Toll-like receptor (TLR) 4 polymorphism Asp299Gly is not associated with disease course in Dutch sarcoidosis patients

Summary

The aetiology of sarcoidosis, a systemic disorder characterized by the formation of non-caseating granulomas in variable organs, remains enigmatic. Clarification is hampered by heterogeneity in disease phenotypes and course, due partly to the influence of a variety of genetic and environmental factors. Multiple studies have pointed towards bacteria as possible causative agents. Toll-like receptors (TLR) are innate immunity receptors important in the immune response against pathogens. TLR-4, together with CD14 and MD-2, is an essential receptor for the recognition of lipopolysaccharide (LPS), unique to the cell wall of Gram-negative bacteria. Recently, an association between TLR-4 polymorphism Asp299Gly, leading to a change in the extracellular domain of the receptor and possible hyporesponsiveness to LPS, and a chronic course of sarcoidosis was found in German patients. In the present study this polymorphism was genotyped in 156 Dutch sarcoidosis patients and 200 healthy Dutch controls using dual-labelled fluorescent oligonucleotides. No differences were found in allelic distributions between patients and controls (P = 0·79) or within the different clinical entities of the sarcoidosis group (P = 0·44). Importantly, there were no differences between the Dutch and German sarcoidosis patients (P = 0·62). However, the allelic distribution of the Asp299Gly polymorphism differed significantly between both control groups (P = 0·04). This study highlights the importance of testing a reported gene association in a distinct population when performing genetic association studies.

Introduction

Sarcoidosis is a systemic disorder characterized by the formation of non-caseating granulomas in variable organs, usually the lungs. In half the cases sarcoidosis follows a benign course, with spontaneous resolution within 12–36 months [1]. However, the disease can be severe due to localization of granulomas in particularly vulnerable organs such as the central nervous system or heart, or as a consequence of the abundance of granulomas in the lungs leading to respiratory insufficiency. Mortality attributable to sarcoidosis is estimated between 0·5 and 5% [2], which is clinically relevant for a disease affecting relatively young people.

The aetiology of sarcoidosis remains unknown, which has been attributed to its heterogeneity in clinical phenotypes that might reflect a collection of different granulomatous diseases, each with its own aetiological factor, rather than a single disease. Sarcoidosis is characterized by a strong cell-mediated immune reaction, which is essential for combating viruses or intracellular bacteria. The possible role of bacteria in sarcoidosis has therefore been studied extensively. In particular, Gram-positive and intracellular bacteria, such as mycobacteria and propionibacteria, have been suggested to play a role in the aetiology of sarcoidosis [3,4]. There is no convincing evidence, however, on a possible role for Gram-negative bacteria in sarcoidosis [5].

Toll-like receptors (TLRs) are innate immunity receptors responsible for the molecular recognition of pathogens. TLRs can initiate inflammatory and anti-microbial innate immune response, thereby dictating the ensuing adaptive immune response. TLR-2 and TLR-4 are the most studied among the TLR family. TLR-2 recognizes a variety of microbial components, such as lipoproteins, peptidoglycan and lipoteichoic acid from Gram-positive bacteria and lipoarabinomannan from mycobacteria, zymosan from fungi and some glycolipids [6]. TLR-4 is an essential receptor for the recognition of lipopolysaccharide (LPS), unique to the cell wall of Gram-negative bacteria [6].

Thus, when considering a causative role of bacteria in the pathogenesis of sarcoidosis, TLRs recognizing microbial components of Gram-positive bacteria, and not TLR-4, might be good candidates for genetic association studies in sarcoidosis.

Surprisingly, Pabst et al. recently published an article in this journal where they demonstrate an association between the TLR-4 polymorphisms Asp299Gly and Thr399Ile and the chronic course of sarcoidosis [7]. These results could indicate that Gram-negative bacteria might nevertheless be candidates for a bacterial agent in sarcoidosis.

In this study, we tried to confirm the association between the TLR-4 polymorphisms and sarcoidosis in Dutch patients. Because Asp299Gly and Thr399Ile alleles co-segregate, and a functional alteration of the receptor was proved only for the Asp299Gly mutation [8], we examined the functional polymorphism in our cohort.

