INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sarcoidosis is a chronic, multisystem disorder of unknown origin, characterized by accumulation of T lymphocytes and mononuclear phagocytes, noncaseating granulomas, and derangement of the normal tissue architecture. All systems can be affected, but the organ most frequently affected is the lung (1, 2). A variety of infectious and noninfectious agents have been implicated to play a role in the etiology of the disease, but there is no proof that any specific agent is responsible (1).

Unlike many diseases in which the lung is involved, sarcoidosis is more common in nonsmokers (3). Nonetheless, it has been observed that there is a slightly higher incidence for sarcoidosis patients to develop malignancies, including lung cancer. The reported relative risk for lung cancer in sarcoidosis patients ranges between 2 to 3 in a number of series (8). The basis of this relation, which is under debate (10), is not known. Although several clinical studies are available regarding sarcoidosis, the molecular biology of the disease remains obscure.

Microsatellite DNA is very short tandem nucleotide repeats that are found scattered throughout the human genomes of eukaryotes (12, 13). Instability of tandem repeat DNA sequences, or microsatellite instability (MI), has been correlated with high mutational rate, and DNA repair processes have been associated with MI (14). MI has been previously reported in malignancies of various origins including lung carcinomas (15).

Additional information on the molecular pathway of cancer development is provided by the identification of novel tumor suppressor genes (TSGs). The inactivation of TSGs plays a critical role in multistage carcinogenesis. At present, loss of heterozygosity (LOH) using highly polymorphic microsatellite markers is the most common methodology employed for the localization of sites in the genome with high probability for the presence of candidate TSGs (18, 19).

The present protocol was designed to study the genetic alterations at the microsatellite level in blood and sputum cytological specimens from patients with pulmonary sarcoidosis, using ten highly polymorphic microsatellite markers located at several chromosomal arms, which could be part of the complex genetic basis of the disease. To the best of our knowledge, this is the first such study in patients with sarcoidosis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Thirty patients with clinical and histological features consistent with sarcoidosis and followed at our clinic were studied. Median age was 53 yr (range, 24 to 66 yr); 10 were male and 20 female, with a mean smoking history of 6.24 pack-years. The mean duration of illness was 4.3 ± 1.8 yr (Table 1). Exclusion criteria were coexistent chronic disease, lung infection, or malignancy.

Thirty normal subjects, matched by age, sex, and smoking history, without clinical, physical, or laboratory evidence of disease, served as the control group. Median age was 60 yr (range, 22 to 64 yr); 14 were male and 16 female, with a mean smoking history of 13.3 pack-years. They had an unremarkable physical examination, normal blood and serum analysis as well as normal chest radiographs and spirometry (Table 1). Patients and control subjects used in the present study were all Greek Caucasoid. The Greek population does not exhibit genetic varieties and, because the study was conducted on the island of Crete, they are all affected by the same environmental factors.

All patients underwent routine pulmonary function testing, including spirometry, lung volumes, diffusion capacity and arterial blood gases at rest and after exercise, chest radiography, high-resolution computed tomography (HRCT) of the thorax, and fiberoptic bronchoscopy. Characteristics of the studied subjects are shown in Table 1.

Morning sputum and peripheral blood specimens were collected from both groups. To ensure that sputum samples were from the lower respiratory tract, they were microscopically examined and considered adequate if squamous epithelial cells were less than 10 per low-power field (20).

DNA Extraction

DNA was isolated from peripheral white blood cells and sputum cells using the IsoQuick Nucleic Acid Extraction kit (ORCA; Research, Inc., Bothell, WA), according to the manufacturer's instructions.

Microsatellite and LOH Analysis

Ten microsatellite markers, located on several chromosomal arms, were used to reveal MI or LOH: namely, THRA1, D17S579, D17S855, D17S250, ANK1, D9S59, D9S290, HXB, D8S133, and D8S137 (21). Regions 8p, 17q, 9p, and 9q were evaluated for the incidence of LOH and MI. The selection of chromosome regions to be evaluated was based on previous studies concerning either lung cancer specimens (18) or benign lesions of the lung such as chronic obstructive pulmonary disease (COPD) (19).

Polymerase chain reaction (PCR) analysis (22) was performed in a 50-µl reaction volume containing 200 ng of genomic DNA, 1 µM of each primer, 250 µM deoxyribonucleoside triphosphate (dNTP), 5 µl of 10× buffer (670 mM TRIS-HCl, pH 8.5; 166 mM ammonium sulfate; 67 mM magnesium chloride; 1.7 mg/ml bovine serum albumin [BSA]; 100 µM -mercaptoethanol and 1% [wt/vol] Triton X-100), and 1 U of Taq DNA polymerase. The reactions were denatured for 5 min at 94° C and the DNA was subsequently amplified for 30 cycles at 94° C, 57° C, and 72° C each step. A volume of 10 µl of the PCR product was analyzed in a 10% polyacrylamide gel and stained with silver staining.

