| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Heart Lung Centre Utrecht, St. Antonius Hospital, Departments of Pulmonology (Mr. Kruit, and Drs. Grutters and van den Bosch), and Clinical Chemistry (Drs. Ruven and van Moorsel), Nieuwegein, the Netherlands; and Institute of Clinical Chemistry and Pathobiochemistry (Dr. Weiskirchen and Ms. Mengsteab), RWTH-University Hospital, Aachen, Germany.
Correspondence to: Jules M. M. van den Bosch, MD, PhD, Department of Pulmonology, St. Antonius Hospital, Koekoekslaan 1, 3435 CM, Nieuwegein, the Netherlands; e-mail: j.vandenbosch{at}antonius.net
Abstract
Study objectives: Pulmonary fibrosis develops in approximately 25% of patients with chronic sarcoidosis. Transforming growth factor (TGF)-ß1 plays a central role in fibrosis, and accruing reports address the implication of TGF-ß2 and TGF-ß3 in this process. We determined whether single-nucleotide polymorphisms (SNPs) in the TGF-ß1, TGF-ß2, and TGF-ß3 genes might contribute to pulmonary fibrosis in sarcoidosis patients.
Setting: A hospital in the Netherlands.
Design: Five SNPs per TGF-ß gene were investigated.
Patients and control subjects: Patients with either acute/self-remitting sarcoidosis (n = 50) and Löfgren syndrome (n = 46) or chronic disease with fibrosis (n = 24) and without fibrosis (n = 34) were assessed over a 4-year follow-up period. The control subjects included 315 individuals.
Measurements and results: Polymorphism frequencies were not discordant between the patients and control subjects. The TGF-ß3 4875 A allele was significantly higher in fibrotic patients (carrier frequency, 0.29) than in patients with acute/self-remitting (0.06) and chronic (0.03) sarcoidosis combined (corrected p = 0.01; odds ratio [OR], 7.9). The TGF-ß3 17369 C allele carrier frequency was significantly higher in fibrotic patients (0.29) compared to acute/self-remitting (0.08) and chronic (0.06) patients combined (corrected p = 0.05; OR, 5.1). Although not significant after correction, the TGF-ß3 15101 G allele carrier frequency was lower in fibrotic patients (0.79) compared to acute/self-remitting (0.94) and chronic (1.00) patients combined (p = 0.02; corrected p = 0.1; OR, 0.15). The TGF-ß2 59941 G allele was more abundant in fibrotic patients (carrier frequency, 0.62) compared to patients with acute/self-remitting (0.41) and chronic sarcoidosis combined (0.28) [p = 0.04; corrected p = 0.2; OR, 2.9]. TGF-ß1 gene polymorphisms were not associated with fibrosis.
Conclusions: This study is the first to suggest the implication of genetic variation of TGF-ß3 in the predilection for pulmonary fibrosis developing in sarcoidosis patients.
Key Words: genetic predisposition • pulmonary fibrosis • single-nucleotide polymorphism • transforming growth factor-ß
Sarcoidosis is a systemic disease of unknown cause that is characterized by the presence of noncaseating granulomas in one or multiple organs.1 In approximately 90% of patients with sarcoidosis, the disease is manifested as pulmonary granulomas.2 Although parenchymal abnormalities often resolve spontaneously, pulmonary fibrosis will develop in approximately 20 to 25% of patients.3 Pulmonary fibrosis is marked by a disproportionate increase in extracellular matrix deposition produced by proliferating fibroblasts that reside in the lungs.45 Isoform 1 of the transforming growth factor (TGF)-ß family has been extensively scrutinized in fibrotic diseases such as sarcoidosis with radiographic stage IV67 and idiopathic pulmonary fibrosis.89
In contrast to their relative, isoforms TGF-ß2 and TGF-ß3 have yet to receive due attention. In the lung, the expression pattern of TGF-ß3 suggests that an imbalance between TGF-ß1 and TGF-ß3 is key in the development of fibrosis on tissue injury.101112
Polymorphisms in the genes encoding for all three TGF-ß isoforms have been identified and linked to variations of protein expression or functionality. Single-nucleotide polymorphisms (SNPs) in the TGF-ß1 gene, present in codon 10 (Leu10Pro) and codon 25 (Arg25Pro), are both associated with interindividual variation in TGF-ß1 production.13 These, as well as other polymorphisms of TGF-ß1, have been shown to confer an increased risk, eg, pulmonary fibrosis.1415
Both TGF-ß2 and TGF-ß3 gene polymorphisms, identified with microsatellite markers, have been associated with cutaneous fibrosis in systemic scleroderma.16 A genetic basis for the pathogenesis of this fibrotic disease is thus strongly supported.
