Search tips
Search criteria 


Logo of plosonePLoS OneView this ArticleSubmit to PLoSGet E-mail AlertsContact UsPublic Library of Science (PLoS)
PLoS One. 2010; 5(10): e13759.
Published online 2010 October 29. doi:  10.1371/journal.pone.0013759
PMCID: PMC2966434

Common Variants of TLR1 Associate with Organ Dysfunction and Sustained Pro-Inflammatory Responses during Sepsis

Rory Edward Morty, Editor



Toll-like receptors (TLRs) are critical components for host pathogen recognition and variants in genes participating in this response influence susceptibility to infections. Recently, TLR1 gene polymorphisms have been found correlated with whole blood hyper-inflammatory responses to pathogen-associated molecules and associated with sepsis-associated multiorgan dysfunction and acute lung injury (ALI). We examined the association of common variants of TLR1 gene with sepsis-derived complications in an independent study and with serum levels for four inflammatory biomarkers among septic patients.

Methodology/Principal Findings

Seven tagging single nucleotide polymorphisms of the TLR1 gene were genotyped in samples from a prospective multicenter case-only study of patients with severe sepsis admitted into a network of intensive care units followed for disease severity. Interleukin (IL)-1β, IL-6, IL-10, and C-reactive protein (CRP) serum levels were measured at study entry, at 48 h and at 7th day. Alleles -7202G and 248Ser, and the 248Ser-602Ile haplotype were associated with circulatory dysfunction among severe septic patients (0.001≤p≤0.022), and with reduced IL-10 (0.012≤p≤0.047) and elevated CRP (0.011≤p≤0.036) serum levels during the first week of sepsis development. Additionally, the -7202GG genotype was found to be associated with hospital mortality (p = 0.017) and ALI (p = 0.050) in a combined analysis with European Americans, suggesting common risk effects among studies.


These results partially replicate and extend previous findings, supporting that variants of TLR1 gene are determinants of severe complications during sepsis.


Sepsis is a devastating clinical condition characterized by systemic inflammation occurring in the setting of a severe infection. Among intensive care unit (ICU) patients, incidence and mortality has been estimated in ~12% and >40%, respectively, for the most severe forms [1]. Over the past three decades, a plethora of experimental and clinical studies have contributed substantially to our understanding of sepsis development and associated complications, including the acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS) [2]. Studies in animal models, and twin and association studies have evidenced that the innate immune responses to pathogens show inter-individual variability that is strongly influenced by genetic factors, which may affect disease susceptibility and severity [3], [4], [5].

Toll-like receptor (TLR) pathways are critical components of the immune response to pathogens [6], and targeting of TLRs have been demonstrated to protect from lethal sepsis [7], [8]. Polymorphisms in genes of the TLR-mediated responses are associated with altered immunity [9], [10], [11], with susceptibility to infections and with related acute inflammatory syndromes, including sepsis and severe complications such as ALI and ARDS [12], [13]. In an elegant study, Wurfel and colleagues screened tagging single nucleotide polymorphisms (SNPs) across 49 TLR-related genes for association with whole blood inflammatory responses to pathogen-associated molecules in normal volunteers [14]. This allowed the identification of various tightly linked SNPs from the TLR1 gene associated with the strongest hyper-inflammatory effects. In a subsequent association study in septic patients focusing on two of the common TLR1 SNPs (-7202A/G and Ser602Ile) revealed associations with organ dysfunction, 28-day hospital mortality, ALI, and the prevalence of gram-positive cultures, constituting the unique association study of TLR1 gene variants with sepsis-associated complications to date.

Given the controversies around the association of genetic factors with sepsis-associated outcomes because of the limited replicability of findings [15], [16], here we have examined the association of common variants of TLR1 gene with sepsis outcomes in a cohort of patients with severe sepsis admitted into a network of ICUs from Spain, in order to provide an independent replication of previous results. We additionally explored the relationship of TLR1 polymorphisms and of particular haplotypes with serum levels of four biomarkers of inflammation taken at three stages of sepsis development. We found that TLR1 SNPs and haplotypes were associated with circulatory dysfunction and the source of infection among severe septic patients, and with hospital mortality and ALI in a combined analysis with data from European Americans. Congruently, risk alleles associated with increased nuclear factor κB (NF-κB) activation upon TLR1 stimulation in previous studies [17], [18], were related with reduced interleukin-10 and elevated C-reactive protein serum levels during the first week of sepsis development.


