A Common Genetic Variant in TLR1 Enhances Human Neutrophil Priming and Impacts Length of Intensive Care Stay in Pediatric Sepsis

Abstract

Polymorphonuclear leukocytes (PMN) achieve an intermediate or primed state of activation following stimulation with certain agonists. Primed PMN have enhanced responsiveness to subsequent stimuli, which can be beneficial in eliminating microbes but may cause host tissue damage in certain disease contexts, including sepsis. As PMN priming by TLR4 agonists is well described, we hypothesized that ligation of TLR2/1 or TLR2/6 would prime PMN. Surprisingly, PMN from only a subset of donors were primed in response to the TLR2/1 agonist, Pam3CSK4, although PMN from all donors were primed by the TLR2/6 agonist, FSL-1. Priming responses included generation of intracellular and extracellular reactive oxygen species, MAPK phosphorylation, integrin activation, secondary granule exocytosis, and cytokine secretion. Genotyping studies revealed that PMN responsiveness to Pam3CSK4 was enhanced by a common single-nucleotide polymorphism (SNP) in TLR1 (rs5743618). Notably, PMN from donors with the SNP had higher surface levels of TLR1 and were demonstrated to have enhanced association of TLR1 with the endoplasmic reticulum chaperone gp96. We analyzed TLR1 genotypes in a pediatric sepsis database and found that patients with sepsis or septic shock who had a positive blood culture and were homozygous for the SNP associated with neutrophil priming had prolonged pediatric intensive care unit length of stay. We conclude that this TLR1 SNP leads to excessive PMN priming in response to cell stimulation. Based on our finding that septic children with this SNP had longer pediatric intensive care unit stays, we speculate that this SNP results in hyperinflammation in diseases such as sepsis.

Figure 1. In contrast to TLR2/6 stimulation, a TLR2/1 agonist primes ROS generation in PMN from only half of donors

ROS generation was measured by LUC-CL as RLUs in PMN primed with no agonist, 100 ng/ml FSL-1 (a TLR2/6 agonist), or 1 μg/ml Pam₃CSK₄ (a TLR2/1 agonist) for 30 min and then stimulated with fMLF. Representative tracings (mean ± SEM of three replicates) for (A) a donor whose PMN primed in response to Pam₃CSK₄ (TLR2/1 high primer) and (B) a donor whose PMN did not prime in response to Pam₃CSK₄ (TLR2/1 low primer). (C) Combined donor data (mean ± SEM) showing direct ROS production (priming stimulus only). n ≥ 16 per group. (D) Peak RLUs (mean + SEM) immediately following addition of fMLF after 30 min of priming in TLR2/1 high primers (black bars) and TLR2/1 low primers (gray bars). n ≥ 16 per group. ***p < 0.0001, Students t test.

Figure 2. Pam₃CSK₄ induces significantly greater direct and primed ROS production by TLR2/1 high primers than TLR2/1 low primers

(A) Direct and (B) primed generation of H₂O₂ measured by Amplex UltraRed in response to no agonist, 100 ng/ml FSL-1, 1 μg/ml Pam₃CSK₄, or 1 ng/ml TNF-α. Mean + SEM, n ≥ 3 per group. **p < 0.01. TLR2/1 low primers did not generate H₂O₂ above no agonist levels when stimulated with Pam₃CSK₄. (C) Intracellular ROS generation was determined by OxyBURST® green. Mean + SEM, n = 6 per group. *p < 0.05. (D) Primed generation of superoxide measured by SOD-inhibitable reduction of ferricytochrome c. Mean + SEM, n ≥ 2 per group. **p < 0.01. P values were calculated by Student’s t tests between TLR2/1 low primers and TLR2/1 high primers.

Figure 3. Pam₃CSK₄ initiates MAPK signaling in PMN from TLR2/1 high primers but not in TLR2/1 low primers

(A) p38 MAPK, (B) ERK1/2, and (C) JNK1 phosphorylation was quantified by immunoblotting and normalized to β-actin. Mean ± SEM, n ≥ 16 per group. *p < 0.05, ***p < 0.001. P values were calculated by 2-way ANOVA with Bonferroni’s multiple comparison test to compare between no agonist, 100 ng/ml FSL-1, and 1 μg/ml Pam₃CSK₄ at each time point. Asterisks indicate significance compared to no agonist.