Materials and methods

Sarcoidosis and control group

A total of 156 unrelated and randomly selected Dutch white patients with sarcoidosis (88 men and 67 women) were included in the study. In 112 patients, the diagnosis of sarcoidosis was established when clinical findings were supported by histological evidence, and after the exclusion of other known causes of granulomatosis. Forty-four patients presented with the classical Löfgren’s syndrome of fever, erythema nodosum, bilateral hilar lymphadenopathy and arthritis. The diagnosis in these patients was made mainly without biopsy proof [9]. Verbal and written consent was obtained from all subjects, and authorization was given by the Ethics Committee of the St Antonius Hospital, Nieuwegein, the Netherlands. The control subjects comprised 200 healthy Dutch Caucasian employees of the St Antonius Hospital (118 men and 82 women). By completing a questionnaire, relevant background information was provided by these volunteers and included medication, ethnicity and hereditary diseases.

Evaluation of pulmonary disease severity

Pulmonary disease severity at presentation was evaluated by chest radiography. Further chest radiographs for each patient were examined and compared to determine disease outcome. Chest radiographs at presentation, 2 years and 4 years were collected for each patient and assessed blind by a pulmonary physician for disease severity using standard radiographic staging for sarcoidosis. In brief, this comprises five stages: stage 0, normal; stage I, bilateral hilar lymphadenopathy (BHL); stage II, BHL and parenchymal infiltration; stage III, parenchymal infiltration without BHL; and stage IV, irreversible fibrosis with loss of lung volume. Radiographic evolution over a 4-year period was available for all 156 patients. In the statistical analysis acute sarcoidosis was defined as Löfgren’s syndrome or a normalized chest X-ray 2 years after presentation. Chronic sarcoidosis was defined as a disease course over at least 2 years.

Analysis of the TLR-4 Asp299Gly polymorphism

Genomic DNA from all subjects and controls, extracted from peripheral blood cells, was genotyped for the Asp299Gly polymorphism using single tube polymerase chain reaction (PCR) and exonuclease degradation of dual-labelled fluorescent oligonucleotides, as described elsewhere [10].

Data analysis

Statistical analysis was performed using χ² contingency table analysis with the appropriate number of degrees of freedom. A P-value of < 0·05 was considered significant.

Results

The genotype of the investigated Asp299Gly polymorphism in Dutch sarcoidosis and control populations are presented, together with the data of the German sarcoidosis study by Pabst et al. [7], in Table 1. Genotype data from all populations were in Hardy–Weinberg equilibrium. Among 200 healthy Dutch controls, 23 were heterozygous for Asp299Gly and one was homozygous for the minor allele (minor allele frequency 6·3%; 25/400, Table 1). Among the sarcoidosis patients, 21 were heterozygous and no homozygous mutation was found (minor allele frequency 6·7%; 21/312, Table 1). No differences were found in allelic distributions between patients and controls (P = 0·79) or when comparing the different clinical entities within the sarcoidosis group (P = 0·44). Also, when comparing the Dutch sarcoidosis patients with the German sarcoidosis patients, no difference was found in allelic distributions (P = 0·62). However, Asp299Gly polymorphisms occurred more frequently in Dutch controls compared to German controls (P = 0·04).

Table 1.

Prevalence of different TLR-4 Asp299Gly genotypes in sarcoidosis patients and controls. The results of the German population are extracted from the paper published by Pabst et al. [7].

	No.	AA	AG	GG	G allele frequency (%)
Sarcoidosis (all patients)
Dutch	156	135	21	0	6·7
German	141	119	22	0	7·8
Acute sarcoidosis
Dutch	62	52	10	0	8·0
German	33	31	2	0	3·0
Chronic sarcoidosis
Dutch	94	83	11	0	5·9
German	108	88	20	0	9·3
Controls
Dutch	200	176	23	1	6·3
German	141	133	8	0	2·8

Taken together, we did not find an association between the prevalence of the Asp299Gly polymorphism and sarcoidosis in a Dutch population.

Discussion

TLR-4 is an innate immunity pattern recognition receptor expressed on a variety of human cells, such as monocytes, mast cells, neutrophils, dendritic cells, T cells and endothelial cells. TLR-4 is an essential receptor for the recognition of LPS, unique to the cell wall of Gram-negative bacteria. However, the main LPS binding receptor is CD14, also a pattern recognition receptor [11]. TLR-4 acts as a co-receptor for CD14, together with MD-2, and is responsible for activating intracellular signalling pathways resulting in the production of proinflammatory cytokines and up-regulation of co-stimulatory molecules, thereby leading to priming of an adaptive immune response [12].

