reducing circulating endotoxin and modifying Kupffer cell function.

Ethanol exacerbates hepatic microvascular dysfunction, endotoxemia, and lethality in septic mice.

The effect of acute ethanol administration on the hepatic microvascular responses to sepsis was studied. Polymicrobial sepsis was induced 30 min after mice had received ethanol (1 g/kg b.w.) or isocaloric maltose-dextrin by gastric gavage. Lethality within 24 h was 91.7% in the ethanol-treated animals and 40.0% in septic controls. Endotoxin levels in ethanol treated animals were 107 pg/ml at 6 hr and 1205 pg/ml at 12 h, compared with 32 pg/ml and 104 pg/ml, respectively in the controls. In vivo microscopy revealed that at 3 h in the ethanol treated septic animals, Kupffer cell phagocytic activity was increased by 41%, whereas the number of sinusoids containing blood flow were reduced by 34% concomitant with a 144% increase in the adherence of leukocytes to the sinusoidal walls when compared with the septic controls. By 6 h, however, Kupffer cell phagocytic activity was reduced by 48% in the ethanol treated animals; this was accompanied by a further deterioration in sinusoidal blood flow. Thus, a small, acute dose of ethanol causes significant impairment of the hepatic microcirculation followed by suppression of Kupffer cell activity. This results in exacerbation of endotoxemia and lethality during polymicrobial sepsis.

Deficient Kupffer Cell Phagocytosis and Lysosomal Enzymes in the Endotoxin‐Low‐ Responsive C3H/HeJ Mouse

Various substances, including lysosomal enzymes, are produced by Kupffer cells and other macrophages; their release has been implicated in the toxic response to endotoxins. C3H/HeJ mice exhibit little or no response to doses of endotoxin that are lethal in syngeneic C3HeB/FeJ mice. To explore the nature of this deficient response, the Kupffer cells of these mice were studied using in vivo microscopic as well as histochemical and electron microscopical methods. In vivo, the rate of phagocytosis of single 0.8 μ m latex particles was measured in individual Kupffer cells as was the number of phagocytic cells per microscopic field. Frozen sections of livers were stained for a variety of lysosomal enzymes and liver specimens also were processed for electron microscopy. In comparison to the endotoxin‐sensitive C3HeB/FeJ mice, the livers of the C3H/HeJ mice contained 60% fewer Kupffer cells that phagocytosed latex. However, the rate of phagocytosis by these cells was not statistically different and ranged from 19–26 sec. The volume density of acid‐phosphatase‐positive Kupffer cells was 40% less in the C3H/HeJ mice. Similar differences were observed with other lysosomal enzymes including cathepsins B and H and dipeptidyl peptidases I and II. However, light and electron microscopy revealed a relatively normal number of Kupffer cells in livers stained for peroxidase, a nonlysosomal enzyme. The results suggest that the insensitivity of C3H/HeJ mice to endotoxin may be related in part to a lysosomal enzyme deficiency and a paucity of phagocytic Kupffer cells in these animals.

Aspects of beneficial endotoxin-mediated effects

SummaryThe status of hyperreactivity and hyporeactivity following the administration of endotoxin in a suceptible host represents phenomena which are of interest in an attempt to understand the role of endotoxins in pathophysiological events in general. Two experimental approaches designed to examine these events are reported herein; i.v. injection with minute concentrations of endotoxin (10 ng of a BOIVIN endotoxin fromE. coli 0111) induces tolerance against lethal doses of endotoxin (0.5 µg or 5.0 µg) within 24 h in hyperreactive NMRI mice that were infected 14 days before with BCG. Transfer of post-endotoxin serum from BCG infected mice, which contains a myriad of macrophage mediators and which induces nonspecific resistance to X-irradiation, renders a strain of mice (C3H/HeJ) that is hyporeactive to endotoxins, susceptible to the lethal effect of endotoxin. Studies of the role of the macrophage and its mediators in the experimental models described here may contribute to a further understanding of the mechanisms underlying endotoxin-induced biological activities.

