Sepsis is the main cause of death in intensive care units (ICUs) worldwide, representing 45% of all-cause mortality, and accounts for more than 5.3 millions of deaths per year (1). It is now well established that during sepsis, both hyperinflammatory and immunosuppressive processes occur, and that both contribute to organ failure and mortality (2). Nevertheless, there is still no adjunctive treatment for this lethal condition, as many (immunomodulatory) therapies have failed to improve survival in clinical trials (3). Therefore, there is an urgent need for a better understanding of the mechanisms involved in the systemic inflammatory response in sepsis, and human models that closely resemble it are instrumental to achieve this goal.
For more than 100 years, the bacterial membrane product endotoxin (lipopolysaccharide [LPS]) has been administered in humans, in the early years with therapeutic purposes (4), and later to elucidate the systemic response in the so-called “experimental human endotoxemia model” (5). When administered i.v. in moderate to high doses (1–4 ng/kg of Escherichia coli LPS), clinical responses are evident within 1 h; subjects experience symptoms that resemble a flu with fever and there is an increase in respiratory frequency and heart rate (6). Within the first 1.5 to 3 h, circulating concentrations of most inflammatory cytokines reach their maximum (7). Furthermore, the immune cell composition in the circulation changes dramatically after endotoxin administration. There is a transient mono and lymphocytopenia in the first hours after endotoxin administration followed by a profound leucoytosis later on, mainly due to an increase in neutrophils (8). In more detail, monocytes virtually disappear from the circulation 1 h after induction of endotoxemia, returning to baseline and elevated levels at 4 to 8 h post administration, respectively (9). Within the lymphocyte compartment, there is a pronounced decline in absolute B and T-cell counts 2 to 8 h after endotoxin challenge (10). All these changes return to normal 24 h after endotoxin exposure. These effects are highly reproducible and capture several pathological features of the early phase of sepsis. Examples of these similarities include increased concentrations of both pro and anti-inflammatory cytokines in the circulation, together with an increment in the numbers of mature and immature neutrophils and a substantial decrease of circulating T and B cells (6). In addition, monocyte deactivation is observed in both sepsis and human endotoxemia, exemplified by a lower cell surface expression of human leucocyte antigen (HLA-DR) (11).
Nevertheless, there remains controversy as to whether human endotoxemia is an appropriate model for sepsis, and if it is suitable to investigate therapeutic strategies for this disease (6). An important issue is that most research to date, especially concerning the adaptive immune system, has focused on the very early phase after endotoxin administration (first 24 h). The long-term effects in the adaptive immune compartment have been largely overlooked, with only very limited information available (12).
In the present study, we aimed to provide additional insights in the endotoxin challenge model and its relevance for sepsis. Therefore, we investigated both the short and long-term effects of endotoxin administration on the immune cell composition, and compared it with the profiles obtained in patients with septic shock.
PATIENTS AND METHODS
Subjects and ethics
All study procedures comply with the Declaration of Helsinki and the Good Clinical Practice guidelines.
Healthy volunteers challenged with endotoxin
Data were obtained from 7 healthy male volunteers who participated in a previously published experimental human endotoxemia study (13). The study protocol was approved by the local Ethics Committee (CMO Arnhem-Nijmegen, CMO 2013–600) and registered at ClinicalTrials.gov (NCT02085590). Written informed consent was obtained from all subjects. All participants were challenged with endotoxin (single intravenous bolus of 1 ng/kg of E. coli LPS). Four subjects also received another intervention (vaccination with gamma-irradiated Bacille Calmette–Guérin), whereas 3 received placebo. Bacille Calmette-Guerin vaccination did not have effects on any of the parameters studied (13). Blood samples were obtained before endotoxin administration (0 h), 4 h, 2 days, and 20 days after endotoxin administration. Further details of the endotoxemia study procedures are reported elsewhere (13).
Twenty-three patients admitted to the ICU with septic shock were enrolled between November 2014 and March 2017 in an ongoing prospective observational study. The protocol was approved by the local Ethics Committee (CMO Arnhem-Nijmegen, CMO 2016–2923). Patients were included in the study according to conventional criteria for septic shock: a suspected infection and 2 or more systemic inflammatory response syndrome criteria as well as requiring vasopressor therapy to maintain blood pressure (14, 15). The ethical approval for the ongoing observational study dictates that samples can only be obtained from patients with an arterial line in place, so venipuncture is not required to obtain samples. Therefore, 10 of the initial 23 patients enrolled, were excluded after day 1 and only 13 remained with an arterial line during the follow-up period of the study. Exclusion criteria were chronic bowel disease, immune suppression, and documented chronic renal disease (serum creatinine concentration >150 μmol/L). The study was carried out in the Netherlands in accordance with the applicable rules concerning the review of research ethics committees and informed consent. All patients or legal representatives were informed about the study details and could refuse to participate. Blood samples were taken less than 24 h after ICU admission (day 1) and 3 to 5 and 7 to 9 days after ICU admission. The follow-up period was based on previous experience; it represents a good trade-off between length of follow-up and adequate numbers of patients who are still on the ICU with an arterial line in place from which samples can be obtained.