In sarcoidosis, studies on gene polymorphisms of the TGF-ß family are scant. Both Leu10Pro and Arg25Pro polymorphisms in the TGF-ß1 gene have been evaluated for implication in sarcoidosis disease progression in Japanese17 and white patients.18 However, no associations were found between polymorphisms and disease progression, susceptibility, or severity in either study.
We hypothesize that TGF-ß plays an important role in the development of pulmonary fibrosis in sarcoidosis and that genetic variation of TGF-ß1, TGF-ß2, and TGF-ß3 underlies the propensity for fibrosis to develop in patients with sarcoidosis. Therefore, we sought to determine whether SNPs and constructed haplotypes of either isoform TGF-ß1, TGF-ß2, or TGF-ß3 might be differentially distributed in sarcoidosis patients with fibrosis, compared to those without fibrosis, as assessed by radiographic evolution over a 4-year follow-up period.
Methods and Materials
Subjects
One hundred forty-five unrelated, Dutch, white patients who received a diagnosis of sarcoidosis between 1965 and 1999 were included in this study (88 men and 66 women; mean ± SD age at diagnosis, 36.5 ± 10.4 years; range, 17 to 71 years). The diagnosis of sarcoidosis was established when clinical findings were supported by histologic evidence and after exclusion of other known causes of granulomatosis in accordance with the consensus of the American Thoracic Society/European Respiratory Society/World Association of Sarcoidosis and Other Granulomatous statement on sarcoidosis.19 In 46 patients, the diagnosis was made without biopsy proof because these patients presented with the classic symptoms of Löfgren syndrome, namely fever, erythema nodosum, arthralgia, and bilateral hilar lymphadenopathy. The medical ethical committee of the St. Antonius Hospital approved the study conducted, and all subjects gave formal written consent.
Chest radiographs were assessed in consensus by chest physicians specialized in diffuse lung diseases to determine disease severity using standard radiographic staging for sarcoidosis, classified according to the Scadding criteria.19 Radiographic data were available of 101 sarcoidosis patients. One patient presented with stage 0 (normal), 44 patients presented with stage I, 28 patients presented with stage II, 25 patients presented with stage III, and 3 patients presented with stage IV. Radiographic evolution over a 4-year follow-up period (presentation, 2 years, and 4 years) was categorized as follows: acute/self-remitting (normalization or improvement toward stage I, n = 50); chronic (persistent stage II/III or progression toward this stage, n = 34); and fibrosis (stable stage IV or progressive toward this stage, n = 24). Of the last group, seven patients with radiographic stage IV had sarcoidosis diagnosed outside of this hospital prior to referral. High-resolution CTs were available for 16 patients with stage IV sarcoidosis at 0 to 6 years following presentation. Of these patients, three main patterns of fibrosis (linear pattern, bronchial distortion, and honeycombing) were semiquantified according to a classification established by Abehsera et al.20 Namely, 11 patients had minimal-to-moderate linear patterns (concomitant moderate-to-severe bronchial distortion, n = 6; both severe linear patterns and severe bronchial distortion, n = 4; and severe honeycombing, a severe linear pattern, and severe bronchial distortion, n = 1).
Löfgren syndrome (radiographic stage I at presentation) was diagnosed in 46 patients. Of this group, 45 patients had received radiographic follow-up. These patients had normalized chest radiograph findings after 2 years or 4 years following presentation, except for one patient, who remained at radiographic stage I after 4 years. The diagnosis, rather than radiographic follow-up, was regarded as a category. Due to limitation of available sample material, four patients of the acute/self-remitting group, three patients of the chronic group, and two patients with Löfgren syndrome could not be genotyped for TGF-ß2. The control group included 315 healthy, unrelated, Dutch, white employees of St. Antonius Hospital in the Netherlands (118 men and 197 women; mean age, 39.0 ± 11.4 years; range, 18 to 68 years).
