INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: tumour necrosis factor-; cytokine; polymorphism (genetics); sarcoidosis

Sarcoidosis is a complex disease and thought to be the product of genetic susceptibility and an unknown antigenic stimulus from the environment (1-3). The main defining pathologic feature of this disease is a chronic inflammation resulting in granuloma formation. Multiple organs can be affected, but the lungs and lymph nodes are among those most commonly involved. The inflammatory response in sarcoidosis is characterized by the production of increased amounts of several proinflammatory cytokines at sites of disease, including tumor necrosis factor (TNF)- and interleukin-1 (4). Normally cytokine production is regulated via the action of opposing cytokines, the release of soluble cytokine receptors, and the production of cytokine receptor antagonists. There is growing evidence for the contribution of genetic polymorphisms to interindividual differences in the regulatory mechanisms of cytokine production. Therefore, particular variant cytokine genotypes might contribute to the predisposition to sarcoidosis, exacerbate granuloma formation, or modulate disease severity.

TNF- is thought to be a principal cytokine in the pathogenesis of sarcoidosis, due to its pivotal role in granuloma formation (5-7). Increased amounts have been found at sites of disease activity (8-10). There is evidence that polymorphisms in the promoter region of the TNF- gene affect the amount of TNF- production, resulting in high TNF- producers and low producers (11-13). Therefore, the TNF gene is an important candidate in the search for genetic variations predisposing to sarcoidosis.

Wilson and colleagues were the first to describe a single base pair transition polymorphism (nucleotide G to A at position 307 [originally misnumbered as 308]) in the promoter region of the human TNF gene (14, 15). Reporter gene assays have suggested a small but significant effect of this polymorphism on TNF- transcription, with the rarer 307A (also called TNF-2) allele being associated with slightly higher levels of gene transcription (16, 17). The relation between this polymorphism and susceptibility to major autoimmune disease has been widely studied, but in general no clear influence on disease susceptibility was demonstrated (13). However, numerous reports have indicated a relationship between the 307 genotype and severity of infectious diseases (13). It is notable that certain infections, characterized by the presence of intracellular bacteria (leprosy, leishmania, and chlamydia), seem to be influenced by this polymorphism (18-20).

In sarcoidosis, a higher frequency of the TNF 307A allele has been found in patients presenting with the Löfgren's syndrome, but no increased frequency was found when comparing the sarcoidosis group as a whole with unaffected control subjects (21, 22). The association between the TNF 307A allele and Löfgren's syndrome was recently confirmed by Labunski and colleagues, who reported an association between this allele and sarcoidosis-associated erythema nodosum, a clinical feature of Löfgren's syndrome (23).

One of the more recently discovered polymorphisms in the promoter region of the TNF gene is a change from C to T at position 857 (14). The transcriptional promoter activity of the rarer TNF 857T allele was shown to be higher than that of the common allele in activated blood mononuclear cells from Japanese donors (24). Another TNF promoter polymorphism, further upstream at position 863, influences the binding of the nuclear factor (NF)-B p50-p50 homodimer. This NF-B dimer acts as a transcriptional repressor after binding to its DNA binding domain in the TNF promoter. The binding of the NF-B p50-p50 homodimer to its domain, however, is significantly inhibited by the A-variant of the TNF 863 promoter polymorphism. The result is an insufficient down-regulation of TNF expression, leading to increased TNF- production in cell models (11).

The aim of the present study was, therefore, to investigate the potentially functional TNF promoter polymorphisms at nucleotide positions 307, 857, and 863, in two clinically well-defined groups of patients with sarcoidosis from different countries, to determine the association between these polymorphisms and sarcoidosis. To extend the mapping of the TNF promoter region, two other previously described TNF promoter polymorphisms were included, one at position 1,031 and one at position 237 (13, 25).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

British Patients and Control Subjects

Ninety-six unrelated British white sarcoid patients were investigated. In all patients, the diagnosis of sarcoidosis was established when clinico-radiologic findings were supported by histologic evidence of noncaseating epithelioid cell granulomas. Verbal and written patient consent was obtained from all subjects, and the Ethics Committee of the Royal Brompton Hospital, London, UK, gave authorization. This hospital is a tertiary referral center taking patients mainly from the southeast of the UK.

