Endotoxin Tolerance Induced by Different TLR Ligands

Endotoxin tolerance is a long-recognised property of macrophages that leads to an altered response to repeated or chronic exposure to Toll-like receptor (TLR) ligands. The physiological role of tolerance is to limit the potential damage to host tissue that may otherwise result from prolonged production of pro-inflammatory cytokines. Endotoxin tolerance is induced by all TLRs tested to date, however, tolerance induced by the TLR4 ligand lipopolysaccharide (LPS) is by far the best studied. LPS tolerance involves a global transcriptional shift from a pro-inflammatory response toward one characterised by the expression of anti-inflammatory and pro-resolution factors. Although largely reversible, LPS-tolerance leads to a hybrid macrophage activation state that is pro-inflammatory in nature but possesses distinct regulatory anti-inflammatory features. Remarkably, a comparative transcriptomic analysis of tolerance induced by different TLR ligands has not previously been reported. Here we describe the transcriptomic profiles of mouse macrophages tolerised with ligands for TLR2, TLR3, TLR4 and TLR 9. While we identified TLR-specific transcriptional profiles in macrophages tolerised with each ligand, tolerance induced by TLR4 was the most comprehensive state, such that each gene tolerised by any of the TLRs tested was also found to be tolerised by TLR4. Pro-inflammatory cytokines are not universally supressed in all tolerant cells but distinct patterns of cytokine expression distinguished TLR-specific tolerance. Analysis of gene regulatory regions revealed specific DNA sequence motifs associated with distinct states of TLR tolerance, implicating previously identified as well as novel transcriptional regulators of tolerance in macrophages. These data provide a basis for the future exploitation of TLR-specific tolerant states to achieve therapeutic re-programming of the innate immune response.


42
Innate immunity is the first line of host defence against infection and is critical for the 43 development of adaptive immunity. Toll-like receptors (TLRs) are key sensors in the innate 44 immune system and recognise conserved structures of microbial-derived molecules or pathogen 45 associated molecular patterns (PAMPs). To date, 13 TLRs have been identified in mammals. 46 These form homo-or heterodimers to recognise a range of PAMPs that spans microbial 47 diversity, and regulation of this TLR repertoire fundamentally alters the tissue response to 48 infection (reviewed in (Huang and Wells, 2014;Leifer and Medvedev, 2016)). TLR activation 49 induces the expression of hundreds of genes that encode inflammatory cytokines, type I 50 interferons, antimicrobial proteins, and regulators of metabolism and regeneration; these 51 molecules in turn mediate inflammation, anti-microbial immunity and tissue regeneration. The negative regulation of TLR-signalling events is critical to ensure that prolonged or repeated 64 exposure to TLR ligands does not lead to uncontrolled or inappropriate inflammation and 65 consequent damage to host tissue. The most important mechanism for controlling TLR activation 66 is a form of tolerance to repeated exposure to a TLR ligand. This has been best described for 67 lipopolysaccharide (LPS) activation of TLR4 and is otherwise known as endotoxin-tolerance 68 (Medvedev, Kopydlowski and Vogel, 2000;Collins and Carmody, 2015). TLR-tolerance can be 69 described as a state of altered responsiveness of cells to the repeated or chronic activation of 70 TLRs, and includes the phenomena of cross-tolerance, where pre-exposure to one TLR-ligand 71 will reduce inflammatory responses to another (Dobrovolskaia et al., 2003;Julian et al., 2015). 72 While convergent signalling via NF-κB is essential for acquisition of LPS-tolerance, it is not 73 known how generalizable this may be to other TLR-ligands (Dobrovolskaia et al., 2003; 74 Carmody et al., 2007). Chromatin changes at tolerised genes allow for persistence of altered 75 responsiveness to re-stimulation of cells, but these changes are reversible over time, or in 76 response to competing signals (Foster,

107
Murine bone marrow isolation 108 Bone marrow was isolated from 6-8 weeks old female C57BL/6 for generation of primary bone Glutamine, 1% non-essential amino acids), and centrifuged at 4°C at 300 x g for 5 minutes. Bone   1μM CpG (1826 sequence, Eurofins). After 24 hours, the media was removed and the cells were 137 washed twice with sterile PBS. The cells were allowed to rest in fresh culture media for 1 hour 138 before a second stimulation for 4 hours with the same TLR ligand (Figure 1 (A)).

