Disturbance of endothelial integrity, especially in the gastrointestinal tract, is the hallmark of sepsis and septic shock. Poor perfusion of mucosal layers causes damage to the mucosal barrier and translocation of bacteria and their toxins into the systemic circulation 12. Increased paracellular permeability leads to extravasation of fluids and macromolecules such as albumin into the intercellular space. The resulting edema causes disturbances of microcirculation and organ function or even organ failure. Edema formation is, therefore, a diagnostic criteria for sepsis 3 and also a main pathomechanism for the development of multiple organ dysfunction syndrome and multiple organ failure.

Being the last step in the coagulation cascade and in the regulation of fibrinolysis, coagulation factor XIII (F XIII) plays a vital role as a fibrin stabilizing agent in fibrin clot formation. The active form of F XIII (F XIIIa) is a transglutaminase that builds cross-links between polypeptide chains, thus protecting fibrin from fibrinolytic enzymes4. F XIII may have multiple effects in addition to its effects in the coagulation system: it plays a role in cell adhesion and migration 5, prevents edema formation due to its influence on endothelial barrier function 678, influences constitution of cellular and extracellular matrix 910 and it seems to promote wound and bone healing 111213. Cytoplasmatic expression of the F XIIIa subunit in macrophages is related very closely to phagocytic activities and thus to leukocyte activation 14.

Some data suggest that F XIII also has an influence on the integrity of gut mucosa that has changed because of inflammation 151617. During sepsis and septic shock, plasma levels of F XIII can decrease 1819 and high levels of tumor necrosis factor alpha are found 20. Tumor necrosis factor alpha activates neutrophil granulocytes 21 and thus is able to induce secretion of lysosomal enzymes, such as human neutrophil elastase. Endothelial cell monolayer permeability is increased by these enzymes 22. Moreover, human neutrophil elastase degrades F XIII and can encourage hemostatic disturbances and organ dysfunction 23. Low F XIII activity in sepsis is associated with the severity of illness and organ damage 19.

In the treatment of sepsis, F XIII may have several potentially beneficial effects due to the inhibition of leukocyte activation, cell adhesion and migration, as well as by diminishing the magnitude of plasma extravasation and protecting the gut mucosal integrity, perfusion and function. We put this hypothesis to the test by investigating the effect of F XIII on intestinal functional capillary density (FCD), leukocyte adherence on venular endothelium as a parameter of leukocyte activation and mesenteric plasma extravasation of fluorescein isothiocyanate (FITC) labeled albumin by intravital fluorescence microscopy (IVM) in experimental endotoxemia.

Hemodynamic changes that occurred in the macrocirculation are given in Tables 1 and 2. Blood pressure and heart rate remained stable in the control group. The endotoxin challenge resulted in a significant fall in mean arterial pressure in the LPS and the F XIII groups after one hour. The transient pressure drop was followed by stabilization in these two endotoxemic groups two hours after LPS administration. Two hours after the endotoxin challenge there was no difference in arterial pressure values in either endotoxemic group compared to the control group. The heart rate was increased significantly in both endotoxemic groups at this point in time.

After three hours the white blood cell count in the control group was 7.8 ± 1.6 × 10³/mm³. At this time point, the white blood cell count was significantly lower in the LPS and F XIII groups compared to the controls (LPS group, 3.0 ± 0.8 × 10³/mm³; F XIII group, 3.7 ± 1.0 × 10³/mm³; p < 0.05 both groups versus control group).

After three hours of endotoxemia, F XIII activity fell to 49.8 ± 6.8% of baseline activity in the LPS group. F XIII administration resulted in an increase of F XIII activity to 111.7 ± 3.7 % after 1.5 hours and after 3 hours of endotoxemia the F XIII group values were in the range of the controls (Figure 1).

The changes in the FCD of the intestinal mucosa are shown in Figure 2. After two hours of endotoxemia we found a significant reduction of mucosal FCD in the LPS group (-18.5%; p < 0.01 versus control group). The F XIII administration attenuated the decrease in mucosal FCD (-3.7% compared to control; p < 0.05 versus LPS group). The FCD changes in the circular and the longitudinal muscle layers of the small intestine were not statistically significant (Table 3).

In the control group we determined a remarkable baseline rolling of the leukocytes along the endothelial lining of collecting (V 1) and postcapillary (V 3) venules. A significant decrease in the flux of rolling leukocytes was found in both the LPS and the F XIII groups (Table 4).

Figure 3 illustrates the count of firmly adherent leukocytes two hours after start of the endotoxin challenge. In the LPS group a 24-fold increase was noticed in the count of sticking leukocytes in the collecting venules (V 1) compared to the control animals (p < 0.05). In postcapillary venules (V3) the increase was 20-fold (p < 0.05). This increase of leukocyte adherence was not influenced by the F XIII treatment in the FXIII group.

Figure 4 depicts plasma extravasation measured by FITC-BSA leakage across the venular endothelium in the mesentery after one hour of endotoxemia. During the experiment we observed a tendency for FITC-BSA extravasation to increase. There was no statistically significant difference between the groups.

The administration of F XIII in septic animals prevented the decrease of mucosal FCD as seen in the untreated LPS group compared to controls. We could not detect an influence of F XIII administration on leukocyte to endothelium interaction and plasma extravasation. After three hours, endotoxemia resulted in an expected decrease in F XIII activity compared to the control group. The decrease of F XIII in our sepsis model correlates with findings of clinical investigations in septic patients 1819. However, through F XIII substitution in the treatment group we achieved F XIII levels comparable to controls after three hours of endotoxemia.