Ethics Statement

This study was approved by the Ethics Committees of Hospital Universitario N. S. de Candelaria and Hospital Universitario Río Hortega (Spain). Written informed consent was obtained from each subject or appropriate surrogates.

Study Subjects

Samples were collected as part of a prospective, observational study of adults admitted during 2003–2005 into a network of Spanish ICUs. The study has been reported elsewhere [19] and included consecutive patients admitted to ICUs older than 18 years old fulfilling the international criteria for severe sepsis (n = 218) [20]. We collected basic demographic data, previous health status, severity of illness scores and clinical information until discharge from the ICU, including source of infection, pathogens, and development of organ dysfunction. All patients were followed prospectively for the development of ALI, as defined by the American-European Consensus Conference [21] and for the development of organ dysfunction included in the sequential organ failure assessment (SOFA) scale [22]. For the purpose of this study, patients with ALI and ARDS were analyzed as a single group of ALI patients. Genomic DNA was isolated from peripheral whole blood using commercial kits (GFX kit, GE Healthcare, Little Chalfont, UK) and stored at −20°C until use. A summary description of cases can be found in Table 1.

Table 1
Demographic and clinical features of patients with severe sepsis.

DNA samples from population-based controls (n = 346), randomly drawn from a population-based study of about 7000 unrelated individuals as a representation of the general Spanish population [23], were also genotyped but data were only used for quality control purposes and not for association studies.

SNP selection and genotyping

Instead of focusing exclusively on the two previously associated variants [14], we selected tagging SNPs (tSNPs) to efficiently study common TLR1 gene variation in order to test whether additional gene variants associated with the expression of the clinical phenotype. TagIT 3.03 software [24] was used to select a set of 7 tSNPs from data available from 23 European-Americans retrieved from the Innate Immunity NHLBI Program for Genomic Applications (PGA) website (, accessed October 2008), forcing the inclusion of previously associated tSNPs by Wurfel and colleagues [14]. A SNP-dropping-with-re-sampling method [24] was used to evaluate the expected properties of the TLR1 7 tSNPs set, indicating an average haplotype r2>0.90.

Genotyping was conducted using the iPLEX™ Gold assay on MassARRAY system (Sequenom, San Diego, CA) by the Spanish National Genotyping Center, Santiago de Compostela Node (CeGen, Briefly, iPLEX™ assays were scanned by MALDI-TOF mass spectrometry and individual SNP genotype calls were automatically generated using Sequenom TYPER 3.4™ software. Two tSNPs that failed in iPLEX™ Gold assays (rs5743618 and rs5743551) were genotyped at Hospital Universitario N. S. de Candelaria using TaqMan™ assays (Applied Biosystems, Foster City, CA) in a 7500 Fast Real-Time PCR System (Applied Biosystems), with automated calls generated using the 7500 software 2.0.1 based on discriminating plots with 95% confidence. Genotyping was blind to sample outcomes. Approximately 10% of the samples were genotyped by duplicate to monitor genotyping quality. Genotypes were assigned using all data from the study simultaneously. Discrepancy rate among duplicates was 0.51% (95% confidence interval [CI]: 0.17–1.49%) for iPLEX™ Gold and 0.00% (95% CI: 0.00–2.11%) for TaqMan™. Call rates and further quality-control measures for genotyped tSNPs, as well as the correspondence of dbSNPs reference numbers with SNP positions relative to the start codon, can be found in Table 2.

Table 2
Quality control measures for the TLR1 tSNPs.

Serum measurements

Serum was available only from a subset of 121 patients as dry-ice transportation to the coordinating center was not warranted for all participating centers. These samples were collected within 24 hours of meeting severe sepsis criteria, at 48 hours and 7 days after study entry, only if the patient remained hospitalized into the ICU. Given this, serum was available for all 121 patients at inclusion, for 96 of these patients at 48 hours, and for 60 of these patients at the 7th day. Serum was obtained by centrifuging peripheral blood samples for 10 min at 3200 rpm within 35 min after sampling and was kept at −80°C until measurements were performed. These samples have been previously measured for four commonly used biomarkers of acute inflammation, interleukin (IL)-6, IL-1β, IL-10, and C-reactive protein (CRP) [25]. Measurements were done by duplicate using commercial kits: IL-6, IL-1β, and IL-10 were assessed in an Immulite analyzer (Siemens Medical Solutions Diagnostics, Caernarfon, UK), and CRP was determined in a Hitachi 917 analyzer (Roche Diagnostic, Basilea, Switzerland).