Genotyping
Biallelic SNPs were determined using sequence-specific primers and polymerase chain reaction (PCR). The identification numbers of the SNP loci and the sequences of SNP-specific primers with their complementary consensus primers are shown in Table 1
. The PCR conditions were as previously described.21 The final primer concentrations used were 7.6 ng/µL except for SNP rs1800472 and rs8179181, which were both 3.8 ng/µL. To control for erroneous genotyping, previously TGF-ß–genotyped samples were inserted blindly to the person performing the assay. No discrepancies were found between the genotypes of the analyzed and reanalyzed samples.
|
Pulmonary Function Tests
Pulmonary function tests were performed at presentation and at 2 years and 4 years following diagnosis. Inspiratory vital capacity (IVC), FEV1, and carbon monoxide diffusing lung capacity (DLCO) were used to assess the presence of lung function impairment at presentation and follow-up of disease. All lung function parameters are expressed as percentage of predicted values. IVC and FEV1 were calculated from volumes in liters and adjusted to body temperature, ambient pressure, and saturated with water vapor in accordance with the European Respiratory Society recommendations.23
Statistical Analysis
Statistical analysis of SNP and haplotype frequency distributions was performed using 
2 contingency table analysis with the appropriate number of degrees of freedom (df) [SPSS for Windows; SPSS; Chicago, IL]. A Fisher exact test was used if expected cell frequencies were < 5. Adjustment for multiple tests was made by multiplying the p value by the number of SNPs or haplotypes of each gene (Bonferroni method). Genotype frequencies were tested for Hardy-Weinberg equilibrium. Haplotypes were determined using Phase, version 2 (Mac OS X; Apple Computer; Cupertino, CA).2425 Statistical power was calculated with an on-line tool (case control for discrete traits test), available at http://pngu.mgh.harvard.edu/
purcell/gpc/.
Multivariate analysis, controlled for radiographic evolution, smoking, and treatment, was used to assess the influence of TGF-ß genotypes and haplotypes on lung function parameter changes over a 4-year follow-up period. A cut-off period of 5 years following cessation of smoking was considered a negative smoking history. Individuals who were treated with corticosteroids for at least 3 months during the 4-year follow-up period were deemed appropriate for treatment. No other treatment regimens had been used in the study group. All values of lung function parameters were log transformed. Statistical significance was denoted at p < 0.05 for all tests performed.
Results
The distributions of allele carrier and genotype frequencies of TGF-ß1, TGF-ß2, and TGF-ß3 in the sarcoidosis patients and healthy control subjects did not deviate from the Hardy-Weinberg equilibrium. Polymorphism frequencies were not discordant between the total patient population and control subjects (data not shown).
Polymorphism frequencies did not differ between patients who did or did not receive treatment, between smoking history, or between different radiographic stages at presentation (data not shown). The C allele carrier frequency of the TGF-ß1 28 T/C polymorphism was lower in Löfgren syndrome patients (0.52) compared to control subjects (0.69), which was ascribed to an increased TT genotype frequency (0.48) and decreased CT genotype frequency (0.35) compared to the control subjects (TT genotype, 0.31; CT genotype, 0.51; p = 0.05; corrected p = 0.25; df = 2). Patients with fibrosis showed a preponderance of the G allele (carrier frequency, 0.62) of the TGF-ß2 59941 A/G polymorphism compared to patients with acute/self-remitting (0.41) and chronic sarcoidosis (0.28) combined (p = 0.04; corrected p = 0.2; odds ratio [OR], 2.9; 95% confidence interval [CI], 1.1 to 7.4; Fig 1 ). Furthermore, control subjects had a TGF-ß2 59941 A/G genotype distribution, AG (carrier frequency, 0.33) and AA (0.64), that was different from that of patients with fibrosis, AG (0.58) and AA (0.38) [p = 0.03; corrected p = 0.15; df = 2].