Figure 4. Pam₃CSK₄ induces integrin activation and secondary granule exocytosis in TLR2/1 high primers but not TLR2/1 low primers

PMN surface expression of (A) CD11b, (B) the active conformation of CD11b, (C) CD66b (a secondary granule marker), and (D) CD63 (a primary granule marker) determined by flow cytometry of PMN from TLR2/1 low primers and TLR2/1 high primers treated with no agonist, 100 ng/ml FSL-1, or 1 μg/ml Pam₃CSK₄. There were no changes from baseline expression levels of CD11b or CD66b in TLR2/1 low primers stimulated with Pam₃CSK₄. Mean + SEM, n ≥ 5 per group. **p < 0.01, ***p < 0.001, Student’s t tests.

Figure 5. Upregulation of active CD11b and CD66b is downstream of p38 MAPK signaling

PMN were pretreated with a p38 MAPK inhibitor or inactive analog control prior to exposure to a TLR2 agonist as indicated previously. Surface (A) active CD11b and (B) CD66b expression were then quantified by flow cytometry in TLR2/1 low primers and TLR2/1 high primers. Data are shown as a percentage of surface upregulation inhibited by the p38 MAPK inhibitor. Mean + SEM, n ≥ 5 per group. **p < 0.01, Student’s t tests.

Figure 6. Select cytokines are upregulated in PMN and monocytes from TLR2/1 high primers only following Pam₃CSK₄ treatment

PMN secretion of (A) IL-8, (B) IL-1Ra, and (C) MCP-1 and monocyte secretion of (D) TNF-α, (E) IL-6, (F) IL-8, (G) IL-1Ra, and (H) MCP-1 following 24 h treatment with no agonist, 100 ng/ml FSL-1, or 1 μg/ml Pam₃CSK₄ was quantified by ELISA in TLR2/1 low primers and TLR2/1 high primers. Mean + SEM, n ≥ 9 per group. *p < 0.05, **p < 0.01, Student’s t tests.

Figure 7. Pam₃CSK₄ delays apoptosis in PMN from TLR2/1 high primers

Following 24 h treatment with no agonist, 100 ng/ml FSL-1, or 1 μg/ml Pam₃CSK₄, the percentages of (A) live and (B) apoptotic PMN were determined by Annexin V staining in TLR2/1 low primers and TLR2/1 high primers. Mean + SEM, n ≥ 4 per group. *p < 0.05, Student’s t tests.

Figure 8. TLR1 expression is significantly increased on the surface of PMN from TLR2/1 high primers compared to TLR2/1 low primers, although total cellular proteins levels are similar

(A) PMN surface expression of TLR1, TLR2, and TLR6 were quantified by flow cytometry and adjusted to an IgG control. Mean + SEM, n = 7 per group. **p < 0.01, Student’s t test. (B) Total TLR1 protein in whole PMN lysates was quantified by gel electrophoresis and immunoblotting in TLR2/1 low primers and TLR2/1 high primers. Mean + SEM, n ≥ 5 per group.

Figure 9. Increased association between TLR1 and gp96 in TLR2/1 high primers

(A) Immunoprecipitation of TLR1 followed by immunoblotting for gp96 demonstrates increased association in TLR2/1 high primers, n = 4. *p < 0.05. (B) Immunoprecipitation of gp96 followed by immunoblotting for TLR1 demonstrates increased association in TLR2/1 high primers, n = 3. ***p < 0.0001, t tests. (C) Immunoprecipitation of TLR1 followed by immunoblotting for TLR2 shows a trend towards increased association in the TLR2/1 high primers, n = 4. (D) Representative immunoblots.

Figure 10. Septic patients homozygous for the 1805T allele with positive blood cultures had significantly longer stays in the PICU

(A) Correlation between individual genotypes and PICU LOS in patients with positive bacterial blood cultures and sepsis. *p < 0.05 for GG vs. TT paired comparison, p = 0.053 for ANOVA. (B) Comparison between PICU LOS and GG vs. GT + TT combined, *p < 0.05. (C) Comparison between genotypes and PICU LOS in septic patients with Gram positive bacteremia. (D) Comparison between PICU LOS and genotype in septic patients with Gram negative bacteremia.