Toll-like receptor 4 promoter polymorphisms: common TLR4 variants may protect against severe urinary tract infection

Abstract

Background: Polymorphisms affecting Toll-like receptor (TLR) structure appear to be rare, as would be expected due to their essential coordinator role in innate immunity. Here, we assess variation in TLR4 expression, rather than structure, as a mechanism to diversify innate immune responses.

Methodology/principal findings: We sequenced the TLR4 promoter (4,3 kb) in Swedish blood donors. Since TLR4 plays a vital role in susceptibility to urinary tract infection (UTI), promoter sequences were obtained from children with mild or severe disease. We performed a case-control study of pediatric patients with asymptomatic bacteriuria (ABU) or those prone to recurrent acute pyelonephritis (APN). Promoter activity of the single SNPs or multiple allelic changes corresponding to the genotype patterns (GPs) was tested. We then conducted a replication study in an independent cohort of adult patients with a history of childhood APN. Last, in vivo effects of the different GPs were examined after therapeutic intravesical inoculation of 19 patients with Escherichia coli 83972. We identified in total eight TLR4 promoter sequence variants in the Swedish control population, forming 19 haplotypes and 29 genotype patterns, some with effects on promoter activity. Compared to symptomatic patients and healthy controls, ABU patients had fewer genotype patterns, and their promoter sequence variants reduced TLR4 expression in response to infection. The ABU associated GPs also reduced innate immune responses in patients who were subjected to therapeutic urinary E. coli tract inoculation.

Conclusions: The results suggest that genetic variation in the TLR4 promoter may be an essential, largely overlooked mechanism to influence TLR4 expression and UTI susceptibility.

Figure 1. Polymorphisms in the promoter region of the TLR4 gene.

A. Eight sequence variants were identified by direct sequencing of PCR products from the distal promoter region of the human TLR4 gene, starting 4.3 kb upstream of the ATG start codon. SNPs were located at bp -4038, -3612, -3002, -2604, -2570, -2081, -2026 and -1607. B-C. Conservation of the human TLR4 promoter compared to Bos Taurus and Sus scrofa (b) Rattus norvegicus and Mus musculus (c). Polymorphisms in the human consensus sequence were defined by comparison with the gene bank sequence (AF177765) and positions are indicated by ovals, numbered relative to the transcription start site in the human gene bank sequence. Filled ovals are located on conserved sequence blocks. D. Pair-wise linkage disequilibria (LD) were calculated, using VG2 software. LD was measured using r² where the color scheme presents the degree of LD (range 0.0 through 1.0) where the maximum LD value (r² = 1) corresponds to the upper boundary of the color spectrum. E. In silico predictions of transcription factor binding using TFSEARCH. Five of the eight SNPs were located in sequences with a high degree of homology to possible transcription factor binding motifs.

Figure 2. Haplotype analysis of TLR4 promoter polymorphisms in pediatric UTI patients and controls.

A. Circle diagram showing the haplotype frequency in the UTI groups and controls. The ABU groups differed significantly from the pediatric controls and between the ABU and the APN groups, but there was no difference between the two control groups or between the APN and controls. B. Haplotype frequency in patients and controls. H4 was more common in the primary and H5 in the secondary ABU patients. There was no difference haplotype frequency between APN and control groups.

Figure 3. TLR4 promoter genotypes.

A. TLR4 promoter GPs. Each column represents a genotype and each row a polymorphic site. Common allele homozygote (blue), heterozygote (red) and rare allele homozygote (yellow) are shown. Genotype patterns marked I-XX were found in ≥2 individuals, while GPs marked with – were found in one individual. B. GP frequencies in UTI prone patients and controls (Fisher's Exact Test). GPX was significantly more common in the primary ABU group and GPs V and VII were slightly more common in primary ABU, while GP XX was more common in secondary ABU. No difference in GP frequency was observed between the controls and the patients with APN but patients with APN group differed significantly from those with ABU.

Figure 4. TLR4 promoter genotype patterns (GPs) in UTI patient and control groups.

A. GP distribution in pediatric (study 1) and adult (study 2) patients and controls. The UTI groups differed significantly compared to controls, but there was no difference between the pediatric and adult controls. B. The combined frequency of the most common GPs in primary ABU (GPs V, VI, VII, X, XIII) was compared to pediatric secondary ABU (p. sec ABU), pediatric APN (p. APN) and adult secondary ABU (a. sec ABU) or adult APN (a. APN)) groups.

Figure 5. TLR4 promoter GPs common in ABU patients reduce transcription efficiency.

A-B. TLR4 promoter activity determined in a luciferase reporter assay. Human renal carcinoma cells (A498) were transfected with plasmids carrying TLR4 promoter sequences in frame of a luciferase reporter gene. The single SNP promoter constructs contained each of the detected SNPs. The multiple SNP constructs corresponding to GPs in the primary or secondary ABU groups. Luciferase levels were compared to GPIV, which was most common in the population and unrelated to UTI susceptibility. All transfected cells showed increased promoter activity compared to the background (Ba) control (means of three independent experiments, mean+/− SEMs). C-D. Change in promoter activity after infection with the uropathogenic E. coli strain CFT073 and schematic diagram showing possible transcription factor binding sites. The change in promoter activity after E. coli CFT073 infection is shown relative to the GPIV control. The histogram shows the fold difference in promoter activity post-infection for ABU-associated GPs and for the corresponding single SNPs. SNPs VII, X, XIII and IX significantly reduced promoter activity in response to infection. SNPs -2081 and -2026 significantly increased luciferase activity while -2604 and -2570 reduced the response to infection. (* = significant, ns = not significant, compared to uninfected cells carrying the same plasmids, Student T test).

Figure 6. Primary ABU associated GPs lower innate immune response to therapeutic inoculation of UTI prone patients.

(A) Patients (n = 15) were subjected to therapeutic intravesical inoculation with the prototype ABU strain E. coli 83972. The innate host response was quantified as the neutrophil numbers (B), IL-8 (C) and IL-6 (D) concentrations in urine at various times after inoculation. The response of patients with the ABU associated GPs (VI, VII, X, XIII) was compared to the most common and not UTI associated GPIV. Patients with the ABU associated GPs showed decreased responses compared to the GPIV controls. Pat.  =  Individual patient number. N = number of samples from each patient.