Endotoxaemia modulates Toll-like receptors on leucocytes in humans


Bernd Jilma, MD, Department of Clinical Pharmacology, Vienna University Hospital School of Medicine, Waehringer Guertel 18–20, A-1090 Vienna, Austria. E-mail: Bernd.Jilma@univie.ac.at


Summary. The modulation of Toll-like receptors (TLR) 1, 2 and 4 was studied during experimental human endotoxaemia. Healthy volunteers received 2 ng/kg of lipopolysaccharide (LPS) endotoxin (n = 10). TLR1, 2 and 4 expression occurred on monocytes and neutrophils, with monocytes expressing higher baseline levels of TLR2. LPS infusion downmodulated TLR4 expression on neutrophils, with maximal downregulation occurring at 24 h (−62% from baseline; P < 0·03 versus baseline). Monocyte TLRs were upregulated in vivo (TLR1 and 2), and in vitro (TLR1, 2 and 4) 8 h after LPS bolus (P < 0·05 versus baseline). Therefore, neutrophils and monocytes differentially express surface TLRs, and endotoxaemia differentially regulates TLR expression.

The Toll-like receptors (TLR) facilitate innate immunity by detecting products that are unique to invading microorganisms (Takeuchi & Akira, 2001). Ten members of the TLR family have been identified in humans, and several of them appear to recognize specific microbial products, including lipopolysaccharide (LPS), bacterial lipoproteins, peptidoglycan and bacterial DNA. Recognition of these ligands leads to the activation of intracellular signalling pathways, which upregulate a wide array of inflammatory modulators that contribute to the early host cell response (Takeuchi & Akira, 2001; Sabroe et al, 2002).

Signals initiated by the interaction of TLRs with specific microbial products induce a subsequent inflammatory response. While the exact expression pattern of each TLR varies, TLR2 and TLR4 are expressed on peripheral blood leucocytes (Mackman, 2000; Guha & Mackman, 2001). TLR1 has previously been detected on monocytes (Visintin et al, 2001) but scarce data exists relating to its expression on other haematopoietic cells (Muzio et al, 2000; Kurt-Jones et al, 2002).

Leucocyte responsiveness to LPS is dependent upon CD14 and receptors of the TLR family. Regulation of TLR2 and TLR4 expression by LPS is considered to be one of the mechanisms by which the overall responses of immune cells to bacteria are controlled. However, little is known about the regulation of TLRs in endotoxaemia. Thus, we investigated the regulation of TLR1, 2 and 4 in human endotoxaemia.

Patients and methods

The study was approved by the Institutional Ethics Committee. Ten healthy men aged 19–35 years were included in the study after written informed consent was obtained. Our LPS model has been described in detail previously (Homoncik et al, 2000). All volunteers were free of medication for at least 3 weeks prior to the study day. All subjects received 2 ng/kg of LPS (National Reference Endotoxin, Eschericha coli; USP, Rockville, MD, USA) as a bolus infusion over 2 min.

Sampling times were selected based on the kinetics of CD11b receptor upregulation in previous trials (Hollenstein et al, 2002). Blood samples were collected by venepuncture into EDTA anticoagulated Vacutainer tubes (Becton Dickinson, Vienna, Austria) before (0 h) and after LPS infusion at 2, 4, 6, 8 and 24 h. TLR1, 2 and 4 receptor expression was quantified by flow cytometry using a FACSCalibur flowcytometer (Becton Dickinson, San Jose, CA, USA) as previously described (Jilma et al, 2000), using antibodies that were phycoerythrin-labelled (EBioscience, San Diego, CA, USA). Results are presented as the binding index (BI), which was computed using the following equation: percentage of positive cells multiplied by the mean fluorescence intensity (MFI)/100. Cell counts were obtained using a cell counter (Sysmex, Milton Keynes, UK).

For the in vitro trial we used LPS at a concentration of 50 pg/ml, which is consistent with the administered dose in the clinical trial. Nine hundred µl of EDTA anticoagulated blood was incubated with 100 µl of LPS diluted in phosphate-buffered saline for 2–4 h at 37°C before staining and flow cytometric analysis.

Data is expressed as the mean and standard error of the mean (SEM). Non-parametric statistics were applied. All statistical comparisons were done using a Friedman analysis of variance (anova) and a Wilcoxon test. A two tailed P-value of < 0·05 was considered significant.


