Abstract

Toll-like receptor 5 (TLR5) mediates innate immune responses to bacterial pathogens by binding to flagellin. A polymorphism in the TLR5 gene introduces a premature stop codon (TLR5392STOP) that is associated with susceptibility to legionnaires disease. Here we investigated whether TLR5392STOP was associated with typhoid fever. The frequency of TLR5392STOP was not significantly different in 565 patients with typhoid fever and 281 ethnically matched control subjects. Furthermore, TLR5 deficiency had no measurable effect on a number of clinical parameters associated with typhoid fever, including fever clearance time, pathogen burden, disease severity, or age at acquisition of disease. TLR5 may not play an important role in TLR-stimulated innate immune responses to human infection with Salmonella enterica serovar Typhi. Initiation of these responses may rely on other TLRs that recognize different bacterial ligands

The toll-like receptor (TLR) family consists of cell surface and endosomal receptors that recognize specific, conserved pathogen-associated molecular patterns present on infectious agents [1]. Stimulated TLRs activate the transcription factor nuclear factor κ-B and a signaling cascade that culminates in the increased expression of immune and proinflammatory genes, and it thereby plays an essential role in innate and adaptive immunity

We reported elsewhere that TLR5 recognizes bacterial flagellin and that ∼10% of individuals have a point mutation that introduces a stop codon within the ligand binding domain (TLR5392STOP) [2, 3]. The TLR5392STOP mutation functions as a dominant negative receptor that severely impairs signaling. This mutation is associated with susceptibility to pneumonia caused by the flagellated bacterium Legionella pneumophila [3]

Salmonella enterica serovar Typhi is a gram-negative, flagellated bacterium that can cause typhoid fever on ingestion. It is estimated that 22 million cases of typhoid fever occur worldwide each year, resulting in 200,000 deaths [4]. There is a significant burden of disease in developing countries, where sanitary conditions can be inadequate. In southern Vietnam, typhoid fever is the major cause of community-acquired septicemia, and the development of multiple-drug resistance in the pathogen is becoming a more serious problem in clinical care. The identification of genetic factors that predispose or protect individuals from typhoid fever can provide insight into its underlying disease mechanisms. We postulated that TLR5 plays an important role in protection against typhoid fever and performed a case-control genetic association study to examine this hypothesis

Subjects, materials, and methodsVenous blood samples were collected from patients with typhoid fever as part of larger epidemiological or treatment studies performed at either the Hospital for Tropical Diseases in Ho Chi Minh City, Dong Thap Provincial Hospital in Cao Lahn in Dong Thap Province, or Dong Nai Pediatric Center in Dong Nai Province, in southern Vietnam, between 1995 and 2002. These case patients were chosen for the study on the basis of blood or bone marrow cultures found positive for S. enterica serovar Typhi. These blood samples were either used in previous genetic association studies [5] or described elsewhere [6]. Venous blood samples were also collected from 281 healthy control subjects from Dong Thap Province and from patients undergoing minor surgery at Dong Nai Pediatric Center. All case patients and control subjects were unrelated and were of Vietnamese Kinh ethnicity. Informed consent was obtained from all individuals who participated in the study. Ethical approval was obtained from the ethical and scientific committees of the Hospital for Tropical Diseases, the Dong Thap Provincial Hospital, and Health Services of Dong Thap Province and from the institutional review board of Dong Nai Pediatric Center

DNA was extracted from 2-mL EDTA-treated blood samples by use of the QIAamp DNA Blood Midi Kit (Qiagen) or the Nucleon BACC 1 DNA extraction kit (Nucleon Biosciences). The TLR5392STOP polymorphism was genotyped by allele-specific polymerase chain reaction (PCR). Primers were designed that bound either to the wild-type sequence beginning with C at position 1174 (TTACAGACCTTGGATCTCC) or to the mutant sequence beginning with T at position 1174 (TTACAGACCTTGGATCTCT). The polymorphism-specific primers were used together with a conserved reverse primer (CAGAATCTGGAGATGAGGTACCCG) and internal control primers specific for human growth hormone (CAGTGCCTTCCCAACCATTCCCTTA [forward] and ATCCACTCACGGATTTCTGTTGTGTTTC [reverse]). Two PCRs that used either the C-specific or the T-specific primer together with the reverse primer (203-bp product) and the internal control primers (480-bp product) were performed on each DNA sample. PCRs contained 0.5 μmol/L primer, 200 μmol/L dNTPs, 1.5 mmol/L MgCl2, 0.45 mol/L betaine, 0.75 U of Taq polymerase, and 90 ng of DNA. Standard PCR cycling conditions were used, with an annealing temperature of 65°C. Allele-specific PCR products were separated on a 1% agarose gel