Reference data from healthy controls
Reference data were obtained from 20 age and gender-matched healthy controls (HC; 12 male and 8 female; median [range] age of 62.5 [29–75]) that were immunophenotyped using the same flow cytometry panels and gating strategy as used in the present study. The HC samples were collected as part of a larger group of 516 individuals that participated in the human functional genomics project, a previously published study (16), approved by the ethics committee of Radboud University Nijmegen (no. 42561.091.12). The inclusion criteria are described in http://www.humanfunctionalgenomics.org.
Blood sample collection
Blood was collected in BD Vacutainer (BD Biosciences, Erembodegem, Belgium) spray-coated potassium 2 ethylene diamine tetraacetic acid tubes. White blood cell counts were determined by a cell counter (Coulter Ac-T Diff cell counter; Beckman Coulter, Fullerton, CA) and used to calculate the absolute numbers of cluster of differentiation (CD)45+ leukocytes identified by flow cytometry as described in detail previously (16).
Briefly, total leukocytes were obtained after erythrocyte lysis in isotonic NH4CL buffer and washed twice with phosphate-buffered saline (PBS). Peripheral blood mononuclear cells (PBMCs) were isolated by means of density gradient centrifugation (Lymphoprep; Axis-Shield PoC AS, Oslo, Norway), washed twice in PBS, resuspended in medium (RPMI 1640 medium supplemented with pyruvate [1 mM], glutamax [2 mM], penicillin [100 U/mL], streptomycin [100 mg/mL; Thermo Fisher Scientific, Waltham], 10% human pooled serum, and 0.3% sodium citrate [Merck, Darmstadt, Germany]), after which cell counts were performed.
To analyze plasma cytokines, ethylenediaminetetraacetic acid-anticoagulated blood was centrifuged at 2,000 × g at 4°C for 10 min immediately after withdrawal, and plasma was stored at −80°C until analysis. Concentrations of tumor necrosis factor (TNF)-α, IL-6 and Il-10 were measured batch wise using a Luminex assay (Milliplex, Billerica, CA).
Half a million of total leukocytes or PBMCs were used to analyze surface and intracellular (only PBMC) markers with the 10-color Navios flow cytometer (Beckman Coulter, Fullerton, CA). In brief, cells were transferred to a V-bottom plate and washed twice with PBS + 0.2% bovine serum albumin (BSA; Sigma-Aldrich, St. Louis), stained for 20 min at room temperature in the dark with the fluorochrome-conjugated monoclonal antibodies (mAbs) of interest and washed twice with PBS + 0.2% BSA. For surface staining of monocytes, natural killer (NK) cells, T and B cells, the following conjugated mAbs were used: CD3-PE/ECD/PB (Beckman Coulter; UCHT1), CD4-PC5.5/PB (Beckman Coulter; 13B8.2), CD4-AF700 (eBioscience, San Diego; RPA-T4), CD8-APC-AF700/APC-AF750 (Beckman Coulter; B9.11), CD14-ECD (Beckman Coulter; RMO52), CD16-FITC (Beckman Coulter; 3G8), CD19-APC-AF750 (Beckman Coulter; J3–119), CD27-PE Cy5.5 (Beckman Coulter; 1A4CD27) CD45-KO (Beckman Coulter; J33), CD45RA-FITC/ECD (Beckman Coulter; ALB11/2H4LDH11LDB9), CD56-APC (Beckman Coulter; N901), HLA-DR-PE (Beckman Coulter; immu-357). For intracellular staining, samples were permeabilized and fixed according to manufacturer's instructions (eBioscience). Cells were labeled with the following conjugated mAbs: interferon (IFN)-γ-PC7 (eBioscience; 4SB3), Ki67-AF488 (BD; B56), and TNF-α-FITC (BD; MAB11) for 30 min at 4°C in the dark. Isotype and fluorescence minus one controls were used for gate settings. The data were analyzed using Kaluza V1.1 software (Beckman Coulter). The gating strategy used for analysis of immune cell populations is depicted in the figure shown in figure S1 (Supplemental Digital Content 1, http://links.lww.com/SHK/A788).
Intracellular cytokine expression
PBMCs were stimulated with phorbol-12-myristate-13-acetate (PMA), ionomycin, and brefeldin A (respectively, 12.5 ng/mL, 500 ng/mL and 5 μg/mL; Sigma-Aldrich, St. Louis) for 4 h at 37°, 5% CO2. Cytokine expression was determined by measuring the intracellular production of IFN-γ and TNF-α by flow cytometry as described above.
Statistical analyses were performed using GraphPad Prism 5 (Graphpad software Inc., La Jolla, CA). Nonparametric Friedman tests followed by Dunn's post-hoc tests were used to analyze the human endotoxemia data. Nonparametric Kruskal–Wallis tests with Dunn's post-hoc tests were performed to compare data of sepsis patients with the reference HC data. Statistical significance is denoted as ***: P < 0.001, **: P < 0.01, and *: P < 0.05 compared to baseline or HC. All indicated values are median values of absolute numbers or percentages ± range.
