The Distinctive Activation of Toll-Like Receptor 4 in Human Samples with Sepsis

Clinical success of Toll-Like receptor-4 (TLR-4) antagonists in sepsis therapy has thus far been lacking. As inhibition of a receptor can only be useful if the receptor is active, stratification of patients with active TLR-4 would be desirable. Our aim was to establish an assay to quantify phosphorylated TLR-4 using the proximity ligation assay (PLA). HEK293 TLR4/MD2/CD14 as well as THP-1 cells were stimulated with LPS and the activation of TLR-4 was measured using the PLA. Furthermore, peripheral blood mononuclear cells (PBMCs) from 25 sepsis patients were used to show the feasibility of this assay in clinical material. Activation of TLR-4 in these samples was compared to the PBMCs of 11 healthy individuals. We could show a transient activation of TLR-4 in both cell lines. Five min after the LPS stimulation, the signal increased 6.7-fold in the HEK293 cells and 4.3-fold in the THP-1 cells. The assay also worked well in the PBMCs of septic patients. Phosphorylation of TLR-4 at study inclusion was 2.9 times higher in septic patients compared to healthy volunteers. To conclude, we established a diagnostic assay that is able to quantify the phosphorylation of TLR-4 in cell culture and in clinical samples of sepsis patients. This makes large-scale stratification of sepsis patients for their TLR-4 activation status possible.


Introduction
Toll-Like receptors (TLRs) are gatekeepers of the immune system. Currently, there are 10 known human variants (TLR1-TLR10) [1]. TLRs are pattern recognition receptors (PRRs) that bind pathogen-associated molecular patterns (PAMPs) and danger-associated molecular patterns (DAMPs) [1]. TLR4 is responsible for recognizing lipopolysaccharide (LPS), a compound of the outer membrane of gram-negative bacteria [2]. On a molecular level, LPS is bound by LPS-binding protein (LBP) and transported to a complex of TLR4 and myeloid differentiation factor 2 (MD-2) with the participation of cluster of differentiation 14 (CD14) [3,4]. This leads to the secretion of proinflammatory cytokines, reactive oxygen species, antimicrobial peptides, chemokines and acute-phase proteins [4,5]. Upon proper binding of LPS to TLR4, the receptor is phosphorylated and further activates a signaling network consisting of myeloid differentiation primary response 88 (MyD88), various interleukin-1 receptor-associated kinases (IRAKs) and subsequently nuclear factor kappa B (NFκB) dimers [3,4]. Therefore, TLR4 is an attractive therapeutic target for immunological disorders such as sepsis. Sepsis is defined as an acute organ dysfunction caused by the dysregulated immune response to a microbial agent [6]. Currently, sepsis is one of the leading causes of death in industrialized nations [7] with a yearly death toll in the millions [8]. As therapeutic options to specifically treat the dysregulated immune response in sepsis have been largely unsuccessful thus far, the TLR4 inhibitors were viewed as a good approach to reduce the mortality in sepsis. Interestingly, this approach worked well in mouse studies [5]. However, no success could be reported when the TLR4 antagonists were administered in human patients [9]. It is noteworthy that the activation state of the TLR4 receptor was not assessed in these clinical trials prior to the recruitment of septic patients, which is understandable as there is currently no way to measure TLR4 activation in clinical material of sepsis patients. The methods that usually work well in cell culture such as immuneprecipitation are too material intensive to be feasible. Furthermore, measuring downstream activation such as phosphorylation of ERK or expression of IL-6 is highly unspecific since both molecules are induced by a range of different receptors [10][11][12][13]. However, since the activation state of the specific receptor is an all-important pre-requisite for an antagonist to work, we wanted to establish an assay based on the proximity ligation assay (PLA, [14]) that can quantify the phosphorylation (i.e., activation) of TLR4 in clinical material.