|
|
Haplotypes were constructed for all TGF-ß genes from five polymorphic loci in each gene. The genotypes assigned to 7 TGF-ß1 and 13 TGF-ß2 haplotypes (data not shown). Six haplotypes were found for the TGF-ß3 gene (Table 3 ). The resulting haplotypes and their frequency distributions were evaluated in patients with classified radiographic evolutions and control subjects. All haplotype frequencies of TGF-ß1, TGF-ß2, and TGF-ß3 were equally distributed between sarcoidosis patients and control subjects, and between patient with different radiographic evolutions, treatment status, or smoking history (data not shown). Only TGF-ß3 haplotype 4 correlated with fibrosis in sarcoidosis patients (Table 3). This haplotype is different from the others by the 4875 G to A substitution and thus showed a similar frequency distribution as the individual 4875 G/A SNP.
|
|
In this study, sarcoidosis patients with pulmonary fibrosis revealed differential allele distributions of SNP loci in the TGF-ß3 gene compared to sarcoidosis patients without pulmonary fibrosis assessed over a 4-year follow-up period. Figure 1 shows the most noticeable results for the SNP allele carrier frequencies in sarcoidosis patients with and without pulmonary fibrosis. TGF-ß3 SNP 4875 G/A and 17369 T/C were the only polymorphisms that retained a p value < 0.05 after the Bonferroni correction was applied, suggesting a genetic influence of TGF-ß3 on the propensity for pulmonary fibrosis to develop in patients with sarcoidosis.
TGF-ß1, TGF-ß2, or TGF-ß3 gene polymorphisms do not appear to contribute to sarcoidosis disease susceptibility as evidenced by the similar polymorphism frequency distributions between patients and healthy control subjects. Functional polymorphisms with moderate-to-high heterozygosity have not yet been mapped for TGF-ß2. One TGF-ß2 SNP, present in exon 1, is known to result in an arginine to histidine amino acid substitution in codon 91.26 Our patient group was also genotyped for this polymorphism but revealed no variation (data not shown).
The TGF-ß2 59941 A/G gene polymorphism located within intron 2 showed a substantial increase of the G allele in the fibrotic patients compared to nonfibrotic sarcoidosis patients that was significant before correction (p = 0.04). This SNP deserves future investigations in association studies with additional patients with sarcoidosis or other fibrotic disorders. Previous reports16 have shown the genetic influence of TGF-ß2 and TGF-ß3 on fibrotic diseases such as scleroderma. Our present study shows that genetic variation of TGF-ß3 may contribute to the development of pulmonary fibrosis in sarcoidosis patients, at least according to radiographic data. With the exception of the TGF-ß3 17282 A/G polymorphism,27 none of the TGF-ß3 SNPs described in our study have been addressed in other association studies. Thus, speculations or conclusions pertaining to any phenotypic changes that may be caused by the 4875 G/A and/or 17369 T/C polymorphisms should be approached with care until functional data becomes available. Kim et al27 found a strong association between the SNP locus 17282 A/G and cleft palate syndrome in Koreans. This polymorphism was not associated with pulmonary fibrosis in our sarcoidosis group, which suggests that either this polymorphism has differential influences on different disease types or that it acts in function of racial background.
The observation that neither any of the fibrosis-associated SNPs nor haplotype 4 had any significant influence on lung function was not unexpected. It has been known that lung function does not always clearly correlate with radiographic staging.28 Although the fibrotic group as a whole had a significantly lower lung function compared to the nonfibrotic groups, a subset of patients within a single category, ie, fibrosis, may have normal IVC and DLCO values as the pattern of pulmonary fibrosis may vary.20 As a consequence, the distribution of the polymorphism frequency in the fibrotic group compared to the nonfibrotic group according to radiography may still poorly correlate with lung function parameters that are commonly associated with fibrosis.