Twenty-three patients with septic shock were included in the study. The median [range] age of the patients was 67 [28–82]. The median [range] Acute Physiology and Chronic Health Evaluation II score was 24 [8–36]. Thirteen patients remained at the ICU for 3 or more days. Patient characteristics are described in Table 1.
Seven healthy male volunteers (median [range] age of 20 [18–25]) were challenged with endotoxin. Endotoxin administration resulted in flu-like symptoms and fever, typical of endotoxemia (data not shown). TNF-α, IL-6, and IL-10 concentrations in plasma were measured in the experimental endotoxemia group during the first 8 h post administration and in sepsis patients on days 1 to 9 after onset of septic shock (see figure S2, Supplemental Digital Content 2, http://links.lww.com/SHK/A789). As expected, a profound increase in the plasma levels of TNF-α, IL-6, and IL-10 was observed 2 to 4 h after endotoxin administration, which returned to baseline levels 8 h post endotoxin. In the sepsis patients, high concentrations of TNF-α, IL-6, and IL-10 were observed at the onset of septic shock (day 1). Thereafter, TNF-α levels gradually declined (Fig. S1 D, http://links.lww.com/SHK/A788), while IL-6 and IL-10 concentrations showed a sharp reduction at day 3 and remained unchanged thereafter (Fig. S1 E and F, http://links.lww.com/SHK/A788).
Immune cell subpopulations change over time in human endotoxemia and sepsis
For both the endotoxemia and sepsis group, we summarized the absolute numbers and/or percentages of innate and adaptive immune cells measured at each time point (see Table, Supplemental Digital Content 3, http://links.lww.com/SHK/A790). Many differences as well as similarities were found between both groups. Below, several immunological patterns are explained in greater detail.
Experimental human endotoxemia results in a long-term increase in the abundance of monocytes and NK cells
First, we identified circulating granulocytes, monocytes, NK cells, T cells, and B cells in both cohorts; the flow cytometry gating strategy is depicted in figure S1 (Supplemental Digital Content 1, http://links.lww.com/SHK/A788).
In our experimental human endotoxemia cohort, we studied changes in immune cell populations 4 h, 2 and 20 days after endotoxin administration. We considered 4 h and 2 days after endotoxin challenge as short-term effects, and 20 days after endotoxin challenge as long-term effects. In line with previous data on short-term effects of endotoxemia (7), granulocyte numbers increased 4 h after endotoxin administration, while T cell and B-cell numbers were decreased at this time point. These cell numbers returned to baseline at 2 days and were unchanged at 20 days post administration (Fig. 1A). Monocyte numbers were unchanged at 4 h and 2 days, while NK cells decreased at 4 h and returned to baseline levels at 2 days post administration (Fig. 1A). Different from the 2 days, at 20 days post administration, we found significantly increased numbers of monocytes and NK cells (Fig. 1A), in particular of NK CD56dim cells (see Table, Supplemental Digital Content 3, http://links.lww.com/SHK/A790).
Next, we investigated whether the effects observed after 20 days in the human endotoxemia model were also present in 13 patients of whom we could obtain samples up to 7 to 9 days after the onset of septic shock. In line with previous work (7), we found similarities between sepsis patients at day 1 and the endotoxemia model at 4 h, i.e., increased granulocyte and decreased NK cell, T cell, and B-cell numbers (Fig. 1B). Nevertheless, we also observed that, except for granulocytes, this pattern did not recover at any time point measured in septic shock patients, which is markedly different from what we found 20 days postendotoxin administration (Fig. 1B).
We explored the dynamics of the different subpopulations of monocytes and T cells present in both the endotoxemia model and sepsis in more detail.
The monocyte subset composition and HLA-DR expression in experimental human endotoxemia and sepsis differ
Because we observed increased monocyte numbers 20 days after induction of experimental human endotoxemia, we wished to obtain further insight into the distribution of the CD14+CD16− classical, CD14+CD16+ intermediate, and CD14lowCD16++ nonclassical monocyte subsets (representative flow cytometry plots shown in Fig. 2A). In line with recent observations from Thaler et al.(17), 4 h after endotoxin administration, we observed a decrease in the absolute numbers of intermediate and nonclassical subsets, while no significant change for the classical monocyte subset was observed (Fig. 2B). Two days after endotoxin administration, the number of classical monocytes showed a decline, although this did not reach statistical significance, while there was a significant increase in the number of intermediate monocytes compared with baseline, which persisted up to 20 days post administration (Fig. 2B).
In sepsis patients, monocyte subset analysis revealed a significant reduction in absolute numbers of all three monocyte subsets compared with HCs, a pattern which persisted on days 3 to 5 and 7 to 9 (Fig. 2C).