Statistical analysis

Departures from Hardy-Weinberg equilibrium (HWE) were only tested in controls by means an efficient exact test [26]. Associations of SNPs with sepsis-derived complications were assessed using logistic regression models by means of SNPassoc [27] assuming additive effects for all SNPs. To replicate previous association findings [14], recessive models were also applied for -7202A/G and Ser602Ile SNPs as necessary. Logistic regressions were also used to obtain SNP effects as odds ratios (ORs) with 95% CI adjusting for age, gender, and the acute physiology and chronic health evaluation score (APACHE) II. Genotype data available from 20 unlinked polymorphisms from different parts of the genome [13] were utilized to adjust individual SNP associations for population stratification by means of a modified Cochran-Armitage trend test [28]. A Mantel-Haenszel test over genotype counts was used to perform the joint stratified analysis of the two SNPs overlapping across our study and the previous one [14], under the principle that if SNPs had consistent effects (i.e. allele and effect-wise) among different studies, the joint analysis would improve the results obtained from the two studies by separate. Data were analyzed under recessive models with -7202GG and 602Ile/Ile genotypes as risk factors to emulate previous association findings [14]. For this analysis, we did not consider non-septic control sample genotype counts from Wurfel et al. [14], and used data from the sepsis cohort to test associations with hospital mortality and the CELEG cohort to test associations with ALI development (Table S1). THESIAS [29] was used to assess the association of particular TLR1 haplotypes with sepsis-derived outcomes. A two-tailed p≤0.05 was considered significant. However, in order to estimate the chance that SNP associations were true positives, we calculated the false-discovery rate (FDR) overall comparisons using QVALUE [30]. The patterns of linkage disequilibrium (LD), in terms of r2 values, were explored using Haploview 3.32 [31].

Differences for serum biomarker levels between genotypes were explored using general linear models (GLM) for repeated measures in a longitudinal analysis over the three time points by means of SPSS 15.0 (SPSS Inc., Chicago, IL). Although the four biomarkers were measured for all 121 serum samples at inclusion, 96 serum samples at 48 h and 60 serum samples at 7th day, the GLM analysis tested for biomarker level differences between genotypes over time. Thus, it only considered the 60 samples for which measures were available for the three time points. Despite dropping the sample size, GLM allowed to explore sustained SNP relationships with biomarkers, under the rationale that these were more likely to identify genuine associations. A post hoc analysis of differences in biomarker levels between genotypes at individual time points was also performed using all available serum measures from each day by means of linear regression. These analyses were performed under recessive models in order to emulate previous functional findings in vitro [17], [32], [33]. Values for IL-6, IL-1β, and IL-10 were logarithmically transformed for the analysis.

Results and Discussion

Among the 218 severe septic patients, the median APACHE II score was 23 (interquartile range: 18–27), 55.0% percent developed ARDS and 31.1% developed ALI during hospitalization (Table 1). Peritonitis was the leading cause of severe sepsis followed by pneumonia, and most pathogens were characterized as Gram-negative bacteria. In agreement with previous studies [34], no pathogens were identified in blood cultures as the causative microorganism for sepsis albeit having an identified site of infection for 48.8% of patients. Hospital mortality was 52.7%.