The polymorphisms in the gene encoding TGF-ß1 did not reveal associations as seen with TGF-ß2 and TGF-ß3. All but one SNP (20743 C/T, located in the intron/exon boundary) are known to influence the expression levels of TGF-ß1.1329 Based on available literature describing the influence of individual SNPs on variation of expression levels, TGF-ß1 haplotypes that were constructed in this study would each have a net effect on the expression levels. None of the TGF-ß1 haplotypes could be ascribed to the different phenotypes of sarcoidosis. Thus, and in accordance with previous reports,1718 the absence of associations between haplotypes and disease phenotype implies that genetically controlled levels of TGF-ß1 neither predict the outcome of, nor contribute to the susceptibility to sarcoidosis. On a critical note, the lack of associations found between the genetic variation of TGF-ß1 and the development of fibrosis may not yet be conclusive. The statistical power was estimated to be 41% with a relative risk of 2, which may not have sufficed to identify the impact of TGF-ß1 polymorphisms on the development of pulmonary fibrosis. A substantially larger group (n = 62 to achieve 80% power) of sarcoidosis patients with pulmonary fibrosis is needed to confirm the negative findings.
It is speculated that the balance between all three TGF-ß isoforms may determine the nature of healing. Isoforms 1 and 2 of the TGF-ß family are generally described as having a profibrotic nature, while TGF-ß3 shows a more diverse character that may either sustain or resolve the progression of fibrosis.3031 The latter is also supported by the observation that the exogenous addition of TGF-ß3 can mitigate the deleterious effects of increased TGF-ß1 and thus prevent fibrosis in the injured lung.32 The outcome of sarcoidosis may be predestined by the genetics that control the immune response and the subsequent repair and remodeling of the injured tissue. Just as in those described in TGF-ß1, functional polymorphisms in TGF-ß3, which may exist in strong linkage with the fibrosis-associated SNPs described in this study, may lead to decreased expression of TGF-ß3. Consequently, this may cause a shift toward TGF-ß1 and possibly TGF-ß2 expression that may be unfavorable to the outcome of sarcoidosis. Such speculations can, however, only be confirmed by functional studies.
The results described in this present study indicate that genetic variation of TGF-ß3 is associated with pulmonary fibrosis that could be discerned by chest radiographs. Available high-resolution CT data were able to confirm fibrotic changes present as three main patterns of fibrosis, according to Abehsera et al.20 Although an association between these different HRCT patterns and TGF-ß3 polymorphisms could not be identified at this point due to a limited sample size, future studies using a larger group of fibrotic patients may lead to additional associations between TGF-ß polymorphisms and different fibrotic patterns.
Five patients who had presented with stage II and nine patients who had presented with stage III progressed toward stage IV 4 years later. A recent study by Akira et al33 showed that HRCT scans failed to discriminate between ground-glass opacities that had either disappeared or evolved into honeycombing of the lung after 7.4 (mean) years. The associations between TGF-ß3 gene variations and fibrosis would not have been found had we not followed up on these patients. Thus, a follow-up period of a significant number of years proves to be imperative in order to establish a category of patients with apparent pulmonary fibrosis for genetic association studies.
Conclusion
Although efforts have been made to find associations between TGF-ß1 polymorphisms and sarcoidosis, the results of this study are the first to suggest the implication of genetic variation of TGF-ß3 in the predilection for pulmonary fibrosis developing in patients with sarcoidosis. Additional studies in other sarcoidosis populations, preferentially in different ethnic groups, are recommended to confirm our findings.
Acknowledgements
The authors thank Natalie Pot and Jan Broess for technical assistance, and Kenneth I. Welsh, Roland M. du Bois, and Hiroe Sato for scientific advice.
Footnotes
Abbreviations: df = degrees of freedom; DLCO = carbon monoxide diffusing lung capacity; IVC = inspiratory vital capacity; OR = odds ratio; PCR = polymerase chain reaction; SNP = single-nucleotide polymorphism; TGF = transforming growth factor
This work was performed at the Heart Lung Centre Utrecht, St. Antonius Hospital, Department of Pulmonology, Nieuwegein, the Netherlands.
Received for publication October 4, 2005. Accepted for publication December 19, 2005.
References
and epidermal growth factor-receptor in idiopathic pulmonary fibrosis. Sarcoidosis Vasc Diffuse Lung Dis 1999;16,57-61[ISI][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