Subsequently, we assessed the activation status of each monocyte subset in both cohorts. It is well known that a decrease in the HLA-DR expression level on monocytes indicates decreased activation and impaired function (18). In the human endotoxemia model, we observed no change in the HLA-DR expression level on classical monocytes at any time point, while the HLA-DR expression levels on the intermediate subset were significantly reduced at 4 h as well as 2 and 20 days post administration (Fig. 2D). Additionally, the nonclassical subset exhibited decreased HLA-DR expression 4 h and 2 days postendotoxin administration, returning to baseline levels at 20 days (Fig. 2D). In contrast to the results obtained in endotoxemic subjects, HLA-DR expression levels on classical monocytes were significantly lower on day 3 to 5 in sepsis patients compared with HCs. There were no changes in HLA-DR expression levels at any time point for the intermediate and nonclassical monocytes (Fig. 2E).
These results show that experimental human endotoxemia results in increased numbers of intermediate monocytes 20 days post administration, and decreased expression of HLA-DR on this subset. In contrast, during sepsis, the numbers of the 3 monocyte subsets in the circulation are consistently and markedly reduced.
CD8+ effector T cells are increased in both experimental human endotoxemia at 20 days post administration and in sepsis
In addition to the innate immune cell composition, we also wished to gain more insight into the adaptive immune cell composition in the experimental human endotoxemia model and compared it to what is observed in septic shock. Human endotoxemia has been shown to decrease CD4+ and CD8+ T-cell numbers shortly after endotoxin administration (19), mirroring the changes observed in sepsis patients (20). Here, we examined both the short and long-term dynamics of circulating naïve (CD45RA+ CD27+), central memory (CM) (CD45RA− CD27+), effector memory (EM) (CD45RA− CD27−), and effector (CD45RA+ CD27−) CD4+ and CD8+ T cells (gating strategy shown in Fig. 3A and B). Four hours after endotoxin administration, decreases in absolute numbers of naïve, CM, EM, and effector CD4+ and CD8+ T cells was apparent (Fig. 3C and E). These early changes in CD4+ T-cell subsets normalized after 2 days and remained at pre-endotoxemia levels until 20 days after endotoxin administration (Fig. 3C). In sepsis patients, decreased CD4+ CM cell numbers were found at all time points compared with HCs, while nonsignificant changes in all the other CD4+ subsets were observed (Fig. 3D).
Notably, an increase in absolute counts of effector CD8+ T cells at 2 days postendotoxin administration was observed, which persisted up to 20 days after administration (Fig. 3E). In sepsis patients, we also found increased numbers of effector CD8+ T cells, which was accompanied by a decrease of CM cells during the entire observation period (Fig. 3F).
These data show that there is a similar effect on effector CD8+ T cells in both the endotoxemia model and in sepsis.
CD8+ T cells show an increased proliferation status at 20 days postendotoxin administration and during sepsis
In previous studies in sepsis patients, an increase in proliferating CD4+ and CD8+ T cells has been demonstrated (21). Here, we examined the division status of CD4+ and CD8+ T cells in subjects challenged with endotoxin and sepsis by analyzing the expression of the intracellular division marker Ki67 using flow cytometry (representative flow cytometry plots are shown in Fig. 4A and B). Within the CD4+ T-cell compartment, we identified an increase in the percentage of Ki67+ cells at 2 days postendotoxin administration, which returned to baseline levels at 20 days (Fig. 4C). The percentage of Ki67+ CD4+ T cells in sepsis patients only showed an increase at day 7 to 9 (Fig. 4D). In addition, the percentage of Ki67+ CD8+ T cells also increased 2 days after endotoxin challenge and remained significantly increased at 20 days (Fig. 4E). Likewise, in sepsis patients a significant increase in Ki67+ CD8+ T cells at all time points was observed (Fig. 4F).
These observations indicate that both during endotoxemia and sepsis increased proliferation is taking place in the T-cell compartment, which is most evident within the CD8+ T-cell compartment.
Cytokine expression is enhanced in CD8+ T cells in both the experimental human endotoxemia and sepsis
It has been shown that CD4+ and CD8+ effector T cells exhibit increased cytokine expression and cytotoxic effector function after antigen encounter (22). Since we observed an increase in effector CD8+ T cells 20 days after endotoxin administration, we next examined the cytokine-expressing capacity of CD4+ and CD8+ T cells in the human endotoxemia model and sepsis patients. Freshly isolated PBMCs were stimulated with PMA/ionomycin in the presence of brefeldin A, and subsequently the intracellular expression of TNF-α and IFN-γ was analyzed by flow cytometry (representative flow cytometry plots for CD4+ and CD8+ T cells are shown in Fig. 5A and B, respectively).
Within the CD4+ T-cell compartment, 4 h after endotoxin administration, a decreased percentage of TNF-α+, IFN-γ+, and TNF-α+/IFN-γ+ double positive cells compared to baseline was observed (Fig. 5C). Two days later, we found that the percentage of TNF-α+ cells had returned to baseline levels, whereas the percentage of IFN-γ+ and TNF-α+/IFN-γ+ double positive cells remained low, but recovered to baseline levels 20 days post administration (Fig. 5C). In sepsis patients, cytokine expression by CD4+ T cells remained comparable with that observed in HCs at all time points, except for a moderate increase of TNF-α+ cells on day 1 (Fig. 5D).
