The inflammatory host response as an amplification loop of Shiga toxin toxicity

PALERMO MS

Buenos Aires

Simposio; 7th International Symposium on Shiga -Toxin (Verocytotoxin)- Producing Eschericchia coli (VTEC2009); 2009

Asociación Argentina de Microbiología

D2) Since epidemic form of HUS  is an important  public health problem in our country and considering that several questions about its pathogenic mechanism remains elusive,  we began to study more than ten years ago the host response looking for the answers mainly to the following questions: 1)      Why do only 10% of STEC-infected children develop HUS?2)      Why is the kidney the main target organ?Although multiple factors probably converge to define the clinical evolution of STEC infections,  host immune response may actively participate since it is challenged by the central pathogenic factors, which are LPS and shiga toxin.    Then, our working hypothesis is, that The innate and inflammatory responses may be able to modulate Shiga toxicity, and to define the evolution of HUS, D3. During this talk I will refer to the results obtained working on two experimental approaches: the mouse model by iv injection of purified Stx2, with or without LPS, and peripheral blood leukocytes isolated from HUS patients at diagnosis and hospitalization, previously to any treatment, including dialysis and blood transfusion  D4.  In the mouse model we found out that the in vivo toxic effect of shiga toxin could be enhanced or decreased by the same dose of LPS depending on the time elapsed between both injections. In this figure, for example 5 ug LPS /mice injected one hour before STX2 accelerated the death of mice, while the same LPS dose  injected 18 hrs before Stx delayed it. LPS is the common bacterial component of gram-negative bacteria with one of the highest pro- inflammatory activity which is proportional in intensity and time duration to the dose injected. But always this inflammatory reaction is followed by a proportional anti-inflammatory phase, leading to host to counteract the inflammation and return to the homeostasis. In this context, these results are interpreted  as when Shiga toxin is injected simultaneously to the onset of the inflammatory reaction, toxicity of Shiga is potentiated. On the contrary, when Shiga toxin is injected during the anti-inflamamtory phase of the response, toxicity of Shiga is decreased. Therefore,  from the mouse model we have the former evidence that Inflammatory response can modulate STx toxicity and could participate during the evolution of HUS. D.5  In addition, we found out in mice injected with purified Shiga toxin several signs of inflammation. As a peak of seric TNF, the increase of activating membrane markers on neutrophils ,  and the increase in their activity (assayed as adhesion and cytotoxicity). More important, when mice were depleted from PMN, the toxicity of shiga toxin was significantly reduced.. In conclusion, During the injury induced by Stx2 inflammatory response and PMN are activated, and this population plays an important role in Stx toxicity D6. However, former studies carried out in circulating PMN from patients, showed decreased expression of all activation markers studied (CD16, CD11b and MPO production, as well as a decreased  citotoxic capacity) both, when compared with healthy or acute uremic children,   D7. These differences even persisted when comparing with PMN from children coursing a non-related bacterial infection D.8  In addition, neutrophils from HUS patients showed a reduced intracelular content  of different types of granules, such as CD66b, b glucuronidase These results suggested that PMN from HUS patients are exhausted  (similar to platelets); probably as a consequence of a previous and strong activation process, which might take place during the window time that elapses between diarrhea and HUS-diagnosis. D9. More significant, through retrospective studies we found a strong correlation between the level of PMN-deactivation at the moment of diagnosis and the severity of the disease, for all degranulation markers.  Since neutrophils are ending cells, activation stimuli induce degranulation and prolongation of its half-life, decreasing its apoptotic rate. As can be seen in this slide, decrease in the apoptocic rate also correlates with severity. That means, the more deactivated neutrophils those children have, the more severe disease they present.  D10. What did we find when monocytes,  as the other cellular component of innate immunity,  were studied??? In mouse model, depletion of hepatic and splenic macrophages diminishes Stx toxicity. On the other hand, we studied peripheral monocytes isolated from HUS patients on the acute period, as stated before, previously to transfusion or dialysis. We found that HUS patients showed an increased percentage of the inflammatory subset of monocytes characterized by the double expression of CD14 and CD16 . This subpopulation is also expanded in the control group of infected children and has been reported to be enhanced in other infections, such as  sepsis and AIDS. This subset tends to disappear simultaneously with clinical recovery. This subset has been reported to be more mature, and  to produce higer amounts of inflammatory cytokines. D 11. When results were retrospectively analyzed, and patients classified in mild, moderate and severe, we found that most severe cases showed highest number of inflammatory monocytes. In addition, monocytes from these patients showed the maximal cytokine production, for both TNF and IL-10. Altogether these results indicate that Monocytes  from HUS patients, at diagnostic moment, show signs of having received a previous activating stimulus which differentiate them to this inflammatory subset D. 12 BUT: Since inflammation and thrombosis are systemic processes Why is