MI was scored by comparing the electrophoretic pattern of the microsatellite markers amplified from the paired DNA preparations (sputum/white blood cells), demonstrating a shift of one or both of the alleles in the sputum DNA specimen. The shift was indicated by either an addition or deletion of one or more repeat units resulting in the generation of novel microsatellite alleles. The appearance of additional novel bands may be explained as the result of alterations in the length of microsatellites, limited only in a cellular subpopulation of the pathological tissue, creating novel cellular clones. Thus, in a microsatellite analysis in total extracted DNA from the studied tissue, one may observe alleles from both affected and unaffected microsatellites.

Gels were scanned and the intensity of the bands corresponding to the microsatellite alleles was quantitated by an ultraviolet photometry (UVP) image analysis system. Allelic loss was scored as a 50% or more decrease in intensity of one allele relative to the other, as determined from comparison of sputum and normal DNA. The analysis in the MI/LOH positive cases was repeated at least twice and the results were highly reproducible.

The study was approved by the medical research ethics committee of our hospital.

Statistical Analysis

Differences in the mean values of quantitative measurements were tested using the Student's t or Mann-Whitney tests. The ² test was used for comparison of percentages. Analysis of covariance (logistic regression) was used for covariates. A p value < 0.05 was considered as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Results of MI and LOH in the studied sputum specimens in sarcoidosis patients are shown in Table 2. Genetic alteration was found only in the sputum specimens from the sarcoidosis patients. We found that 14 of 30 (47%) sarcoidosis patients showed genetic alterations, either MI or LOH in at least one of the studied markers. Six (20%) sarcoidosis patients exhibited MI in at least one microsatellite marker. The most commonly affected microsatellite marker was ANK1, which exhibited MI in 2 of 25 evaluated specimens (8%) on chromosome 9p. One of the samples (Patient 16) showed MI in two microsatellite markers (THRA1 and ANK1).

LOH in at least one marker was found in nine (30%) sputum samples of the patients. The most commonly affected microsatellite markers were THRA1, which exhibited LOH in 3 of 30 evaluated specimens (10%) on chromosome 17q, and ANK1, which exhibited LOH in 3 of 25 evaluated specimens (12%) on chromosome 9p. Four of the patients (13%) showed LOH in more than one marker (Table 2). One of the patients showed complete deletion of chromosomal arm 17q11.2-q21, which is probably a result of mitotic recombination events.

In total, the most commonly affected microsatellite markers were ANK1, which exhibited MI or LOH in 5 of 25 specimens (20%), THRA1 in 4 of 30 (13.3%), and D9S59 in 3 of 26 specimens (11.5%). Commonly affected regions for allelic loss were 9q (16%) and 17q (16%).

Microsatellite analysis in sputum specimens from matched healthy subjects revealed no evidence of MI or LOH in any of the genetic loci examined. Representative examples of specimens with MI and LOH are shown in Figures 1 and 2, respectively. Subgroups of sarcoidosis patients, positive and negative for MI or LOH, were compared. No statistically significant difference was found between the two subgroups in relation to age, duration of illness, smoking habit, arterial blood gas, and spirometric measurements.

View larger version (54K): [in this window] [in a new window]   Figure 1.   Representative electrophoretic patterns of five microsatellite markers in patients exhibiting MI. Arrows indicate the shift in the mobility of the alleles. The numbers above the locus name represent the patient numbers. N = normal (blood) DNA; S = sputum DNA.

View larger version (54K): [in this window] [in a new window]   Figure 2.   Representative electrophoretic patterns of four microsatellite markers in patients exhibiting LOH. Arrows indicate a significant decrease in the intensity of the alleles. The numbers above the locus name represent the patient numbers. N = normal (blood) DNA; S = sputum DNA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study we investigated the genetic alterations at the microsatellite level in 30 patients with sarcoidosis. We detected two genetic alterations in sarcoidosis microsatellite DNA: MI, reflecting an elevated mutational rate affecting 20% of the specimens, and LOH, indicating the presence of TSGs affecting 30% of the specimens.

Genetic alterations, such as MI, have been detected in almost all human tumors (15) as well as in benign diseases, such as human atherosclerotic plaques (23) and in human pterygium (24), leading to the suggestion that these diseases possess similarities with neoplasia and thus should be considered as neoplastic benign lesions. Sarcoidosis is thought to be a common benign lesion. However, there are reports of a higher tendency for these patients to develop lung cancer (8).