Microvascular dysfunction in hepatic ischemia–reperfusion injury in pigs

Although hepatic ischemia-reperfusion (I/R) injury has been investigated for more than two decades, histopathological documentation is limited. As a result, three pig livers with I/R injury and three control livers were injected with colored media, cut into 14 segments, and examined by light microscopy together with microscopic map making. In livers with I/R injury, lobules were identified as being occluded or unoccluded. The proportion of the occluded lobules increased in a caudocephalic fashion, while that of the unoccluded lobules decreased (chi(2) for linear trend, P<0.0001). Especially in the occluded lobules, swollen hepatic plates displayed various forms of cellular distortion. Collapsed sinusoids containing leukocyte aggregation and shrunken central veins were observed together with reduced caliber of the contiguous sublobular veins. Portal vein constriction with loosening of the surrounding stroma suggestive of edema and hepatic artery dilation were also seen. Isolated arterioles and transintimal vasal outlets of the hepatic veins vasa venarum were dilated and frequently observed. In conclusion, I/R injury affected the liver parenchyma, the microvasculature, and its surrounding stroma. The heterogeneous distribution of occluded and unoccluded lobules is suggested due to the difference of vascular structure in various liver segments. The constrictive/obstructive changes in the portosinusoidal-hepatic vascular profile suggest a definite increase in resistance at presinusoidal, sinusoidal, and proximal postsinusoidal levels, resulting in an expansion of the arterial shunt circulation.

Histopathology of mixed anaerobic intraabdominal infection in mice

SummaryThis study characterized acute peritonitis and chronic abscess formation resulting from experimental mixed anaerobic infection withBacteroides melaninogenicus andFusobacterium necrophorum. At intervals after infection liver and spleen samples were obtained, fixed, and processed for histological examination. An acute to chronic infection progressed in mice infected with this mixture of anaerobic bacteria, whereas, no infection resulted when either organism was injected alone. Acute inflammatory cell infiltrates were noted in tissue samples at 12 h postinfection. Small, discrete areas of liver cell necrosis with neutrophilic infiltrates were observed as early as 24 h. By 48 h after infection the liver parenchyma was infiltrated with both acute and chronic inflammatory cells, with moderate to severe hepatocyte degeneration recognized at 72 h. Large intrahepatic abscesses were present in the subphrenic (upper lobe) area 2 to 6 weeks after experimental infection.

First observations using the scanning electron microscope in the early phase of endotoxin effect on terminal blood vessels

ZusammenfassungMittels REM wurde an Hamsterbackentasche und Kaninchenmesenterium die terminale Strombahn 50 min nach intravenöser Injektion von Endotoxin untersucht. Die hierbei erhobenen Befunde wurden zu den vitalmikroskopisch erkennbaren Veränderungen an der Mikrozirkulation in Parallele gesetzt.Wie schon aus vitalmikroskopischen Untersuchungen bekannt, treten erhebliche Deformierungen der Zelloberfläche besonders bei den Erythrocyten endotoxinbedingt auf. Aggregation von Thrombocyten, die in der terminalen Strombahn nach Endotoxingabe entstehen, konnten mittels der Stereoscantechnik als irreversible Thrombocytenthromben erkannt werden. In Aggregaten von Erythrocyten, die sich vielerorts in den kleinen Gefäßen häufig wandadhärent entwickeln, konnte im Rasterelektronenmikroskop Fibrin mit Thrombocyten nachgewiesen werden.SummaryBy Scanning Electron Microscopy the terminal bloodflow of the hamster cheek pouch and the rabbit mesenterium was studied for 50 minutes after intravenous injection of endotoxin. The resulting findings were compared with vitalmicroscopic observed changes of the microcirculation.Caused by endotoxin the surface of the cells especially of the erythrocytes undergo considerable deformations as already known from the vitalmicroscopic studies. Aggregation of platelets, which are formed in the terminal blood flow after application of endotoxin were recognized as irreversible aggregates (thrombus) of platelets by means of the Stereoscan-technique. Fibrin and platelets were identified in aggregates of erythrocytes which often appear disseminate in many of the small vessels adhaering to the wall.