the kidney the main affected organ?? It is possible that the specific damage in the capillary beds of kidney is not merely the result of mechanical trapping of platelets and leukocytes, but involves specific interactions with the endothelial wall. Chemokines and their receptors have been proposed to play a role in homing of leukocytes to selected sites or organs. Taking this into account a very interesting candidate is Fractalkine, since it is preferentially expressed in inflamed or injured endothelium and epithelium of kidney and central nervous system. According to its structure, FKN has a dual function: as a transmembrane protein it has  proadhesive functions, but also Fractalkine can be cleaved into a soluble fragment containing the chemokine domain and it is capable of attracting immunocompetent cells carrying the fractalkine Receptor Moreover,   * The expression of fractalkine and the presence of CX3CR1-expressing cells in the kidney, have been demonstrated in patients with various types of nephropathies. * anti-CX3CR1 antibody treatment dramatically blocked leukocyte infiltration into the glomeruli, improving renal function. In order to answer whether FKN is involved in HUS nephropathy D13. we evaluated blood leukocyte populations expressing the receptor for FKN in  HUS patients. As can be seen in this figure, the whole population of  cells expressing the receptor for FKN is significantly reduced in HUS patients.  These cells include mainly: monocytes, Natural Killer cells and a small subset of cytotoxic T cells. All these cells have a high cytotoxic potential, regardless its origin or recognition pathway.  Very striking is the fact that none of the other clinical groups show reduction of these cells carrying the receptor for FKN. Neither infected children who have a similar percentage of the inflammatory subset of monocytes. D14. When we analyzed other membrane markers in monocytes, we found a significant reduction of the major circulating Mo subset in healthy people, which are positive for FKN-R, and CD62, and negative for CD16. This reduction is mainly at the expense of an increase in the Mo population lacking the receptor for FKN.   D15. Similarly, we observed a significant decrease in the percentage of NK cells. Again this is an observation unique for HUS patients, distinct from infected or uremic patients. As can be seen in this picture, the decrease in NK cells was seen only in the CD56dim subset, which carry the  fractalkine receptor. D16. The specific reduction in the subsets of leukocytes carrying  the receptor for FKN suggests that this chemokine over expressed on injured renal endothelium and epithelium plays a pathogenic role attracting cytotoxic cells that  bear FKN-receptor enhancing the Stx –initiated damage. In fact, in this picture we can see immunstainning of cells carrying the fractalkine receptor or a marker of macrophagic population that have been entrap in the renal glomerular and tubular tissues, D17. To directly evaluate if Mo in contact   Shiga toxin, are able to interact with endothelial and renal epithelial cells through the FKN way and contribute to tissue damage, we did in vitro studies using HUVEC and renal epithelial cells. We incubated TNF-sensitized endothelial cells with Shiga toxin, basal Mo and Mo preincubated with Shiga toxin, and extensively washed. After overnight incubation we assayed the toxic effects by counting living cells. We observed the direct effect of Shiga, but also a high cytotoxicity  mediated  by Mo preincubated with Shiga toxin, which was partially blocked by Shiga toxin antibody or soluble fractalkine. D18: A similar effect was observe in human renal tubular epithelial cells. D.19.In order to determine if migration and adhesion of inflammatory cells to reanl endothelium is involved in  the pathogenic mechanism of HUS, we tested in vivo toxicity of STx2 in mice KO for different chemokine receptors specific for monocytes and PMN?. As we can see in this figure >KO mice for CCR1, MCP1 and fractalkine were protected against Stx toxicity. These preliminary but promising results encourage us the search of chemokine- blockingn agents  as therapeutic weapons  to prevent HUS development. D20: This slight shows a schematic representation of the pathogenic mechanisms of HUS. The direct cytotoxic effect of STx on endothelium and epithelium. But also, the contribution of all components of the innate response: cytokines, which induce or enhance receptors, adhesion molecules and chemokines. And the cellular populations: monocytes, PMN, and NK cells, which upon activation exert their cytotoxic potencial enhancing the endothelial injury initiated by STX. In conclusion , during HUS a unique scenario is present, distinctive from other infective disease, in which, an inflammatory reaction (also potentiated by the stimulous that represents an injured endothelium that has also triggered the thrombotic and complement cascade) act as an amplification loop, this inflammatory reaction through cells expressing a high cytotoxic activity, enhances the endothelial damage. D21. As concluding remarks I want to point out that We have demonstrated the strong inflammatory response that takes place in STEC-infected children previously to clinical diagnosis of HUS. And We suggest that the qualitative and quantitative features of this inflammatory process is determinant in the evolution to HUS. This conclusion leads us to propose The search of new prognosis parameters in STEC-infected children that allow to predict HUS evolution; and the development of New therapeutic approaches might be applied previous to HUS, since at that moment pathogenic mechanisms have been completely developed All this investigation could be possible thanks to the work of all this people, and the central collaboration of  Children Hospitals. MUCHAS GRACIAS