The clinical parameters investigated in this study (fever clearance time, quantitative blood count, quantitative bone marrow count of S. enterica serovar Typhi, white blood cell count, level of C reactive protein [CRP], level of antiflagellum antibody, and level of antilipopolysaccharide [anti-LPS] antibody) were measured in conjunction with other clinical studies, and the details of the procedures that were used in those investigations were published elsewhere [6–8 ]

Statistical analysis was performed using SPSS (version 10.0.5; SPSS). Pearson’s χ2 test was used to test associations between disease phenotypes (in case patients and control subjects and in groups according to disease severity) and allelic or genotypic frequencies. Fisher’s exact test was used when an expected value in the contingency table was <5. To compare genotypes with various clinical parameters that were not normally distributed, we used the Mann-Whitney U test. P⩽.05 was considered to be significant

ResultsThe aim of this study was to investigate the importance of the TLR5392STOP mutation in protection against typhoid fever. We recruited 565 patients with blood culture–confirmed typhoid fever. The median age of case patients was 10 years (range, 1–68 years). Two hundred eighty-one healthy subjects from the community and patients undergoing minor surgery who did not have a clinically apparent infectious disease were used as control subjects. The median age of control subjects was 6 years (range, 0–29 years)

We developed an allele-specific PCR assay to genotype the TLR5392STOP mutation, to investigate whether this mutation was associated with susceptibility to typhoid fever. Table 1 shows the allelic and genotypic frequencies of TLR5392STOP in 565 case patients and 281 control subjects. The allelic and genotypic frequencies in case patients and control subjects were not significantly different (P=.688 and P=.881, respectively). The allelic and genotypic frequencies in the Vietnamese individuals in the present study were similar to those found in Dutch white individuals [3]. Case patients were also grouped on the basis of genotype into those who had the TLR5392STOP mutation (C/T and T/T) and those who did not have the mutation (C/C), representing patients who had either mutant or wild-type TLR5 respectively. There was no association between having mutant TLR5 and having typhoid fever (P=.674; table 1)

Table 1

Allelic and genotypic frequencies of the 392 stop codon in toll-like receptor 5 in case patients and control subjects

Table 1

Allelic and genotypic frequencies of the 392 stop codon in toll-like receptor 5 in case patients and control subjects

These results suggested that the TLR5392STOP mutation was not associated with having typhoid fever. We then investigated whether the presence of the TLR5392STOP mutation had a more subtle effect on the innate immune response to infection with S. enterica serovar Typhi. We postulated that TLR5-deficient individuals would acquire their first fulminant infection at a younger age, that fever clearance times in these patients would be longer, that the bacterial load in their blood and bone marrow would be higher, and that their white blood cell counts and CRP levels would be lower, compared with individuals who have TLR5. We compared these clinical parameters in case patients who had either mutant or wild-type TLR5 (C/T and T/T combined vs. C/C). There was no significant difference between the 2 groups in terms of age at acquisition of disease, fever clearance time, quantitative blood counts of S. enterica serovar Typhi, quantitative bone marrow counts of S. enterica serovar Typhi, white blood cell counts, or CRP levels (table 2)

Table 2

Comparison of clinical parameters in case patients with and without the 392 stop codon in toll-like receptor 5 (TLR5)

Table 2

Comparison of clinical parameters in case patients with and without the 392 stop codon in toll-like receptor 5 (TLR5)

Case patients were grouped by genotype (C/T and T/T combined vs. C/C) and were compared on the basis of disease severity (case patients with severe disease, n=67; case patients with nonsevere disease, n=359). Severe disease was classified as intestinal perforation, gastrointestinal bleeding, abnormal state of consciousness, hemodynamic shock, jaundice/hepatitis, cholecystitis, myocarditis, renal impairment, pneumonia, or severe anemia. There was no association between having mutant TLR5 and having severe disease (χ2=0.746; P=.388). Furthermore, individuals with the TLR5392STOP mutation did not present with a specific type of severe disease (data not shown)

Titers of serum IgG antibody against flagella and against LPS from S. enterica serovar Typhi were measured in 58 case patients before the genotyping was completed [7]. We then compared the antiflagella and anti-LPS IgG titers in case patients who had either mutant TLR5 (C/T and T/T; n=9) or wild-type TLR5 (C/C; n=49). The length of illness was not significantly different between the genotyped groups (mutant TLR5: median, 4 days; interquartile range [IQR], 3–11 days; wild-type TLR5: median, 6 days; IQR, 4–10 days; P=.231, by Mann-Whitney U test). The anti-LPS titers in case patients who had wild-type TLR5 (median, 3200; IQR, 800–6400) was not significantly different than the anti-LPS titers in case patients who had mutant TLR5 (median, 400; IQR, 250–6400; P=.399, by Mann-Whitney U test). There was a trend for the antiflagella titers to be lower in case patients who had mutant TLR5 (median, 40; IQR, 40–200) than in patients who had wild-type TLR5 (median, 200; IQR, 45–800), but this difference did not reach statistical significance (P=.090, by Mann-Whitney U test)