Table 2 shows that completion rates were >97% for all 7 tSNPs and that all followed HWE expectations in control samples. None of the tSNPs was significantly associated with hospital mortality or with sepsis-induced ALI (Table 3). Borderline associations were observed for the SNP Arg80Thr (rs5743611) with hospital mortality (p = 0.080), and for SNPs -7202A/G (rs5743551) and Asn248Ser (rs4833095) with sepsis-induced ALI (p = 0.100 and p = 0.090, respectively). The association of -7202A/G with hospital mortality or with ALI did not improve if a recessive model was used. However, analyses combining samples from this and the previous study by Wurfel et al. [14], enabled the detection of a weak risk of the -7202GG genotype for hospital mortality and for ALI consistent across the two studies. The significance of the association of -7202A/G with hospital mortality improved to p = 0.017 (p = 0.028 in the previous study [14] and p = 0.566 in this study). Similarly, significance of the association of -7202A/G with ALI among septic patients improved to p = 0.050, while separate findings were non-significant (p = 0.171 in the previous study [14] and p = 0.108 in this study). For Ile602Ser (rs5743618), analyses combining samples from this and the previous study by Wurfel et al. [14], did not enable the detection of consistent significant effects (results not shown). Contrary to previous findings [14], none of the tSNPs showed a significant association with prevalence of Gram-positive infections versus other types of pathogens, but the SNPs -2299C/T (rs5743594) and -7202A/G showed borderline associations (p = 0.067 and p = 0.110, respectively). Nevertheless, in our study, three tSNPs had nominal significance in association with pulmonary infection among septic patients, which is frequently caused by Gram-positive bacteria [1], [35]: -5531A/G (rs5743565) with p = 0.004, -2076C/T (rs5743596) with p = 0.004, and Asn248Ser with p = 0.037. However, when adjusted for covariates using regression models, results remained significant for -5531A/G (OR for -5531A: 1.84, 95% CI: 1.10–3.10, p = 0.023) and for -2076C/T (OR for -2076C: 1.90, 95% CI: 1.07–3.39, p = 0.029), but not for Asn248Ser (OR for 248Ser: 1.44, 95% CI: 0.92–2.26, p = 0.112). Similar findings were obtained after applying a conservative correction for population stratification, as only the association of -5531A/G with pulmonary infection remained significant (p = 0.047, p = 0.057 and p = 0.127 for -5531A/G, -2076C/T and Asn248Ser, respectively).

Table 3
Association p-values of TLR1 tSNPs with severe sepsis complications.

The strongest associations of TLR1 tSNPs with sepsis-derived complications were found with circulatory dysfunction (Table 3). Three tSNPs, -7202A/G, Asn248Ser, and Ser602Ile, showed p-values ranging from 0.001 to 0.024. The association of -7202A/G remained significant after adjusting for covariates (OR for -7202G: 2.68, 95% CI: 1.40–5.12, p = 0.003) and population stratification (p = 0.031). The associations for Asn248Ser and Ser602Ile also persisted when adjusting for covariates (OR for 248Ser: 2.13, 95% CI: 1.16–3.90, p = 0.015; OR for 602Ile: 1.94, 95% CI: 1.10–3.44, p = 0.023) and showed borderline significance after the conservative adjustment for population stratification (p = 0.061 for Asn248Ser and p = 0.064 for Ser602Ile). In the context of the multiple tests performed, some of the above findings might represent true positives. In particular, FDR was below 5% for the association of -7202G/A with circulatory dysfunction, which partially replicates the study by Wurfel et al. [14], which reported the association of this TLR1 SNP with circulatory, neurologic, coagulation, renal and hepatic dysfunction among septic patients. However, one must take into account the power limitation of this association study. Assuming a minor allele frequency of 30% (as is the case for -7202A/G) and an OR of 2.0, in the range of values reported by Wurfel et al [14], statistical power ranged from 39% to 69% for the outcomes tested, which might be responsible for the lack of replication in the association of TLR1 variants with some of the outcomes.

Many evidences point to the SNP Ser602Ile as the main responsible for the inter-individual variation in TLR1-mediated responses, 602Ser allele being related to a decrease in NF-κB activity [14], [17], [18], [32]. Albeit the SNP Asn248Ser have demonstrated no functional relevance [14], [18], previous association studies have indicated that it might be important for disease susceptibility [14], [36], [37], [38]. Thus, we were interested in testing the association of the ancestral haplotype 248Ser-602Ile [18] with sepsis-derived complications, motivated by the observation that this haplotype consistently tends to associate with the highest NF-κB activation upon TLR1 stimulation in independent studies [17]. In particular, we were interested in evaluating whether haplotypes of these two SNPs were more strongly associated with circulatory dysfunction than the two SNPs by separate. However, our results showed that, although 248Ser-602Ile haplotype homozygous subjects were more frequent among patients with circulatory dysfunction (17.5%) than in the rest of patients (5.6%), being a risk factor for circulatory dysfunction (OR: 2.07, 95% CI: 1.12–3.82, p = 0.020), statistical significance did not improve considerably compared to those obtained for Asn248Ser and Ser602Ile individually.