Patient Recruitment
The samples used in this study were part of the SepsisDataNet.NRW cohort [15]. Patients were prospectively recruited at the Universitätsklinikum Knappschaftskrankenhaus Bochum (DRKS00018871; Ethics-Registration No. 18-6606-BR) and the University Hospital Münster (Ethics-Registration No. 2017-513-b-S) according to the Sepsis-3 definition [6]. In the SepsisDataNet.NRW study biomaterials, serum and peripheral blood mononuclear cells (PBMCs) were collected at day one, day four and day eight after study inclusion. In addition, blood was drawn from self-declared healthy subjects and the PBMCs were isolated (Ethics-Registration No. 21-7278, Ethics committee Ruhr University Bochum).

PBMC Isolation
For this study, 20 septic patients with either a hospital-acquired pneumonia or a peritonitis were identified in the SepsisDataNet.NRW cohort. In addition, five septic patients with severe COVID-19 were selected from the cohort. For isolation of the PBMCs, blood samples from sepsis patients and healthy subjects were analyzed using a Ficoll density gradient centrifugation (GE Healthcare Europe, Freiburg, Germany). The phase containing the PBMCs was collected and washed with PBS. After subsequent lysis of residual erythrocytes, the cells of healthy subjects were seeded in a 12-well plate in an appropriate cell number in RPMI medium and stimulated with LPS as described above. The PBMCs of septic and COVID-19 patients as well as stimulated cells from healthy subjects were directly spun to a microscope slide using a cytospin (Cellspin II, Tharmac, Wiesbaden, Germany) according to the manufacturer's instructions. The cells were then fixed using a 4% formaldehyde solution, dried and frozen at −80 • C until use.

Measurement of Cytokines in Serum
The serum samples were used to quantify the concentration of thirteen cytokines at the time of recruitment. The LegendPlex Human Inflammation Panel 1 (BioLegend, San Diego, CA, USA) was used according to manufacturer's instructions. Briefly, the serum samples were incubated with LegendPlex beads for antigen capture, washed and incubated with detection antibodies. Subsequently, the fluorescence was quantified in a flow cytometer (Canto II, BD Biosciences, San Jose, CA, USA). When the recorded concentration of a cytokine was below the lower limit of detection (LOD), the value was set to 0 pg/mL; additionally, if a value was recorded as higher than the upper LOD, it was set to the upper LOD.

Proximity Ligation Assay
The PLA was performed as described previously [16]. In short, the cells were rehydrated, permeabilized with 0.1% Triton X (Carl Roth, Karlsruhe, Germany) and treated with 1% SDS (Carl Roth, Karlsruhe, Germany). After blocking of unspecific binding sides with Duolink Block (Sigma-Aldrich), primary antibodies were applied. For the cell culture, primary antibodies were diluted 1:200 TLR4 (sc-293072, Santa Cruz Biotechnology, Dallas, TX, USA) and 1:100 pan phospho-tyrosine antibody (p-Tyr-1000, #8954, Cell Signaling Technology, Denver, MA, USA) in antibody diluent (Sigma-Aldrich) and incubated at 4 • C over night. For the PBMCs, the antibodies were diluted 1:50 TLR4 and 1:50 p-Tyr-1000 in antibody diluent and incubated at 4 • C over night. After washing of the unbound primary antibody, the slides were incubated with proximity probes Duolink Mouse PLUS and Duolink Rabbit Minus (Sigma-Aldrich) according to the manufacturer's instructions. After ligation and amplification using a compaction oligonucleotide described before [16], the slides were imaged in a widefield microscope.

Image Analysis
Image analysis was performed with FIJI and CellProfiler (Stirling, Swain-Bowden et al. 2021). First, the maximal intensity projections were calculated using FIJI and the images compatible for CellProfiler analysis were created. For PBMCs, the DAPI channel was used to identify and mark intact nuclei. Images were analyzed in CellProfiler and subjected to a pipeline containing the following modules: identify primary objects (nuclei), identify secondary objects (cells), enhance or suppress features, identify primary objects (PLA signals) and relate objects (PLA signals to cells). For each condition, a minimum of 50 cells were analyzed with at least three images per condition.