Regarding the CD8+ T-cell compartment, the percentages of TNF-α+, IFN-γ+, and TNF-α+/IFN-γ+ double positive cells were reduced at 4 h postendotoxin administration and returned to baseline 2 days afterwards (Fig. 5E). Furthermore, at 20 days post administration the percentage of TNF-α+ CD8+ T cells was significantly increased (Fig. 5E). In sepsis patients, significantly higher levels of TNF-α+, but also IFN-γ+ and TNF-α+/IFN-γ+ double positive CD8+ T cells were observed at all time points compared with HCs (Fig. 5F).
These findings reveal that while CD4+ T cell TNF-α expression and effector phenotype tend to return to basal levels 2 days after endotoxin challenge, the capacity to produce IFN-γ was still diminished and only recovered at the 20 days time point. On the other hand, our data signify that elevated levels of TNF-α expressing CD8+ T cells are present at 20 days after human endotoxemia and in sepsis patients throughout the time points measured.
Presence of low-density granulocytes in the PBMC fraction of subjects challenged with endotoxin and in sepsis patients
The increased cytokine expression potential that we demonstrated in CD8+ T cells of sepsis patients was unexpected, as it was previously reported that production of cytokines such as IFN-γ is reduced in sepsis patients (23). However, in these previous experiments, cytokine production was measured in cell culture supernatants after stimulation of isolated PBMCs, while we analyzed cytokine expression focused on CD4+ and CD8+ T cells after PBMC stimulation using flow cytometry. We wished to clarify whether these methodological differences could explain the discrepancy between the previous observations and our work. First, we noticed that cytokine expression of the entire CD45+ leukocyte population, instead of CD4+ or CD8+ T cells, resulted in identical results for the experimental human endotoxemia model (see figure S4, Supplemental Digital Content S4 A and C, http://links.lww.com/SHK/A791). However, for sepsis patients, this strategy results in a reduced percentage of TNF-α+, IFN-γ+, and TNF-α+/IFN-γ+ double positive CD4+ (all time points) and CD8+ (day 1) compared with HCs (see figure S4, Supplemental Digital Content S4 B and D, http://links.lww.com/SHK/A791), which is in line with the previous observations made by others (23).
The difference between cytokine-expressing cells following gating on either CD4+ and CD8+ T cells or CD45+ leukocytes is likely the consequence of an altered cellular composition of the PBMC fraction in sepsis as compared to healthy individuals. Therefore, we performed a more detailed analysis of the cellular composition of the PBMC fraction in the endotoxemia and sepsis cohorts (representative flow cytometry plots in Fig. 6A). In the human endotoxemia model, we found a significant increase in the percentage of low-density granulocytes (LDGs) and a significant decline in CD4+ and CD8+ T-cell percentages 4 h post administration (Fig. 6B). Two days after endotoxin administration, the percentage of LDGs returned to baseline levels and remained so until 20 days postendotoxin administration.
As demonstrated before (24), there was a significant increase of LDGs within the CD45+ cell fraction of PBMCs obtained from sepsis patients compared with HCs at all 3 time points, which was accompanied by a substantial decrease in CD4+ and CD8+ T cells (Fig. 6C). This increase in LDGs and concomitant reduction in T-cell percentages explain the previously shown reduction of cytokine production in PBMC cultures of septic patients (23).
Sepsis is the first cause of death in the ICUs worldwide (1), and more than 100 clinical trials have failed to yield therapies to improve patient survival (3). Therefore, increased understanding of, and novel therapeutic strategies for this disease are highly warranted. The heterogeneity of the sepsis patient population remains one of the biggest challenges to overcome. The use of in vivo human sepsis models might have the potential to increase our knowledge of sepsis pathophysiology and could be used to study therapeutic interventions. The experimental human endotoxemia model, in which healthy volunteers are i.v. challenged with endotoxin, has been shown to be an appropriate model to investigate the mechanisms underlying systemic inflammation (25) and has increased our understanding of the early (1–24 h after endotoxin administration) changes in the innate and adaptive immune response upon challenge with a bacterial compound.
In the present work, we investigated the long-term effects (here defined as 20 days after LPS administration) of experimental human endotoxemia on several aspects of the innate and adaptive immune system, and compared these to the effects observed in patients during the first 7 to 9 days after onset of septic shock. Naturally, comparing two different time frames is challenging and may represent a limitation of this work. However, because sepsis and human endotoxemia are also different entities with different kinetics, it is difficult to define optimal timeframes for comparison. We would like to stress that we do not consider the 9-day measurement period in sepsis patients “long-term.” For instance, in a recent review, Delano and Ward (26) explain in detail that long-term effects are observed 60 to 90 days, or even 3 years after sepsis. We also acknowledge that the considerable difference in age (median age of 20 vs. 62.5) between the endotoxemia group and the sepsis patients (and their age-matched HCs) represents a limitation of the current work, as it is well known that the numbers of innate and adaptive immune cells have a positive and negative correlation with age, respectively, in healthy subjects (16).