A number of studies have explored the functional effects of TLR1 variants in vitro [14], [17], [32]. However, no study have explored if TLR1 variants are associated with amplified immune responses to pathogens in vivo. Given that serial serum measures for four biomarkers of inflammation were available from these patients during the first week after the onset of severe sepsis (Table S2), we correlated tSNPs genotypes with biomarker serial measures using GLM (Table 4). None of the tSNPs was found to be associated with serially-measured IL-1β and IL-6 serum levels, despite previous studies have demonstrated that TLR1 gene polymorphisms determined IL-6 levels in response to specific gram-negative bacterial products [14]. On the other hand, CRP and IL-10 levels varied significantly among genotypes for four of the TLR1 variants, with p-values ranging from 0.002 to 0.047. However, given the assessment used in this study to explore TLR1 SNP relationships with biomarkers, statistical power was limited for most comparisons, particularly for those involving IL-6 and IL-1β for which it was below 30% (Table 4). Nevertheless, congruent with our previous results, homozygote patients for -7202G or 248Ser alleles, both constituting risk alleles for sepsis-associated complications, showed significantly increased levels for CRP (p = 0.036 and p = 0.012, respectively) and lower levels for the compensatory anti-inflammatory cytokine IL-10 (p = 0.011 and p = 0.047, respectively) (Table 4), regardless of the in vitro findings suggesting higher IL-10 levels for the -7202GG genotype in response to TLR1/TLR2 agonists [14]. This sustained pro-inflammatory state was also observed for patients homozygous for the risk haplotype 248Ser-602Ile (n = 8), associated with the highest NF-κB activation upon TLR1 stimulation [17], compared to the rest of patients (n = 52; p = 0.024 and p = 0.015 for CRP and IL-10 serial measures, respectively; p≥0.45 for IL-1β and IL-6) (Figure 1), slightly improving significance levels over individual SNP comparisons. Results for the longitudinal analysis barely changed when tests were performed under additive models or when adjusted for age, gender, and APACHE II (not shown).

Figure 1
IL-6, IL-1β, IL-10, and CRP serum levels according to the TLR1 248Ser-602Ile haplotype status.
Table 4
P-values and statistical power (%) from the longitudinal analysis of differences in serum biomarker levels by TLR1 tSNPs.

It is worth noting that no demographical or clinical differences were found between samples considered in the GLM analysis of TLR1 SNP genotypes and serial measures of biomarkers, i.e. those patients who were not discharged from the ICU or were not dead before seven days, and patients without serum measures at the three time points, except for mortality (p = 0.033). Thus, it is reasonable to assume that this GLM biomarker analysis might be biased towards the expected for survivor patients. We explored such possibility by a post hoc analysis of differences in biomarker levels at individual time points among genotypes using all available serum measures. This analysis supported the strong correlation of TLR1 SNPs with CRP and IL-10 levels observed before, while revealed barely significant relationships of TLR1 SNPs with IL-6 and IL-1β levels (Table S3), indicating that GLM results were not as biased as predicted.

Due to the remarkable effects on NF-κB activation, Ser602Ile has been put forward as the principal functional variant affecting TLR1 activity across the whole chromosome region encompassing TLR10-TLR1-TLR6 genes [18]. Nevertheless, we found stronger associations for SNPs -7202G/A and Asn248Ser with sepsis complications and with biomarker serum levels, in agreement with previous association studies [14], [36], albeit no functional properties have been described for them individually [14]. The LD between -7202G/A and Asn248Ser was strong (r2 = 0.84), while the LD of these two with Ser602Ile was moderate in our sample (Figure 2), as has been found elsewhere [18], [39]. Furthermore, Asn248Ser have been associated with Leprosy and Malaria in populations where Ile602Ser is rare or absent [37], [38], indicating that it (or other variant in LD with it) might be relevant for susceptibility to infections. Given this, and the fact that associations found for Ser602Ile were weak unless considered into haplotypes, the evidences suggest that a more complex scenario, possibly involving other rare and common functional variants, underlie TLR1 gene effect on the phenotype. Alternatively, unexplored epistatic interactions with other gene variants relevant for the TLR-mediated responses (e.g. TLR4 Asp299Gly, TIRAP/Mal Ser180Leu, etc) might be needed to be able to detect Ser602Ile effects on this devastating clinical condition.