Microsatellites and MI have been studied for multiple purposes by geneticists and are being analyzed in tumor research. However, no consensus exists in how many loci should be analyzed and how many of them should show alterations to be classified as "high" MI. A number of publications have classified tumors as MI-positive, when as few as one of two loci appeared unstable (25). Other investigators have used the absence of alterations in three (26), four (27), or even five (28) markers as the criterion for labeling the tumor specimens as "low" MI. However, because genetic instability, once it occurs, can happen so randomly across the genome, such changes could include any percentage of the many thousands of microsatellite loci. Thus, a cancer repeat locus demonstrating stability might occur near an unstable cancer repeat locus, a fact that would remain undetectable unless the appropriate probes were used. Nevertheless, a common suggestion tends to be established, classifying as "low" MI, specimens unstable for one genetic locus, and as "high" MI, specimens unstable for two or more loci (21).

In our study, six of 30 specimens (20%) exhibited MI in at least one microsatellite marker. Considering the suggested criteria for "low" and "high" MI, we could classify only one specimen showing MI for  2 microsatellite markers as "high" MI, whereas the remaining five specimens, being unstable for one loci, were classified as "low" MI.

The repetitive unit of all the microsatellite markers used in this study was a dinucleotide. It has been shown previously that the rate of spontaneous changes in six-bases tandem repeats is threefold higher than that in two bases tandem repeat units (29). In addition, the rate of spontaneous mutation in individual tri- and tetranucleotide microsatellite markers can be 50-fold greater than that for dinucleotide microsatellites (30, 31). Therefore, the two systems (di- versus tri- and tetranucleotide microsatellites) are not directly comparable and dinucleotide microsatellites are likely to be more useful as monitors of underlying genomic instability than tri- or tetranucleotide microsatellites. The use in our study of dinucleotide microsatellites may have possibly led to underestimation of the real incidence of genome instability. Although the number of the specimens in our study is small, the detected incidence of MI is considerable. The sputum, on the other hand, contains a high amount of normal DNA, which may eliminate the signal of a mutant allele and may produce false-negative results (19), leading to underestimation. Also, examining the specimens with additional markers might increase our figures.

In this study we also report the presence of another essential malignant feature, the incidence of LOH in the sputum cells of the patients with sarcoidosis. LOH is an event that takes place in tumor cells or in premalignant cells that progress toward malignancy. This fact constitutes strong evidence that there may be transformed cells in the sarcoid tissue. According to Knudson's "two hit hypothesis" (32), the phenomenon of LOH is correlated with the existence of TSG implicated with the disease.

Significant incidence of LOH was found in the locus 17q11.2-q21 suggesting that important TSGs for the development of sarcoidosis may be located on this chromosomal region. Deletions at 17q occur frequently in a variety of neoplasms. These include ovarian tumors (flanked by THRA1 and D17S75) (33), esophageal tumors between probes C117-316 and C117-710 (34), and laryngeal tumors between D17S250 and D17S579 (35). Nonsmall cell lung (36) and prostate (37) cancers are associated with deletions at 17q near the BRCA1 region. The wide spectrum of human cancers affected by alterations of the candidate TSG(s) of 17q suggests a significant role for these genes in the development of neoplasia. Fine mapping of this area is required to establish the precise location of candidate TSG(s) and their role in the pathogenesis of sarcoidosis.

Epidemiological and clinical studies demonstrated a hereditary susceptibility to the pathogenesis of sarcoidosis (38, 39). These studies support the concept that genetic factors may predispose to this disorder or determine clinical expression of the disease. Monozygotic twins appear more likely to both have sarcoidosis than dizygotic twins, strongly suggesting a genetic component to the disease (39). The lack of a clear genetic pattern suggests that susceptibility to sarcoidosis is likely polygenic and may possibly be associated with environmental factors.

Smoking, however, seems not to be related to the previously described genetic abnormalities, because smokers among the sarcoidosis patients did not exhibit MI or LOH in greater numbers compared with nonsmokers. Additional evidence to support this finding is provided by a previous study, in which we found that MI was detected in patients with COPD but not in a matched group of non-COPD smokers (40).

In order to assess whether MI or LOH is an index of the severity of sarcoidosis, two subgroups of MI/LOH-positive and MI/LOH-negative sarcoidosis patients were compared. The results showed that there were no significant differences between these two subgroups in terms of duration of illness, smoking habit, arterial blood gases, lung function, chest radiography classification, and need for treatment. This suggests that MI and LOH are not related to the severity of the disease.

The precise significance of these findings remains obscure, because the information regarding the genetic basis of the disease is limited. However, we may speculate that the relatively high mutational rate of sarcoidosis patients, as reflected in the instability of the microsatellite sequences, indicates a destabilization of the genome, which may affect other genes resulting in the dysregulation of the cells harboring these mutations.

In conclusion, the results of this study showed for the first time that MI and LOH of DNA are detectable phenomena in sarcoidosis and might be implicated in the etiopathogenesis of the disease. Further studies are needed to evaluate the clinical significance and the prognostic value of these genetic alterations for possible lung cancer development.