Our data reveal an increase in the number of circulating monocytes 20 days after endotoxin administration, which could mainly be attributed to an increase in numbers of intermediate monocytes. Thaler et al.(17) demonstrated an increase in the number of intermediate and nonclassical monocytes 24 h after endotoxin administration. Recently, these results were confirmed, and the authors also demonstrated that this subset remained increased at 72 h post endotoxin and returned to baseline levels at 7 days (27). In our sepsis cohort, we observed lower numbers of the monocyte subsets compared with HCs up to 9 days into the disease. Intermediate monocytes have been identified as a pro-inflammatory monocyte subset of which the abundance is increased in several inflammatory diseases (28). The increase in intermediate monocytes can be explained by an emergency release from the bone marrow of classical monocytes, which are the precursors of intermediate monocytes in the circulation (27). The function of intermediate monocytes in sepsis or other inflammatory diseases has not yet been explored. Our data indicate that experimental human endotoxemia may serve as an important tool to further elucidate this.
HLA-DR expressed by monocytes and other antigen-presenting cells is essential to initiate the adaptive immune response (29). Here we determined the level of expression of this molecule on each monocyte subset in our 2 study cohorts. We show for the first time that nonclassical and intermediate monocytes display decreased HLA-DR expression 4 h, 2 days (both subsets), and 20 days (only the intermediate subset) post endotoxin administration. In our sepsis cohort, we found that classical monocytes exhibited decreased HLA-DR expression on days 3 to 5, whereas expression remained stable on intermediate and nonclassical monocytes. The fact that classical monocytes retain their HLA-DR expression during experimental endotoxemia when using a bolus of 1 ng/kg endotoxin while expression is significantly decreased in sepsis patients is likely due to the obvious differences in severity between the model and the actual disease.
It has been established that classical and intermediate monocytes exhibit potent antigen-presenting and T-cell stimulating properties, with the highest HLA-DR expression of all subsets found on intermediate monocytes (28). It is reasonable to assume that lower HLA-DR expression is reflective of reduced antigen presentation, and thus, poor adaptive immune activation. In this regard, it has been shown both in vitro and in vivo in the experimental endotoxemia model as well as in sepsis patients that monocytes develop a phenomenon called “endotoxin tolerance” which is related to functional disruption, exemplified by impaired cytokine production as well as poor antigen presentation and T-cell activation (30, 31). Moreover, several studies demonstrate that persistently lower HLA-DR expression on monocytes during sepsis correlates with occurrence of secondary infections and poor outcome (32). Therefore, the endotoxemia model is of interest to further our understanding of the “endotoxin-tolerant” phenotype in monocytes and its consequences on the immune response.
Another important long-term effect that we observed in the endotoxemia model was related to adaptive immunity, specifically the T-cell compartment. In line with other reports (19), we showed a pronounced decline in T-cell numbers 4 h after endotoxin administration. The reason for the disappearance of T cells from the circulation during endotoxemia has not been well established. In sepsis, however, the same phenomenon was found to be due to massive apoptosis of T cells in the circulation and tissues of sepsis patients (21). It is well known that under lymphopenic conditions, homeostatic proliferation of naïve and memory T cells replenish the T-cell pool in the circulation. The key feature of homeostatic proliferation is increased proliferation of remaining T cells in the circulation that acquire an effector memory or effector phenotype and reveal an oligoclonal T cell receptor (TCR) repertoire. This phenomenon has been observed both in lympho-depleted mice and in patients under alemtuzumab-induced lymphopenia (33, 34). In the endotoxemia model, we show for the first time that after an acute lymphopenic event as observed 4 h post endotoxin administration, there is an increase in the number of effector CD8+ T cells and an increase in the percentage of proliferating (i.e., Ki67+) CD8+ T cells in the circulation. Interestingly, in the sepsis cohort, we also identified an increase in the percentage of Ki67+ CD8+ T cells. These findings suggest that homeostatic proliferation takes place in both the endotoxemia model and in sepsis. Consistent with this hypothesis, it was shown that homeostatic proliferation replenishes CD8+ T cells in the circulation in a mouse model of sepsis (35). Effector and effector-memory CD8+ T cells that are present following homeostatic proliferation have the ability to express TNF-α and IFN-γ (34). This feature was also observed 20 days after endotoxin administration and throughout days 1 to 9 in our sepsis cohort. It is likely that these cells produce cytokines upon stimulation with their cognate antigen; however, it has to be kept in mind that reduced TCR clonality after lymphopenia in sepsis mouse models and in alemtuzumab-treated patients could impair this functionality (34, 35). Therefore, TCR oligoclonality could contribute to the reduced T-cell activation and function observed in sepsis.