The UK control population was comprised of 354 white subjects, again mainly collected from the southeast of the UK. All had been checked for health (including medical history, physical examination, and routine laboratory blood testing) at regular intervals during a 10-year period before taking blood for DNA extraction and gave their written consent.

Dutch Patients and Control Subjects

One hundred unrelated Dutch white sarcoid patients were included in the study. In 95 patients, the diagnosis of sarcoidosis was established when clinical findings were supported by histologic evidence, and after exclusion of other known causes of granulomatosis. In five patients, the diagnosis was made without biopsy proof because these patients presented with the classic Löfgren's syndrome of fever, erythema nodosum, arthralgia, and bilateral hilar lymphadenopathy (26). Verbal and written patient consent was obtained from all subjects and authorization was given by the Ethics Committee of the Sint Antonius Hospital, Nieuwegein (region Utrecht).

The Dutch control group comprised 222 white donors from the Blood Transfusion Service in Utrecht, which takes donors mainly from the region Utrecht. All donors were routinely checked for health before donation and gave their written consent.

Sequence-Specific Primers and Polymerase Chain Reaction

Polymorphisms were determined using sequence-specific primers (SSPs) and polymerase chain reaction (PCR) that utilizes SSPs with 3'-end mismatches and identifies the presence of specific allelic variants through PCR amplification. A total of five biallelic TNF promoter single nucleotide polymorphisms were identified: 1,031(T/C), 863(C/A), 857(C/T), 307(G/A), 237(G/A). Additional information on the nomenclature is given in the online data supplement. For identification of the polymorphism at position 1,031, we used the sequence-specific forward primers 5'-CAAAGGAGAAGCTGAGAAG AT and 5'-CAAAGGAGAAGCTGAGAAGAC in combination with the consensus reverse primer 5'-CCGGGAATTCACAGACCCC at a final concentration of 20 ng/µl, with an expected PCR product size of 433 bp. The polymorphism at position 863 was identified using the sequence-specific forward primers 5'-CGAGTATGGGGACCCC CC and 5'-GAGTATGGGGACCCCCA at a final concentration of 20 ng/µl. For the identification of this polymorphism, the same consensus primer was used as for position 1,031, leading to an expected PCR product size of 263 bp. The promoter polymorphism 857(C/T) was identified with the sequence-specific reverse primers 5'-CTA CATGGCCCTGTCTTCG and 5'-TCTACATGGCCCTGTCTTCA in combination with the consensus forward primer 5'-AAGGAT AAGGGCTCAGAGAG at a final concentration of 10 ng/µl, with an expected PCR product size of 270 bp. In all primer mixes, we included the control primers 5'-TGCCAAGTGGAGCACCCAA and 5'-GCA TCTTGCTCTGTGCAGAT at a final concentration of 1.6 ng/µl. Finally, for identification of the polymorphisms at nucleotide positions 307 and 237, the primer sequences (with minor modifications) and primer mixtures were used previously described by Fanning and colleagues (27).