Figure 2
Linkage disequilibrium (LD) plot of r2 values between TLR1 SNPs.

In conclusion, despite the limited power of the study, here we have replicated the association of TLR1 gene variants, with established increased NF-κB activation in vitro, with risk for circulatory dysfunction during severe sepsis, and associated TLR1 variants with elevated CRP and decreased IL-10 levels in serum during the first week of sepsis development. In line with previous evidences, our results suggest that genetic factors favouring an exacerbated host pro-inflammatory response to an infection predisposes to sepsis complications.

Supporting Information

Table S1

Genotype counts for the TLR1 SNPs used in meta-analysis.

(0.04 MB DOC)

Table S2

Mean (±SEM) serum levels of IL-6, IL-1β, IL-10 and CRP stratified by TLR1 SNP genotypes.

(0.05 MB DOC)

Table S3

P-values from the post hoc analysis of differences in IL-1β, IL-6, CRP and IL-10 levels among TLR1 genotypes at individual time points using all available serum measures from each time point (at inclusion, 48 hours, and 7th day).

(0.04 MB DOC)


GEN-SEP investigators are:

Flores C,a,b Villar J,a,c Espinosa E,d Muros M,a,e Pérez-Méndez L,a,b, Pino-Yanes Ma,b, Torres A,a,f,g Ferrer M,a,f,g Zavala E,a,h Adalia R.a,h

aCIBER de Enfermedades Respiratorias, Instituto de Salud Carlos III, Madrid, Spain; bResearch Unit, Hospital Universitario N.S. de Candelaria, Santa Cruz de Tenerife, Spain; cMultidisciplinary Organ Dysfunction Evaluation Research Network, Research Unit, Hospital Universitario Dr. Negrín, Las Palmas de Gran Canaria, Spain; dDepartment of Anesthesia, Hospital Universitario N.S. de Candelaria, Santa Cruz de Tenerife, Spain; eDept. of Clinical Biochemistry, Hospital Universitario N.S. de Candelaria, Santa Cruz de Tenerife, Spain; fPneumology Department, Clinic Institute of Thorax, Hospital Clinic, Barcelona, Spain; gInsitut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain; hSurgical Intensive Care Unit, Department of Anesthesia, Hospital Clinic, Barcelona, Spain.

GRECIA investigators are:

Blanco J,a,b Muriel A,a Sagredo V,c Ballesteros JC,d Taboada F,d Muñiz G,d Gandía F,e Bobillo F,e Tamayo L,f Labattut AG,g Collado J,h Valledor M,i Antuña MT,i López MJ,j Cortina JJ,j Saldaña T,k Caballero A,k Álvarez T,k De Frutos M,l Guerra J,m Álvarez B,n Sandoval J.n

aMedicina Intensiva, Hospital Universitario Rio Hortega, Valladolid, Spain; bCIBER de Enfermedades Respiratorias, Madrid, Spain; cMedicina Intensiva Hospital Clínico Universitario, Salamanca, Spain; dMedicina Intensiva, Hospital Central de Asturias, Oviedo; eMedicina Intensiva, Hospital Clinico Universitario, Valladolid, Spain; fMedicina Intensiva, Hospital Universitario Río Carrión, Palencia, Spain; gMedicina Intensiva, Hospital General de Soria, Soria; hMedicina Intensiva, Complejo Hospitalario de León, León, Spain; iMedicina Intensiva, Hospital San Agustín, Aviles, Spain; jMedicina Intensiva, Hospital General de Segovia, Segovia, Spain; kMedicina Intensiva, Hospital Virgen de la Concha, Zamora, Spain; lMedicina Intensiva, Hospital General Yagüe, Burgos, Spain; mMedicina Intensiva, Hospital de Cabueñes, Gijón, Spain; nMedicina Intensiva, Hospital del Bierzo, Ponferrada, Spain.


Competing Interests: The authors have declared that no competing interests exist.

Funding: This work was supported by grants from Ministry of Science of Spain (SAF 2004-06833), Fundación Canaria de Investigación y Salud (53/04), and by a specific agreement between Instituto de Salud Carlos III and Gobierno de Canarias (EMER07/001). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.