DiscussionIn this study, we report that individuals who have the TLR5392STOP mutation are no more susceptible to typhoid fever than are individuals who have wild-type TLR5. Furthermore, the presence of the TLR5392STOP mutation had no effect on a number of clinical parameters associated with typhoid fever

For several reasons, TLR5 is a good candidate gene to use in genetic association studies of typhoid fever. First, S. enterica serovar Typhimurium flagellin binds to TLR5 and activates proinflammatory responses in the intestinal epithelia in vitro [9]. Gerwirtz et al. [9] reported that TLR5 is basolaterally expressed on intestinal epithelia and that translocation of S. enterica serovar Typhimurium flagellin across epithelia activated the expression of proinflammatory genes. Second, murine in vivo studies, including experiments that involved Salmonella infections and that were based on studies of the murine model of typhoid fever [2, 10], indicate that flagellin is an important stimulant of innate and adaptive immune responses. Third, the murine TLR5 gene lies within a locus that is associated with susceptibility to Salmonella infection [11]. In addition, TLR5 is associated with legionnaires disease, which is caused by infection with Legionella pneumophila. Legionella and Salmonella are similar in that they are both gram-negative, flagellated pathogens, but, most importantly, they occupy the same intramacrophage niche within the host. Finally, the presence of antiflagellin antibody responses in patients with typhoid fever clearly indicates that flagellin is expressed in vivo in humans and therefore is available for interaction with TLR5 [7]. Despite these reports of the importance of flagellin as a stimulant of the innate immune response in mice and humans, our results indicate that protective human immune responses to S. enterica serovar Typhi do not require TLR5

TLRs regulate innate immune responses and also play a crucial role in the initiation of adaptive immunity [12, 13]. Recent reports have shown that TLRs influence the activation of adaptive immune responses by 2 mechanisms. Primarily, TLRs initiate signaling by up-regulating costimulatory molecules, and this leads to maturation of dendritic cells [13, 14]. In addition, TLR-induced cytokines, most critically interleukin 6 (IL-6), are essential if T helper cells are to overcome the suppressive effect of CD4+CD25+ regulatory T cells and to generate pathogen-specific adaptive immune responses [15]. Interestingly, individuals who have 1 or 2 copies of the TLR5392STOP mutation have significantly decreased IL-6 production after stimulation with flagellin [3], and this decreased production may affect this mechanism of TLR-activated adaptive immunity. This effect may be related to our results that showed a trend for the antiflagella IgG titer to be lower in case patients who have mutant TLR5 than in case patients who have wild-type TLR5 (although this titer was not statistically significantly lower)

There may be sufficient redundancy in the TLR pathway to obviate the requirement of TLR5 for a protective immune response to S. enterica serovar Typhi. TLRs other than TLR5 may be important activators of innate immunity to typhoid fever. Numerous reports have detailed the importance of TLR4-mediated activation in S. enterica serovar Typhimurium infections [16, 17] but this has not been investigated in S. enterica serovar Typhi infections. In white individuals, a defective TLR4 mutation (Asp299Gly) is associated with endotoxin hyporesponsiveness [18] and gram-negative sepsis [19]. This TLR4 mutation is not present, or is present at a very low frequency, in the Vietnamese population we have studied (S.J.D., N.T.H., T.T.H., J.J.F., unpublished data). We are currently identifying mutations within TLR4 in the Vietnamese Kinh, to investigate the importance of TLR4 in immunity to typhoid fever. In addition to TLR4, TLR9, which binds to CpG DNA, may also contribute to TLR-stimulated innate immune responses in typhoid fever

In the case of typhoid fever, results from studies of murine infection with S. enterica serovar Typhimurium cannot always be extrapolated for use in studies of human infection with S. enterica serovar Typhi. Our study has shown that the investigation of potential susceptibility loci within the human host is important if we are to understand typhoid fever

Acknowledgments

We acknowledge the directors and the staff of the Hospital for Tropical Disease, Dong Thap Provincial Hospital, and Dong Nai Pediatric Hospital, for the clinical and microbiological work associated with this study. We also acknowledge the contribution made by Christine Luxemburger in the collection of samples in Dong Thap Province. We thank the Vietnamese individuals who took part in this study