PCR Conditions

All PCR reactions were run under identical conditions and as previously described, in a final volume of 13 µl overlaid with 10 µl of mineral oil (28). Each reaction mixture consisted of 5 µl of the appropriate primer mix and 8 µl of PCR reaction mixture (the final concentration of the PCR reaction mixture was ×1 PCR buffer [Bioline, London, UK], 160 µM of each deoxynucleotide triphosphate [Bioline], 2 mM MgCl₂, 0.3 U Taq polymerase [Bioline], and 0.08 µg DNA per well in 96-well plates). PCR amplifications were done in an MJ Research (Waltham, MA) PTC-200 machine. The cycling parameters for the 13 µl reactions were 96° C for 1 minute, followed by five cycles of 96° C for 25 seconds, 70° C for 45 seconds, and 72° C for 25 seconds; 21 cycles of 96° C for 25 seconds, 65° C for 50 seconds, 72° C for 30 seconds; and four cycles of 96° C for 30 seconds, 55° C for 60 seconds, and 72° C for 90 seconds. To the completed PCR reaction, we added 8 µl of Orange G loading buffer and loaded the entire product onto a 2% agarose-×0.5 Tris-borate-ethylenediamine tetra-acetic acid gel containing 0.14 µg/ml ethidium bromide. Electrophoresis was performed for 20 minutes at 200 V/cm², and the gel was photographed under ultraviolet light (320 nm). The presence of an allele-specific band of the expected size, in conjunction with a control band, was considered to be positive evidence for each particular allele. The absence of an allele-specific band and the presence of a control band were considered to be evidence for the absence of an allele.

Data Analysis

The genotype frequencies, phenotype frequencies (i.e., number of individuals carrying the allele either in both [homozygous] or in only one [heterozygous] chromosome), and the frequency of an allele in the chromosomal pool of each population (allele frequency) were determined by direct counting for both control and sarcoidosis groups. All genotype frequencies were tested for Hardy-Weinberg equilibrium. Haplotypes were identified using the estimate haplotype frequencies program (EH; http://linkage.rockefeller.edu/ott/eh.htm). Subsequently, the carrier frequency of each haplotype was determined by direct counting.

Statistical analysis was performed using chi-square contingency table analysis with the appropriate number of degrees of freedom (df; SPSS for Windows; SPSS Inc., Chicago, IL). Fisher exact test was used if expected cell frequencies were lower than 5. Adjustment for multiple tests was made using the formula p_c = p × n, where p_c is the corrected value, p the uncorrected value, and n the number of tests performed (Bonferonni method). A value of p < 0.05 was considered significant.

The population attributable risk percentage (PAR%) was defined as the excess rate of disease in individuals with a mutation compared with those without. This was estimated by the method of Schlesselman, and to calculate this, the prevalence of sarcoidosis was estimated at 10-40/100,000 (in white individuals from both the UK and The Netherlands) and the frequency of the mutation in the control population assumed to reflect that of the general population (29, 30).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 summarizes the allele frequencies of the investigated TNF promoter polymorphisms in the British sarcoid and control population, and Table 2 summarizes the results for the Dutch subjects. All populations were in Hardy-Weinberg equilibrium for all genotype frequencies. In both the British and the Dutch patient population, we observed a consistent increase in the uncommon TNF 857T allele, with increasing level of significance when adding the two patient populations together (² = 13.27 with 1 df, p = 0.0003 [p_c = 0.002]; Table 3). The PAR% of this polymorphism was estimated 13.3. 

British Population Results

For the TNF promoter polymorphism at position 857, we observed a significant increase in the rarer TNF 857T allele frequency in the British sarcoid patients compared with the British control subjects (² = 7.39 with 1 df, p = 0.007 [p_c = 0.04]). A significant difference was also found in the genotype and phenotype frequency of this polymorphism (² = 10.41 with 2 df, p = 0.006 [p_c = 0.03] and ² = 4.73 with 1 df, p = 0.03 [p_c = 0.15], respectively; data not shown). In the British sarcoidosis group, 26.0% of the individuals carried the rarer TNF 857T allele, compared with 14.1% in the control group (Figure 1).

View larger version (37K): [in this window] [in a new window]   Figure 1.   TNF 857T phenotype frequency in the British (UK) sarcoidosis and control population, the Dutch (NL) sarcoidosis and control population, and all sarcoidosis patients and control subjects. *p = 0.003, pc = 0.02 (df = 1).

The other investigated TNF promoter polymorphisms, at nucleotide positions 1,031, 863, 307, and 237, revealed no differences in genotype, phenotype, and allele frequency between British patients and control subjects.