1. Blanco J, Muriel-Bombin A, Sagredo V, Taboada F, Gandia F, et al. Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study. Crit Care. 2008;12:R158. [PMC free article] [PubMed]
2. Matthay MA, Zimmerman GA. Acute lung injury and the acute respiratory distress syndrome: four decades of inquiry into pathogenesis and rational management. Am J Respir Cell Mol Biol. 2005;33:319–327. [PMC free article] [PubMed]
3. Wurfel MM, Park WY, Radella F, Ruzinski J, Sandstrom A, et al. Identification of high and low responders to lipopolysaccharide in normal subjects: an unbiased approach to identify modulators of innate immunity. J Immunol. 2005;175:2570–2578. [PubMed]
4. de Craen AJ, Posthuma D, Remarque EJ, van den Biggelaar AH, Westendorp RG, et al. Heritability estimates of innate immunity: an extended twin study. Genes Immun. 2005;6:167–170. [PubMed]
5. Aziz RK, Kansal R, Abdeltawab NF, Rowe SL, Su Y, et al. Susceptibility to severe Streptococcal sepsis: use of a large set of isogenic mouse lines to study genetic and environmental factors. Genes Immun. 2007;8:404–415. [PubMed]
6. Trinchieri G, Sher A. Cooperation of Toll-like receptor signals in innate immune defence. Nat Rev Immunol. 2007;7:179–190. [PubMed]
7. Roger T, Froidevaux C, Le Roy D, Reymond MK, Chanson AL, et al. Protection from lethal gram-negative bacterial sepsis by targeting Toll-like receptor 4. Proc Natl Acad Sci U S A. 2009;106:2348–2352. [PubMed]
8. Alves-Filho JC, Freitas A, Souto FO, Spiller F, Paula-Neto H, et al. Regulation of chemokine receptor by Toll-like receptor 2 is critical to neutrophil migration and resistance to polymicrobial sepsis. Proc Natl Acad Sci U S A. 2009;106:4018–4023. [PubMed]
9. Arbour NC, Lorenz E, Schutte BC, Zabner J, Kline JN, et al. TLR4 mutations are associated with endotoxin hyporesponsiveness in humans. Nat Genet. 2000;25:187–191. [PubMed]
10. Khor CC, Chapman SJ, Vannberg FO, Dunne A, Murphy C, et al. A Mal functional variant is associated with protection against invasive pneumococcal disease, bacteremia, malaria and tuberculosis. Nat Genet. 2007;39:523–528. [PMC free article] [PubMed]
11. Shey MS, Randhawa AK, Bowmaker M, Smith E, Scriba TJ, et al. Single nucleotide polymorphisms in toll-like receptor 6 are associated with altered lipopeptide- and mycobacteria-induced interleukin-6 secretion. Genes Immun. In press. [PMC free article] [PubMed]
12. Arcaroli J, Silva E, Maloney JP, He Q, Svetkauskaite D, et al. Variant IRAK-1 haplotype is associated with increased nuclear factor-kappaB activation and worse outcomes in sepsis. Am J Respir Crit Care Med. 2006;173:1335–1341. [PMC free article] [PubMed]
13. Flores C, Perez-Mendez L, Maca-Meyer N, Muriel A, Espinosa E, et al. A common haplotype of the LBP gene predisposes to severe sepsis. Crit Care Med. 2009;37:2759–2766. [PubMed]
14. Wurfel MM, Gordon AC, Holden TD, Radella F, Strout J, et al. Toll-like receptor 1 polymorphisms affect innate immune responses and outcomes in sepsis. Am J Respir Crit Care Med. 2008;178:710–720. [PMC free article] [PubMed]
15. Clark MF, Baudouin SV. A systematic review of the quality of genetic association studies in human sepsis. Intensive Care Med. 2006;32:1706–1712. [PubMed]
16. Flores C, Pino-Yanes MM, Villar J. A quality assessment of genetic association studies supporting susceptibility and outcome in acute lung injury. Crit Care. 2008;12:R130. [PMC free article] [PubMed]
17. Hawn TR, Misch EA, Dunstan SJ, Thwaites GE, Lan NT, et al. A common human TLR1 polymorphism regulates the innate immune response to lipopeptides. Eur J Immunol. 2007;37:2280–2289. [PubMed]
18. Barreiro LB, Ben-Ali M, Quach H, Laval G, Patin E, et al. Evolutionary dynamics of human Toll-like receptors and their different contributions to host defense. PLoS Genet. 2009;5:e1000562. [PMC free article] [PubMed]