We did not find clear evidence of homeostatic proliferation within the CD4+ compartment, neither in the endotoxemia model nor in sepsis patients. Replenishment of CD4+ T cells by homeostatic proliferation in lymphopenic humans has been demonstrated to take place at a significantly lower rate than that of CD8+ T cells (36), which could explain why features of homeostatic proliferation in CD4+ T cells were not found. Furthermore, it is well known that reconstitution of CD4+ T cells through thymopoiesis occurs more prominently in young subjects (37). Therefore, it is likely that thymic replenishment of CD4+ T cells was dominant in the young endotoxemia-challenged subjects.
In conclusion, our data show that the experimental human endotoxemia model can contribute to improved understanding of the longer-term distribution and function of innate and adaptive immune cells during systemic inflammation and sepsis. The long-term effects of endotoxin administration on monocyte subset distribution and HLA-DR expression, which are not similar to the changes observed in sepsis, could nevertheless, be used to study the function of monocytes in vivo under other inflammatory conditions. Furthermore, our results on the long-term effects of endotoxemia on the CD8+ T-cell compartment suggest a role of homeostatic proliferation in the replenishment of this cell subtype to the circulation, an effect that was also observed in patients with septic shock. As such, the endotoxemia model may be instrumental to study the consequences of acute lymphopenia in sepsis, and the mechanisms employed by the immune system to surpass it. Taken together, the experimental human endotoxemia model has an important place in the arsenal of models for sepsis, which otherwise consists of animal models, such as administration of live bacteria or cecal ligation and puncture. Although this human model is for obvious reasons less severe than its animal counterparts, it is instrumental for studying the (patho)physiology of systemic inflammation in humans in vivo, and for establishing proof-of-principle that immunomodulatory interventions which have shown great promise in animal models are effective in the human setting.
The authors would like to thank Linda Hamers and the IC research nurses for sample collection and Jelle Gerretsen for plasma cytokine determinations.
1. Fleischmann C, Scherag A, Adhikari NK, Hartog CS, Tsaganos T, Schlattmann P, Angus DC, Reinhart K. International Forum of Acute Care Trialists. Assessment of global incidence and mortality of hospital-treated sepsis current estimates and limitations. Am J Respir Crit Care Med
2. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche J-D, Coopersmith CM, et al. The third international consensus definitions for sepsis and septic shock
3. Hutchins NA, Unsinger J, Hotchkiss RS, Ayala A. The new normal: immunomodulatory agents against sepsis immune suppression. Trends Mol Med
4. Bahador M, Cross AS. From therapy to experimental model: a hundred years of endotoxin administration to human subjects. J Endotoxin Res
5. Mechanic RC, Frei E, Landy M, Smith WW. Quantitative studies of human leukocytic and febrile response to single and repeated doses of purified bacterial endotoxin. J Clin Invest
6. Suffredini AF, Noveck RJ. Human endotoxin administration as an experimental model in drug development. Clin Pharmacol Ther
7. Copeland S, Warren HS, Lowry SF, Calvano SE, Remick D. Inflammation and the Host Response to Injury Investigators. Acute inflammatory response to endotoxin in mice and humans. Clin Diag Lab Immunol
8. Kamp VM, Leentjens J, Pillay J, Langereis JD, de Kleijn S, Kox M, Netea MG, Pickkers P, Koenderman L. Modulation of granulocyte kinetics by GM-CSF/IFN-γ in a human LPS rechallenge model. J Leukoc Biol
9. Tak T, Van Groenendael R, Pickkers P, Koenderman L. Monocyte subsets
are differentially lost from the circulation during acute inflammation induced by human experimental endotoxemia
. J Innate Immun
10. Kiers D, Koch RM, Hamers L, Gerretsen J, Thijs EJM, Van Ede L, Riksen NP, Kox M, Pickkers P. Characterization of a model of systemic inflammation
in humans in vivo elicited by continuous infusion of endotoxin. Sci Rep
11. Hotchkiss RS, Monneret G, Payen D. Immunosuppression in sepsis: a novel understanding of the disorder and a new therapeutic approach. Lancet Infect Dis
12. Kox M, de Kleijn S, Pompe JC, Ramakers BP, Netea MG, van der Hoeven JG, Hoedemaekers CW, Pickkers P. Differential ex vivo and in vivo endotoxin tolerance kinetics following human endotoxemia. Crit Care Med