In the present study, we tried to confirm a recent report of an association between the TLR-4 polymorphism Asp299Gly and severity of sarcoidosis in a clinically well-defined population from the Netherlands. At the genomic location of this polymorphism, an A→G substitution leads to the replacement of aspartic acid with glycine at amino acid 299, located in the extracellular domain of TLR-4. This substitution has been associated with LPS hyporesponsiveness in humans [8]; however, some strong conflicting studies have been published [13,14].

We did not find an association between the allelic distribution of the Asp299Gly polymorphism in sarcoidosis and controls, nor between different clinical subsets of disease severity. Comparing the data of multiple studies regarding the role of bacteria in initiating the immune response in saroidosis, it is to be expected that a pattern recognition receptor involved in the immune response against Gram-negative bacteria will probably not play a role. In our opinion, these results are therefore in line with current concepts of the role of bacteria in the aetiopathogenesis of sarcoidosis.

However, TLR-4 has also been shown to be involved in the recognition of endogenous ligands such as heat shock proteins (HSP60 and HSP70) [6] and cross-reactivity with mycobacterial HSPs is suggested as a disease mechanism in sarcoidosis [15]. Therefore, a possible role for TLR-4 in this disease can still not be excluded fully.

Importantly, this study again demonstrates the importance in studying multiple populations in the search for genetic variants in sarcoidosis. Pabst et al. [7] found a significant association between patients with a chronic course of sarcoidosis and the Asp299Gly polymorphism, while this association could not be replicated in another white European sarcoidosis cohort. The allele frequency of this polymorphism was almost the same in Dutch and German sarcoidosis patients. However, a remarkable and significant difference was found when comparing the allele frequency in both healthy control populations. The minor allele of the Asp299Gly polymorphism occurred more frequently in Dutch controls compared to German controls. The discrepancy between these control groups is striking.

In contrast with the restricted fragment length polymorphism (RFLP) analysis [16] used by Pabst and colleagues, in the present study the Aps299Gly polymorphism was determined with dual-labelled allele-specific oligonucleotides. Van Rijn et al. [10] compared both techniques and demonstrated that they reveal identical Asp299Gly genotyping results, indicating that this cannot explain the observed differences between the Dutch and German controls. Furthermore, it is well known that allele frequencies of single nucleotide polymorphisms, such as Asp299Gly, vary between different ethnic populations. The TLR-4 Asp299Gly polymorphism has different minor allele frequencies in, for example, Chinese (0%) [17], American (6·4%) [18] and West African populations (9·8%) [19]. An explanation for the discrepancy therefore could be, although it is not probable, that there is a difference in Asp299Gly minor allele carriership between the Netherlands and Germany. We searched for recent publications with control groups of over 150 subjects in order to compare Dutch and German Asp299Gly minor allele frequencies (Table 2). No differences in allele frequencies were seen between healthy controls from the Netherlands and Germany (5·9% versus 5·4%, respectively). Furthermore, our results on the Asp299Gly minor allele frequency are comparable with other healthy Dutch controls (Table 2). The Asp299Gly minor allele frequency found by Pabst et al. seems, as they mention in their article, slightly lower than other healthy German controls (Table 2). Therefore, the calculation of allele frequency using relatively small numbers seems susceptible for under- or overestimation, stressing the importance of being reserved when drawing conclusions.

Table 2.

Minor allele frequency of the Asp299Gly polymorphism in different Dutch and German control groups from several published studies.

Study	Country	Control population no.	Allele frequency (%)
Present study	Netherlands	200	6·3
Netea et al. [21]	Netherlands	200	5·5
Hawn et al. [22]	Netherlands	495	6·5
Ouburg et al. [23]	Netherlands	170	5·3
Van der Paardt et al. [24]	Netherlands	169	4·7
Pabst et al. [7]	Germany	141	2·8
Kroner et al. [25]	Germany	350	4·4
Reismann et al. [26]	Germany	343	5·1
Brand et al. [27]	Germany	199	3·8
Rohde et al. [28]	Germany	444	7·8

In conclusion, the TLR-4 Asp299Gly is not associated with sarcoidosis or disease course in Dutch subjects. Therefore, it is unlikely that TLR-4 genetic variants play a major role in sarcoidosis pathogenesis. However, before excluding a role for this receptor, functional data, such as influences of this or other TLR-4 polymorphisms on TLR-4 cell surface expression and interaction with pathogens, are necessary. Also, the influence of CD14 and MD-2 should be studied. Importantly, this study highlights one of the most critical aspects of genetic association studies, i.e. subsequent testing of a reported gene association in a distinct population, as was also pointed out clearly by Colhoun and colleagues [20].