In our study we found a higher FCD in the F XIII group compared to untreated endotoxemic animals, indicating that perfusion in capillaries of the intestinal mucosa was protected by F XIII administration. It appears that not only the preservation of microcirculation but also pro-angiogenic properties are related to the effects of factor XIII. In vitro F XIII caused a dose-dependent enhancement of array formation in a Matrigel tube formation model, while in a rabbit cornea model, injection of F XIII led to neovascularization 28. As these effects occur over rather a long time (measured after 16 or 48 hours, respectively), the preservation of perfusion in the capillaries in our study can not be explained by these mechanisms. F XIII plays an important role in the final stage of the coagulation cascade, cross-linking fibrin monomers and stabilizing the fibrin clot. Thus, even an impairment of capillary perfusion as a result of augmented fibrin clot formation after the administration of this procoagulative substance is conceivable. In a sepsis model in rabbits it was shown that F XIII plays a role in promoting LPS induced disseminated intravascular coagulation with resulting organ damage 29. In our study, however, we saw an increase in functional perfused capillaries in the gut mucosa.

Perfusion in the capillaries could be impaired directly by the formation of tissue edema. This is why FCD should also be discussed in the context of plasma extravasation with consecutive edema formation. In our study endotoxemia did not result in the expected increased mesenteric plasma extravasation. We found that the LPS group tended to have higher plasma extravasation, although this difference was not statistically significant between the groups. Due to the high discrepancy in plasma extravasation values, a significant difference between the groups could be expected after testing a greater number of animals.

A protective effect of F XIII on the endothelial barrier function was shown in several other studies 678303132. Hirahara and colleagues 32 investigated the effect of F XIII on permeability in guinea pig endothelial cells, enhanced by intradermal injection of anti-endothelial cell antiserum administered into the dorsal skin. Antiserum or a mixture of antiserum and F XIII was injected after Evans blue injection, and later the quantity of Evans blue was determined at each injection site. F XIII had a suppressive effect on dye leakage and on swelling induced by the antiserum. The authors assumed that F XIII plays an important role in inflammatory sites and that it may act as an anti-inflammatory protein. Due to its anti-inflammatory effects, F XIII also might contribute to preservation of perfusion in the capillaries. In a model of cultured monolayers of porcine aortic endothelial cells and in saline-perfused rat hearts, F XIII reduced the albumin permeability of endothelial monolayers 6. The increase in myocardial water content in ischemic-reperfused rat hearts was prevented, indicating that activated F XIII reduces endothelial permeability 6.

The protective effect of F XIII on vascular endothelium integrity has also been documented in clinical investigations. In a prospective investigation of perioperative cardiac edema formation requiring a delayed sternal closure in children, F XIII or placebo was substituted preoperatively. The substitution of F XIII reduced the incidence of myocardial swelling and the authors concluded that the clinical application of F XIII may have a valuable therapeutic benefit in cases of leakage syndrome during extracorporeal circulation in congenital heart surgery 8.

F XIII administration in our model tended only to lower plasma extravasation in comparison to the endotoxin group and this was not statistically significant for mesenteric venules.

After administration of F XIII, a tendency to attenuate the leukocyte adherence could be noticed in our study. Tissue transglutaminase and F XIIIa are expressed on the surface of monocytic cells and some evidence indicates the involvement of transglutaminases in cell adhesion, for example, they can influence adhesion of monocytic cells on fibronectin 33. Factor XIIIa promoted adhesion and spreading of different cells, such as human liver cells, human leukemia cells, human melanoma cells and bovine aortic endothelial cells, to F XIIIa coated surfaces in vitro 5, indicating that F XIIIa itself mediates cell adhesion. In an adhesion assay, it was shown in vitro that F XIIIa mediates adhesion of the human microvascular endothelial cell line HMEC-1 and that F XIIIa binds to HMEC-1 cells in solution. This has been verified by a flow cytometric analysis 34.

Integrins as transmembrane receptors mediating cell-cell interactions are expressed on leukocytes. They play an important role in leukocyte transendothelial migration. Some integrins, especially a₄ß₁ and a₉ß₁ integrins, are ligands for F XIII 35. Taking these findings into consideration, as well as the above mentioned sealing effect of F XIII, enhancement of the interaction between leukocytes and the vascular endothelium by F XIII could be expected. The role of F XIII in relation to the interaction of leukocytes to the vascular endothelium, however, requires further investigation.

Factor XIII can protect mucosal capillary perfusion against endotoxin-induced impairment in an experimental sepsis model in rats. Because of the importance of preservation of intestinal perfusion in the early treatment of sepsis and septic shock, early F XIII administration might be considered in septic patients but requires clinical approval.

• Factor XIII protected the perfusion in mucosal capillaries against endotoxin-induced impairment in an experimental sepsis model in rats.

• In septic patients, F XIII administration might be favourable but requires clinical approval.

BSA = bovine serum albumin; F XIII = factor XIII; FCD = functional capillary density; FITC = fluorescein isothiocyanate; IVM = intravital fluorescence microscopy; LPS = lipopolysaccharide.

The authors declare that they have no competing interests.

JB and OVH coordinated the study and drafted the manuscript. CL, OR, MG, TU, KM and DP performed the IVM, collected the data and helped to draft the manuscript. SZ performed the estimation of F XIII levels and helped to draft the manuscript. CS, WJK and ChL conceived and designed the study and performed the statistical analysis.