13. Hamers LA, Kox M, Arts RJW, Blok B, Leentjens J, Netea MG, Pickkers P. Gamma-irradiated bacille Calmette-Guerin vaccination does not modulate the innate immune response during experimental human endotoxemia in adult males. J Immunol Res
14. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G. SCCM/ESICM/ACCP/ATS/SIS. 2001 SCCM/ESICM/ACCP/ATS/SIS international sepsis definitions conference. Crit Care Med
15. Nguyen HB, Rivers EP, Abrahamian FM, Moran GJ, Abraham E, Trzeciak S, Huang DT, Osborn T, Stevens D, Talan DA. Severe sepsis and septic shock
: review of the literature and emergency department management guidelines. Ann Emerg Med
16. Aguirre-Gamboa R, Joosten I, Urbano PC, van der Molen RG, van Rijssen E, van Cranenbroek B, Oosting M, Smeekens S, Jaeger M, Zorro M, et al. Differential effects of environmental and genetic factors on t and B cell immune traits. Cell Rep
17. Thaler B, Hohensinner PJ, Krychtiuk KA, Matzneller P, Koller L, Brekalo M, Maurer G, Huber K, Zeitlinger M, Jilma B, et al. Differential in vivo activation of monocyte subsets
during low-grade inflammation through experimental endotoxemia
in humans. Sci Rep
18. Muller Kobold AC, Tulleken JE, Zijlstra JG, Sluiter W, Hermans J, Kallenberg CG, Tervaert JW. Leukocyte activation in sepsis; correlations with disease state and mortality. Intensive Care Med
19. Krabbe KS, Bruunsgaard H, Qvist J, Fonsmark L, Møller K, Hansen CM, Skinhøj P, Pedersen BK. Activated T lymphocytes disappear from circulation during endotoxemia in humans. Clin Diagn Lab Immunol
20. Monserrat J, de Pablo R, Reyes E, Díaz D, Barcenilla H, Zapata MR, De la Hera A, Prieto A, Álvarez-Mon M. Clinical relevance of the severe abnormalities of the T cell compartment in septic shock
patients. Crit Care
21. Roger PM, Hyvernat H, Ticchioni M, Kumar G, Dellamonica J, Bernardin G. The early phase of human sepsis is characterized by a combination of apoptosis and proliferation of T cells. J Crit Care
22. Mahnke YD, Brodie TM, Sallusto F, Roederer M, Lugli E. The who's who of T-cell differentiation: human memory T-cell subsets. Eur J Immunol
23. Boomer JS, Shuherk-Shaffer J, Hotchkiss RS, Green JM. A prospective analysis of lymphocyte phenotype and function over the course of acute sepsis. Crit Care
24. Guérin E, Orabona M, Raquil MA, Giraudeau B, Bellier R, Gibot S, Béné MC, Lacombe F, Droin N, Solary E, et al. Circulating immature granulocytes with T-cell killing functions predict sepsis deterioration. Crit Care Med
25. Andreasen AS, Krabbe KS, Krogh-Madsen R, Taudorf S, Pedersen BK, Møller K. Human endotoxemia as a model of systemic inflammation
. Curr Med Chem
26. Delano MJ, Ward PA. The immune system's role in sepsis progression, resolution, and long-term outcome. Immunol Rev
27. Patel AA, Zhang Y, Fullerton JN, Boelen L, Rongvaux A, Maini AA, Bigley V, Flavell RA, Gilroy DW, Asquith B, et al. The fate and lifespan of human monocyte subsets
in steady state and systemic inflammation
. J Exp Med
28. Wong KL, Yeap WH, Tai JJ, Ong SM, Dang TM, Wong SC. The three human monocyte subsets
: implications for health and disease. Immunol Res
29. Fooksman DR. Organizing MHC class II presentation. Front Immunol
30. Draisma A, Pickkers P, Bouw MP, van der Hoeven JG. Development of endotoxin tolerance in humans in vivo. Crit Care Med
31. Shalova IN, Lim JY, Chittezhath M, Zinkernagel AS, Beasley F, Hernandez-Jimenez E, Toledano V, Cubillos-Zapata C, Rapisarda A, Chen J, et al. Human monocytes undergo functional re-programming during sepsis mediated by hypoxia-inducible factor-1α. Immunity
32. Lukaszewicz A, Faivre V, Payen D. Is monocyte HLA-DR expression monitoring a useful tool to predict the risk of secondary infection? Minerva Anestesiol
33. Hamilton SE, Jameson SC. The nature of the lymphopenic environment dictates protective function of homeostatic-memory CD8+ T cells. Proc Natl Acad Sci U S A
34. Jones JL, Thompson SA, Loh P, Davies JL, Tuohy OC, Curry AJ, Azzopardi L, Hill-Cawthorne G, Fahey MT, Compston A, et al. Human autoimmunity after lymphocyte depletion is caused by homeostatic T-cell proliferation. Proc Natl Acad Sci U S A
35. Unsinger J, Kazama H, McDonough JS, Hotchkiss RS, Ferguson TA. Differential lymphopenia-induced homeostatic proliferation
for CD4+ and CD8+ T cells following septic injury. J Leukoc Biol
36. Li CR, Santoso S, Lo DD. Quantitative analysis of T cell homeostatic proliferation
. Cell Immunol
37. Hakim FT, Memon SA, Cepeda R, Jones EC, Chow CK, Kasten-Sportes C, Odom J, Vance BA, Christensen BL, Mackall CL, et al. Age-dependent incidence, time course, and consequences of thymic renewal in adults. J Clin Invest