Dutch Population Results

Similar to our findings in the British population, in the Dutch sarcoidosis population a higher frequency of the rarer TNF 857T allele was found in comparison with Dutch control subjects (² = 4.83 with 1 df, p = 0.03 [p_c = 0.15]; Table 2). As in the British sarcoid population, the same trend in genotype and phenotype frequency of this polymorphism was observed. In the Dutch sarcoidosis group 25.0% of the individuals carried the rarer TNF 857T allele, compared with 14.0% in the control group (² = 3.46 with 1 df, p = 0.06; Figure 1).

The TNF gene promoter polymorphisms at nucleotide positions 1,031, 863, 307, and 237 showed the same genotype, phenotype, and allele frequency in Dutch sarcoid patients as in control subjects.

TNF Promoter Polymorphisms in Löfgren's Syndrome

Eleven of the Dutch and four of the British sarcoid patients presented with a classic Löfgren's syndrome. We assessed whether there were differences in allele frequencies of the studied TNF promoter polymorphisms between patients with Löfgren's syndrome (n = 15) and the other patients with sarcoidosis (n = 181). As reported in previous studies, we found an increased frequency of the rarer TNF 307A allele in Löfgren compared with non-Löfgren patients (² = 7.46 with 1 df, p = 0.006 [p_c = 0.03]; Table 4). Remarkably, the allele frequency of the TNF 857T allele tended to be lower in Löfgren patients (6.7%, versus 14.1% in the non-Löfgren sarcoid patients); however, this difference did not reach statistical significance. The allele frequencies of the other investigated TNF promoter polymorphisms did not differ between the Löfgren and non-Löfgren patients and are summarized in Table 4.

Construction of Haplotypes

From the investigated TNF promoter polymorphisms we were able to deduce six haplotypes, which are shown in Table 5. Haplotype 4, including T at nucleotide position 857, was significantly over-represented in the sarcoidosis group, whereas no differences were found in the frequencies of the other haplotypes: 25.5% of the sarcoid patients carried haplotype 4 versus 14.1% of the controls (RR 2.1, 95% CI 1.4  3.1, p = 0.0002 [p_c = 0.001]; Table 5).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we have investigated the association between five potentially functional TNF promoter polymorphisms and sarcoidosis in two clinically well-defined groups of subjects from the UK and The Netherlands. We detected a significant increase in the allele frequency of the rarer TNF 857T allele in patients with sarcoidosis by comparison with ethnically and geographically matched controls. Further, on constructing TNF promoter haplotypes, only one haplotype, including the T at nucleotide position 857 (haplotype 4), showed a clear association with sarcoidosis. The fact that this result was found in two different populations adds weight to the relevance of the finding and points toward a novel susceptibility marker for sarcoidosis or even causality between the TNF gene and this disease.

It is possible that the TNF 857T allele is a marker for the known HLA association in sarcoidosis. In any area of tight linkage disequilibrium (LD) such as the MHC, however, it is difficult to distinguish the primary candidate. In addition there is, as yet, no published linkage disequilibrium data for haplotypes bearing TNF 857T. Nevertheless, from our results, we could deduce TNF promoter haplotypes, which included positions 307 and 237, positions for which there are published HLA class I and HLA class II data. With this approach, we can conclude that HLA-DR11 and 14, associated with sarcoidosis in certain populations, is likely to be in LD with TNF 857T, but we cannot determine which of these is dominant. This will require the typing of a much larger European sarcoidosis cohort and/or trans-racial gene mapping.