References

1.
Akira
S
Takeda
K
Toll‐like receptor signalling
Nat Rev Immunol
 , 
2004
, vol. 
4
 (pg. 
499
-
511
)
2.
Hayashi
F
Smith
KD
Ozinsky
A
, et al.  . 
The innate immune response to bacterial flagellin is mediated by Toll‐like receptor 5
Nature
 , 
2001
, vol. 
410
 (pg. 
1099
-
103
)
3.
Hawn
TR
Verbon
A
Lettinga
KD
, et al.  . 
A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires’ disease
J Exp Med
 , 
2003
, vol. 
198
 (pg. 
1563
-
72
)
4.
Crump
JA
Luby
SP
Mintz
ED
The global burden of typhoid fever
Bull World Health Org
 , 
2004
, vol. 
82
 (pg. 
346
-
53
)
5.
Dunstan
SJ
Stephens
HA
Blackwell
JM
, et al.  . 
Genes of the class II and class III major histocompatibility complex are associated with typhoid fever in Vietnam
J Infect Dis
 , 
2001
, vol. 
183
 (pg. 
261
-
8
)
6.
Vinh
H
Wain
J
Vo
TN
, et al.  . 
Two or three days of ofloxacin treatment for uncomplicated multidrug‐resistant typhoid fever in children
Antimicrob Agents Chemother
 , 
1996
, vol. 
40
 (pg. 
958
-
61
)
7.
House
D
Wain
J
Ho
VA
, et al.  . 
Serology of typhoid fever in an area of endemicity and its relevance to diagnosis
J Clin Microbiol
 , 
2001
, vol. 
39
 (pg. 
1002
-
7
)
8.
Wain
J
Pham
VB
Ha
V
, et al.  . 
Quantitation of bacteria in bone marrow from patients with typhoid fever: relationship between counts and clinical features
J Clin Microbiol
 , 
2001
, vol. 
39
 (pg. 
1571
-
6
)
9.
Gewirtz
AT
Navas
TA
Lyons
S
Godowski
PJ
Madara
JL
Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression
J Immunol
 , 
2001
, vol. 
167
 (pg. 
1882
-
5
)
10.
McSorley
SJ
Ehst
BD
Yu
Y
Gewirtz
AT
Bacterial flagellin is an effective adjuvant for CD4+ T cells in vivo
J Immunol
 , 
2002
, vol. 
169
 (pg. 
3914
-
9
)
11.
Sebastiani
G
Leveque
G
Lariviere
L
, et al.  . 
Cloning and characterization of the murine toll‐like receptor 5 (Tlr5) gene: sequence and mRNA expression studies in Salmonella‐susceptible MOLF/Ei mice
Genomics
 , 
2000
, vol. 
64
 (pg. 
230
-
40
)
12.
Pasare
C
Medzhitov
R
Toll‐like receptors and acquired immunity
Semin Immunol
 , 
2004
, vol. 
16
 (pg. 
23
-
6
)
13.
Akira
S
Takeda
K
Kaisho
T
Toll‐like receptors: critical proteins linking innate and acquired immunity
Nat Immunol
 , 
2001
, vol. 
2
 (pg. 
675
-
80
)
14.
Schnare
M
Barton
GM
Holt
AC
Takeda
K
Akira
S
Medzhitov
R
Toll‐like receptors control activation of adaptive immune responses
Nat Immunol
 , 
2001
, vol. 
2
 (pg. 
947
-
50
)
15.
Pasare
C
Medzhitov
R
Toll pathway–dependent blockade of CD4+CD25+ T cell–mediated suppression by dendritic cells
Science
 , 
2003
, vol. 
299
 (pg. 
1033
-
6
)
16.
Bernheiden
M
Heinrich
JM
Minigo
G
, et al.  . 
LBP, CD14, TLR4 and the murine innate immune response to a peritoneal Salmonella infection
J Endotoxin Res
 , 
2001
, vol. 
7
 (pg. 
447
-
50
)
17.
Weiss
DS
Raupach
B
Takeda
K
Akira
S
Zychlinsky
A
Toll‐like receptors are temporally involved in host defense
J Immunol
 , 
2004
, vol. 
172
 (pg. 
4463
-
9
)
18.
Arbour
NC
Lorenz
E
Schutte
BC
, et al.  . 
TLR4 mutations are associated with endotoxin hyporesponsiveness in humans
Nat Genet
 , 
2000
, vol. 
25
 (pg. 
187
-
91
)
19.
Agnese
DM
Calvano
JE
Hahm
SJ
, et al.  . 
Human toll‐like receptor 4 mutations but not CD14 polymorphisms are associated with an increased risk of gram‐negative infections
J Infect Dis
 , 
2002
, vol. 
186
 (pg. 
1522
-
5
)
Financial support: The Wellcome Trust