The mean baseline neutrophil level was 2·5 ± 0·14 × 109/l. As expected, neutrophil counts dropped sharply 1 h after LPS infusion, and increased 3·5-fold above baseline from 2 h onwards (data not shown). Monocytes were barely detectable at 2 h as determined by flow cytometric analysis.

As previously described (Hollenstein et al, 2002), endotoxin increased the expression of activation marker CD11b on neutrophils and monocytes by approximately 200% (BI) after 2 h (P < 0·05 versus baseline); the increase in CD11b expression persisted for at least 6 h (data not shown).

Both monocytes and neutrophils expressed TLR1, TLR2 and TLR4 on their surface (Fig 1). LPS significantly downmodulated TLR4 expression on neutrophils by 62% ± 10%, with a maximal downregulation observed 24 h after LPS infusion (Fig 2). Only a trend-wise downregulation (P = 0·13) of TLR1 and TLR2 on neutrophils was observed. TLR1, 2 and 4 BI decreased on monocytes at 2 h, but the surface expression of TLR1 and 2 significantly increased at 8 h after LPS bolus (P < 0·05 versus baseline; Fig 2), and returned to baseline at 24 h.

Figure 1.

The basal expression of Toll-like receptors (TLR) 1, 2 and 4 on neutrophils and monocytes. The forward scatter (FFS) and side scatter (SSC) pattern of peripheral blood leucocytes was used to gate neutrophils (A) and monocytes (B). The basal surface expression of the Toll-like receptors (TLR) 1, 2 and 4 on neutrophils (C, D and E, respectively, red-filled area) and monocytes (F, G and H, respectively, green-filled area) compared with isotype control (blue). An LPS stimulated sample is shown [2-h incubation with 50 pg/ml, red line (neutrophils) and green line (monocytes)]. Subpopulations of leucocytes were gated according to their FSC/SSC position.

Figure 2.

The effects of endotoxin infusion on cell counts of neutrophils (▴, A), and monocytes (▪, F) and the surface expression of the Toll-like receptors (TLR) 1, 2, 4 and CD11b on neutrophils (B, C D and E respectively) and on monocytes (G, H, I and J respectively). Volunteers received 2 ng/kg of E. coli endotoxin (n = 10). Data is presented as mean ± SEM binding index (BI). P < 0·05 versus baseline for all parameters.

In order to clarify whether the modulation of surface expressed TLRs on leucocytes was a direct effect of endotoxin, we incubated whole blood with 50 pg/ml of LPS in vitro. LPS downregulated TLR4 on neutrophils after 2 h and 4 h incubation in vitro (P < 0·05 versus control and baseline). The decrease of TLR4 on neutrophils in vivo was reproducible in vitro (−55% versus baseline, P < 0·05). LPS upregulated all three receptors on monocytes in vitro after 2 and 4 h (+70% in TLR1, +46% in TLR2, +76% in TLR4 at 2 h, P < 0·05; compare Fig 1).


Our data shows that un-activated monocytes and neutrophils expressed all three TLRs assessed. Monocytes expressed relatively high levels of cell surface TLR2 when compared to neutrophils (Fig 1 and 2).

Endotoxaemia induced a marked and protracted downregulation in TLR4 expression on neutrophils in vivo (−62%, P < 0·05), which is currently regarded as the main signalling receptor of LPS (Mackman, 2000) on neutrophils.

A similar degree of downregulation was also observed in vitro, indicating a direct effect of LPS. Such downregulation could possibly result in a decrease in neutrophil responsiveness to subsequent LPS stimulation, which merits examination in further studies.

LPS infusion initially decreased the binding of antibodies against TLR1, 2 and 4 on circulating monocytes (Fig 2). However, in vitro experiments showed a marked upregulation of these receptors in response to LPS after only 2 h. Thus, the initial dip in the in vivo expression of TLRs could be explained by monocytopenia as a subpopulation of monocytes expressing lower levels of TLRs are less likely to be activated and thus remain in the circulation rather than marginate.

Another interesting finding in the monocyte subpopulation is the similarity between the time course of TLR1 and 2. In vitro experiments suggest that TLR1 alone is not able to mount an inflammatory response and is co-dependent on TLR2 (Wyllie et al, 2000). This may provide a biological explanation for the homology in the regulation of TLR1 and TLR2 on monocytes by LPS in our in vivo trial. The later in vivo upregulation of TLR1 and 2 is consistent with both the magnitude of increase observed in vitro after LPS incubation, and the reappearance of monocytes in the circulation.

In conclusion, endotoxaemia differentially regulates TLR1, 2 and 4 on both neutrophils and monocytes.