As the TNF gene is a strong candidate gene for sarcoidosis, it is tempting to speculate on the role of the TNF 857(C/T) promoter polymorphism in the pathogenesis of sarcoidosis. Promoter polymorphisms of the TNF gene are of great immunogenetic interest, as there is in vitro evidence that they may account for interindividual variations in TNF- production in the immune response (11-13). These variations are thought to be subtle in physiologic states, but under the influence of certain pathogenic stimuli might contribute to inflammatory diseases. One of the most thoroughly investigated TNF promoter polymorphisms is the change from C to A at nucleotide position 863 (11). Udalova and colleagues demonstrated a clear effect of this nucleotide change on the relative binding affinities of different forms of the NF-B complex (11). It was shown that the p50-p50 homodimeric form of this complex had a significantly decreased affinity to its DNA binding site for 863A. As the p50-p50 homodimer acts as a transcriptional repressor on binding to its regulatory site in the promoter region of the TNF gene, decreased binding is thought to result in an inadequate down-regulation of TNF gene expression, and thus increased TNF- production (11). It is possible that the change from C to T at nucleotide position 857 might have similar consequences regarding the affinity of other nuclear transcription regulatory factors, but there are no data available on this at this time.

Another interesting result of this study was the significant difference in allele frequency of the TNF promoter polymorphism at position 307 (G to A transition) in a subgroup of patients with sarcoidosis, i.e., patients presenting with classic Löfgren's syndrome. As previously described, our study showed a significant increase in the rarer TNF 307A allele in patients presenting with this syndrome compared with the other group of patients (22). However, no increase in this allele was found in the sarcoidosis group as a whole compared with the control population. This suggests that this polymorphism might play a role only in this particular subgroup of sarcoidosis. In addition, we found a lower frequency of the rarer TNF 857T in Löfgren patients compared with the non-Löfgren sarcoid patients, although this difference did not reach statistical significance. Power calculation showed that 43 Löfgren patients (80% statistical power, 95% CI) would be needed to test if they indeed have a significantly lower frequency of the rarer 857T allele compared with non-Löfgren sarcoid patients. Therefore, we hypothesize that in different sarcoidosis subgroups different TNF promoter polymorphisms might play a role, either as susceptibility genes or as linkage markers for other genes that affect susceptibility.

The observed associations between Löfgren's syndrome and the TNF 307 (G/A) polymorphism might also be caused by linkage disequilibrium (LD). Sarcoidosis is known to be associated with the presence of the HLA-DRB1*0301 allele (DR17) in white patients (31). As previously described, DR3 alleles (including DR17) are in tight LD with the TNF 307A (TNF-2) allele (32). Tight LD between the HLA-DRB1 and TNF loci make a determination of the relative roles of each gene in the immunogenetics of sarcoidosis extremely difficult. More high resolution HLA typing data in combination with further fine mapping data of the TNF gene are needed to resolve these problems and to elucidate the genetic basis of the Löfgren's syndrome in patients with sarcoidosis.

The present study has also demonstrated the importance of studying multiple distinct geographic populations together. The fact that the same relationship between specific TNF promoter polymorphisms and sarcoidosis was found in two groups of patients with a different geographic background, strengthened the analysis and added weight to the interest of the results. As sarcoidosis is a disease with variable prevalence, incidence, and severity in different races and ethnic groups, comparison of genetic results can be very difficult. This phenomenon is well recognized in studies of the MHC. Because of the increasing interest in the genetic basis of complex diseases such as sarcoidosis and the expanding possibilities for extensive mapping of candidate genes and performing chromosomal genomic screens, these problems are also likely to arise in research on other genes. Studying multiple ethnic populations simultaneously might provide a powerful tool in overcoming some of these difficulties.

In conclusion, this is the first study of multiple TNF promoter polymorphisms in sarcoidosis in two distinct populations. In both British and Dutch patients with sarcoidosis, we demonstrated a clearly significant increase in the rarer TNF 857T allele in comparison with the control groups. Furthermore, a significant increase in the rarer TNF 307A allele was observed in a subgroup of sarcoid patients, i.e., patients presenting with Löfgren's syndrome. The results on the TNF 857(C/T) polymorphism indicate a novel susceptibility marker for sarcoidosis or even a possible causal gene for the disease, as this polymorphism might be relevant for the transcription of the gene. Further studies on the functionality of this polymorphism are necessary, as well as high resolution HLA-class II typing in combination with further fine mapping across the MHC region to clarify the eventual role of the TNF gene in the immunogenetics of sarcoidosis.