140
Immunoblotting 141 Whole-cell proteins were extracted using RIPA lysis buffer supplemented with protease   were calculated using the ΔΔCT method.

161
Microarray data was processed as follows: Data was normalised using Limma (3.26.9), including 162 RMA background correction quantile normalisation (Ritchie et al., 2015). Analysis workflow is 163 described in Figure 1    Tolerised genes were defined as those significantly upregulated at 4 hours acute stimulation, but 317 demonstrating at least 1.5-fold lower activation on re-stimulation. A total of 1644 genes matched 318 this pattern in one or more condition (Supplementary table S1 390 ligands.

392
By definition acute induction was a pre-requisite of tolerance, however not all acutely-induced 393 genes were tolerised. Arguably, tolerance represents one form of innate 'priming' or 'training', 394 where pre-exposure to a PAMP alters macrophage responses to re-exposure. We identified 174 395 acutely induced genes that demonstrated further induction on re-stimulation with one or more 396 TLR ligand (Figure 6, Supplementary table 1). Over a third of these genes were predicted to be

424
The set of 70 genes acutely induced in all conditions were overwhelmingly subjected to a 425 tolerance pattern, with a small number exhibiting ligand-specific super-induction (Figure 7 (A)). 426 This may indicate that coordinate mechanisms determine whether the expression of an 427 inflammatory mediator is tolerised or trained. Indeed, the core members of NF-κB activation 428 were themselves subjected to altered expression on re-stimulation with a TLR-ligand (Figure 7   429 (B)). Increased levels of the NF-κB p50 subunit and the IĸB protein BCL-3 were induced by all 430 TLR ligands tested (Figure 7 (C)) underlying the previously established role for the p50:BCL-3 431 transcriptional repressor complex in limiting TLR responses (Carmody et al., 2007). NF-κB 432 target genes were highly represented in BMDM genes with a 'memory' status of the original 433 stimulation, with the vast majority of these categorised in the tolerised group (Figure 7 (D)).  Historically, the study of endotoxin tolerance was performed using lipopolysaccharides. Early 458 research on endotoxin tolerance in vivo relied upon the fever response as a readout for 459 responsiveness to endotoxin and led to the concept that tolerance was a hypo-responsive state 460 due to a desensitisation to endotoxin. However, as the factors that mediate the innate 461 inflammatory response were identified it became apparent that endotoxin tolerance is a state of 462 altered responsiveness to stimulation rather than simply hypo-responsiveness. Initial In summary, this study defines the transcriptional responses of macrophages tolerised with 535 ligands for TLR2, TLR3, TLR4 and TLR9. Our data supports the concept that TLR tolerance 536 promotes a shift away from a pro-inflammatory transcriptional response towards a response that 537 is pro-resolution and anti-inflammatory in nature. The repression of transcription is generally 538 associated with NF-κB target genes while genes with IRF motifs are more likely to be super-539 induced in tolerant cells. However, this study also reveals the differential repression of cytokines 540 and chemokines in macrophages tolerised by specific TLR ligands. These patterns of expression 541 may have functional relevance to stimulus specific inflammatory responses and may also be 542 relevant to the study of innate immune training.

544
Accessions: 545 The data is available from (GSE81291) and Stemformatics  828  829  830  831  832  833  834  835  836  837  838  839  840  841  842  843  844  845  846  847  848  Table S2: Top 5 GO terms (biological process) enriched for each set of TLR-ligand activated 876 genes. Enrichment for each term is shown by the proportion of the GO list overlapping the TLR-877 induced gene lists (5), the number of terms in the TLR gene list /the total number of terms in the 878 GO category. Significance of enrichment indicated by an adjusted p-value. Full list available in 879   Table S1.