Our study demonstrates increased expression levels of TLR2, TLR4, CD14 and Trem-1 protein expression on blood monocytes of patients with E. coli sepsis.
An interesting association was found for the Trem-1 expression on blood neutrophils with the cytokine production: Low Trem-1 expression on neutrophils was associated with low cytokine inducibility after stimulation with bacterial antigen. Trem-1 receptor on neutrophils might be involved in the induction of cytokines during sepsis [19
]. Low Trem-1 expression on neutrophils might be associated with hyporesponsiveness during severe human sepsis.
The host defence needs to detect invasion of pathogenic organism to induce an inflammatory response in order to control the infection. Pathogen recognition receptors (PRR) as TLR’s and CD14 are activated via pathogen associated molecular pattern, namely LPS for gram negative and LTA for gram positive bacteria. The activation of the immune cells causes via induction of gene transcription the production of cytokines.
Trem-1 is a relatively new discovered receptor expressed on neutrophils and monocytes, also involved in the innate immune response to bacterial pathogens [13
]. Activation of Trem-1 together with TLR ligands as LPS or LTA causes increased production of TNFα [20
]. It has been shown that in patients with sepsis and in mice with experimental LPS induced septic shock, Trem-1 is upregulated on neutrophils [14
]. In line with this data, Trem-1 was upregulated on neutrophils and monocytes in our study (Figure ). On the other hand, activation of Trem-1 and TLR4 caused a reduction of the production of the antiinflammtory cytokine IL-10 [21
In patients with severe E
sepsis we found a trend towards lower Trem-1 on neutrophils but higher soluble Trem-1 (compared to patients with simple sepsis, Table ). This data are in line with recent data from Oku et al., who have shown reduced Trem-1 on neutrophils but increase soluble Trem-1 in sepsis patients compared to SIRS patients [9
]. A correlation between soluble Trem-1 and sepsis severity has also been shown in recent papers [22
Why do we and others find a differential pattern in sepsis with high soluble Trem-1 and low neutrophil-surface TREM-1? Monocyte surface Trem-1 is shed by a metalloprotease [25
]. Pina et al. speculate that surface TREM-1 might also be shed from neutrophils explaining low surface Trem-1 in severe sepsis [25
Neutrophil surface Trem-1 might be involved in the capacity to produce cytokines. We found a positive correlation between Trem-1 expression on neutrophils and cytokine inducibility. Low Trem-1 expression on neutrophils was associated with low cytokine production after stimulation. In addition we found a negative correlation of Trem-1 on neutrophils and markers of immune activation. The higher levels of TNF-α in serum were found, the lower Trem-1 expression on neutrophils was measured. Since surface Trem-1 is activated early during sepsis [14
] and downregulated during severe sepsis, a negative loop might exist.
Our result might be interesting for the phenomenon of endotoxin hyporesponsiveness during sepsis [26
]. Sepsis can induce monocyte hyporesponsiveness, causing an anergic state of the immunesystem to gram negative and gram positive bacteria [26
]. Patients with hyporesponsiveness are immunosuppressed, and are prone to secondary infections associated with increased morbidity and mortality [26
]. In our study we can show a hyporesonsiveness in severe sepsis seen as reduced TNF inducibility (Table ) to E
endotoxin and to S
antigen (lipotheichonic acid, LTA). Since reduced surface Trem-1 on neutrophils was associated with reduced TNF inducibility, surface Trem-1 on neutrophils might be involved in hyporesponsiveness.
In addition to above data, as shown before, HLA-DR expression on monocytes was decreased in severe sepsis, probably influencing the capacity of antigen presentation [29
Beside a negative effect with immunosuppression, some data show a protective effect of Trem-1 inhibition via reduction of the overwhelming toxic proinflammation during septic shock. Bouchon et al. were able to show a protective effect of antibody blockade of Trem-1 in Mice [14
]. Injection of antibody 1 hour before endotoxin challenge was protective and caused reduction in mortality from 94 to 24%. The same effect was seen in experimental E
sepsis in mice [14
]. The effect was less pronounced when antibodies were given after the endotoxin injection. On the other hand complete silencing of Trem-1 via SI-RNA decreased bacterial clearance and increased mortality in a mouse model of infection [16
]. Partial inhibition of Trem-1 in the bacterial peritonitis model produced a significant survival benefit [16
]. Downregulation of the cytokine production during severe human sepsis might have a protective effect reducing the dangers of overwhelming cytokine production in the beginning and a detrimental effect later on since immunosuppression increase the risk of secondary infections [30
]. Hyporesponiveness might protect the person of overwhelming cytokine production but increases the risk of secondary infections.
Our study has several limitations. The parameters were only measured at sepsis diagnosis (transversal study), serial evaluation (longitudinal study) was not performed. Since the study population was small, more subjects are required for future trials. Newly discovered factors (e.g. Trem-2, Trem-like transcript 1 and 2) might also be involved.