19. Villar J, Perez-Mendez L, Flores C, Maca-Meyer N, Espinosa E, et al. A CXCL2 polymorphism is associated with better outcomes in patients with severe sepsis. Crit Care Med. 2007;35:2292–2297. [PubMed]
20. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med. 2003;31:1250–1256. [PubMed]
21. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, et al. The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med. 1994;149:818–824. [PubMed]
22. Vincent JL, de Mendonca A, Cantraine F, Moreno R, Takala J, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsis-related problems” of the European Society of Intensive Care Medicine. Crit Care Med. 1998;26:1793–1800. [PubMed]
23. Cabrera de Leon A, Rodriguez Perez Mdel C, Almeida Gonzalez D, Dominguez Coello S, Aguirre Jaime A, et al. [Presentation of the “CDC de Canarias” cohort: objectives, design and preliminary results]. Rev Esp Salud Publica. 2008;82:519–534. [PubMed]
24. Ahmadi KR, Weale ME, Xue ZY, Soranzo N, Yarnall DP, et al. A single-nucleotide polymorphism tagging set for human drug metabolism and transport. Nat Genet. 2005;37:84–89. [PubMed]
25. Villar J, Perez-Mendez L, Espinosa E, Flores C, Blanco J, et al. Serum lipopolysaccharide binding protein levels predict severity of lung injury and mortality in patients with severe sepsis. PLoS One. 2009;4:e6818. [PMC free article] [PubMed]
26. Wigginton JE, Cutler DJ, Abecasis GR. A note on exact tests of Hardy-Weinberg equilibrium. Am J Hum Genet. 2005;76:887–893. [PubMed]
27. Gonzalez JR, Armengol L, Sole X, Guino E, Mercader JM, et al. SNPassoc: an R package to perform whole genome association studies. Bioinformatics. 2007;23:644–645. [PubMed]
28. Wang K. Testing for genetic association in the presence of population stratification in genome-wide association studies. Genet Epidemiol. 2009;33:637–645. [PubMed]
29. Tregouet DA, Escolano S, Tiret L, Mallet A, Golmard JL. A new algorithm for haplotype-based association analysis: the Stochastic-EM algorithm. Ann Hum Genet. 2004;68:165–177. [PubMed]
30. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci U S A. 2003;100:9440–9445. [PubMed]
31. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. [PubMed]
32. Johnson CM, Lyle EA, Omueti KO, Stepensky VA, Yegin O, et al. Cutting edge: A common polymorphism impairs cell surface trafficking and functional responses of TLR1 but protects against leprosy. J Immunol. 2007;178:7520–7524. [PubMed]
33. Misch EA, Macdonald M, Ranjit C, Sapkota BR, Wells RD, et al. Human TLR1 deficiency is associated with impaired mycobacterial signaling and protection from leprosy reversal reaction. PLoS Negl Trop Dis. 2008;2:e231. [PMC free article] [PubMed]
34. Baudouin SV, Saunders D, Tiangyou W, Elson JL, Poynter J, et al. Mitochondrial DNA and survival after sepsis: a prospective study. Lancet. 2005;366:2118–2121. [PubMed]
35. Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med. 2006;34:344–353. [PubMed]
36. Ma X, Liu Y, Gowen BB, Graviss EA, Clark AG, et al. Full-exon resequencing reveals toll-like receptor variants contribute to human susceptibility to tuberculosis disease. PLoS One. 2007;2:e1318. [PMC free article] [PubMed]
37. Schuring RP, Hamann L, Faber WR, Pahan D, Richardus JH, et al. Polymorphism N248S in the human Toll-like receptor 1 gene is related to leprosy and leprosy reactions. J Infect Dis. 2009;199:1816–1819. [PubMed]
38. Hamann L, Bedu-Addo G, Eggelte TA, Schumann RR, Mockenhaupt FP. The toll-like receptor 1 variant S248N influences placental malaria. Infect Genet Evol. 2010;10:785–789. [PubMed]
39. Georgel P, Macquin C, Bahram S. The heterogeneous allelic repertoire of human toll-like receptor (TLR) genes. PLoS One. 2009;4:e7803. [PMC free article] [PubMed]

Articles from PLoS ONE are provided here courtesy of Public Library of Science