Statistics
Statistical analysis was performed using SPSS (IBM, Armonk, NY, USA, Version 28). The PLA signals per cell and other continuous data were compared using a Kruskal-Wallis test for multiple comparisons and a Mann-Whitney U test for pair wise comparisons. Categorical parameters were analyzed using the Fisher Exact test. Correlations were calculated with the Pearson correlation analysis.  (Figure 1a). At the ground state, the HEK cells showed a mean PLA signal of 1.7 (+/−1.9) signals per cell, which increased after 5 min of 1 µg/mL LPS to 11.4 (+/−6.9) signals per cell (n = 3, each p < 0.001, Figure 1b). Activation of TLR4 could further be confirmed by an increase in the expression of interleukin 8 (IL-8) after the LPS stimulation to 34-fold (+/−4; p < 0.001, Figure 1c).

Validation of the Phospho-TLR4 PLA Assay on Endogenous Protein Expression
As the HEK cells ectopically express TLR4 and its relevant components, we tested the PLA assay on the THP-1 cells, which endogenously express TLR4. Upon stimulation with LPS, we could again show an increase in the phosphorylation level of TLR4 from 0.4 (+/− 0.3) signals per cell (unstimulated) to 1.7 (+/− 0.16) signals per cell (5 min after LPS stimulation, n = 3, each p < 0.001, Figure 2b). In the THP-1 cells, we could also confirm the activation of the TLR4 signaling cascade using qPCR by showing an increase in interleukin 6 (IL-6) by 4148-fold (+/− 2005, p < 0.001) after 24 h of incubation with 1 µg/mL LPS (Figure 2c

Translation of the Assay to Peripheral Blood Mononuclear Cells (PBMCs) of Healthy Individuals
The PBMCs from healthy individuals equally showed an increase in the PLA signals from 0.4 signals per cell (+/− 0.8, unstimulated) to a maximum of 1.6 signals per cell (+/− 0.1, n = 3, each p < 0.001, Supplementary Figure S1) after the LPS stimulation. The kinetics of the three individuals differed slightly with a maximum of TLR4 activation between 5 and 15 min after stimulation.

Cohort Description
In order to test the assay in the clinical material, we quantified the TLR4 activation in the PBMCs of 25 critically ill patients. The median age of the cohort was 63 years and 60% were male. The median sequential organ failure assessment (SOFA) score of the cohort at study inclusion was 9. Of the 25 critically ill patients, 20 were diagnosed with a polymicrobial sepsis and five patients suffered from a SARS-CoV-2 induced sepsis. At study inclusion, the median serum concentration was 156 pg/mL (IQR: 531 pg/mL) for IL-6 and 8 pg/mL (IQR: 12 pg/mL) for IL-10. The median C-reactive protein (CRP) was 18 mg/dL and the median procalcitonin (PCT) concentration was 12.2 ng/mL at study inclusion. A more in-depth characterization is shown in Table 1. As a control group, we measured the TLR4 phosphorylation in 11 self-declared healthy individuals. The median age of this cohort was 48 and 36% were male.

TLR4 Activation in Septic Patients
The mean phosphorylation of TLR4 at study inclusion was 2.0 signals per cell (+/− 3.3) in septic patients compared to 0.7 signals per cell (+/− 0.5) in healthy volunteers (p = 0.788, Figure 3a,b). Interestingly, with 4.2 signals per cell (+/− 0.9) the COVID-19 patients showed a significantly higher phosphorylation than both the bacterial septic (p = 0.017) and the control patients (p = 0.002, Figure 3c). In addition, while phosphorylation dropped in the septic patients over the next four days (day four: 1.1 signals per cell; +/− 1.1, Figure 3b) it remained high in the COVID-19 patients (3.9 signals per cell; +/− 1.2). Notably, the variability of the signals per cell was visibly higher in the septic group than in the control group. The coefficient of variation (CV) of the healthy subject group was 79%, which was less than half of the septic group (167%). Interestingly, we found a time-delayed correlation of the TLR4 phosphorylation to IL-6 serum concentrations in the septic patients similar to the cell culture experiments. TLR4 phosphorylation at study inclusion correlated well to IL-6 serum concentration at day four (r = 0.642 p = 0.003) and phosphorylation at day four correlated to IL-6 concentration at day eight (r = 0.827 p = 0.022). While we saw a small detrimental effect of TLR4 activation (cutoff = 2 signals per cell) on the 30-day survival in the entire cohort (about 10% lower survival in the activated group), it did not rise to the level of significance (p = 0.435).

Discussion
In this work, we could establish an in situ proximity ligation assay able to quantify TLR-4 activation in the cell culture and the clinical material of critically ill patients. TLR4 has long been proposed to be a promising target for immune-modulative sepsis therapy.
Evaluation of the activity state of this receptor prior to treatment was thus far not possible. This may be the reason for the failure of clinical trials with the TLR4 antagonists [9]. In our work, we introduced a PLA assay to measure and quantify the activation of TLR4 by examining its phosphorylation in cell culture models and clinical samples of sepsis patients. The combination of a specific receptor antibody with a generic phospho-tyrosine antibody in a PLA assay has been successfully employed before [17]. The transient activation pattern found in all cell lines that we evaluated with a maximum activation between 5 and 15 min is comparable to the results described in the literature [18]. This, as well as the qPCR results showing a stimulation of the immune network subsequent to TLR4 phosphorylation, strengthens our conclusion that the assay accurately captures the activation of TLR4. Interestingly, the LPS stimulation series with the PBMCs from healthy donors showed some inter-individual variability in the time it took to reach maximal activation of the PBMCs. Although we only evaluated three donors, these results already point towards the high heterogeneity we typically see with the immune system. This heterogeneity can also be seen in the cohort of critically ill patients as the CV was notably higher in the sepsis patients than in the control group. Interestingly, we could show that most polymicrobial sepsis patients did not show an elevated activation of TLR4. This helps to explain the failure of the TLR4 antagonists in clinical trials. We could, however, identify a small subgroup of patients with elevated activation of TLR4 for which anti-TLR4 treatment might be warranted. These patients showed no obvious association with the gram status of the underlying infection, as one might have suspected, and can thus far only be identified by measuring the activation of TLR4. However, there are multiple possible confounders. There are often different types of pathogens present in sepsis patients, especially in patients with impaired immunity. In addition, blood cultures do not always identify all bacteria in the sample. Therefore, we can imagine a bias where a gram-negative super-or co-infection was not detected in the clinic. At the same time, a long gram-negative infection might lead to the exhaustion of the signaling potential of the TLR4 and subsequently result in an impaired activation [18][19][20]. We also see no significant association with 30-day survival of the patients. However, as our study was not powered to study survival effects, further work will be needed to evaluate the effect that activation of TLR4 might have in septic patients. Furthermore, a much-needed subgroup analysis of the patients will have to be conducted in larger patient cohorts. Interestingly, a set of patients with severe COVID-19, originally included as a negative control, showed high-activation levels of TLR4 at both time points. It has been proposed recently that SARS-CoV-2 could directly activate TLR4 [21] and this explanation would fit well to our data. However, because critically ill COVID-19 patients often harbor bacterial superinfections [22], we cannot exclude this here. Again, larger studies specifically on COVID-19 are needed to shed more light on this intriguing finding.

Conclusions
To conclude, in this study we show the feasibility of a PLA assay to study the activation of TLR4 in cell culture and the PBMCs of sepsis patients. We propose that measuring the activation state of this receptor should be a pre-requisite for clinical trials evaluating TLR4 inhibitors. Larger studies are needed in order to determine the extent of elevated TLR4 activation in sepsis patients and the impact of this activation on the outcome.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cells11193020/s1, Figure S1: Activation of TLR4 in PBMCs of healthy individuals upon LPS stimulation. (a) Microscopic images of PBMC from one individual. Nuclei are depicted in blue. PLA signals are shown as red dots. (b) Quantification of TLR4 phosphorylation in all three individuals. Some heterogeneity in the kinetics can already be observed in this very small set of probands.