Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The pathology of interstitial lung diseases is characterized by accumulation of inflammatory cells and a variable extent of pulmonary fibrosis. Sarcoidosis is a systemic granulomatous disease of unknown etiology that commonly involves the lung and is characterized by the infiltration of abundant CD4⁺ lymphocytes and alveolar macrophages in the alveolar septa and the formation of noncaseating epithelioid cell granulomas in the interstitium. Interstitial pneumonia associated with collagen vascular disease (CVD-IP) and idiopathic pulmonary fibrosis (IPF) appear to share some common pathogenic mechanisms (1). Increased numbers of inflammatory cells, immunoglobulins, and immune complexes have been demonstrated in lung tissue and bronchoalveolar lavage fluid (BALF) in both diseases (3). Interactions between activated inflammatory cells, and cells comprising the alveolar structures, orchestrate a complex process leading to chronic inflammation and progressive pulmonary fibrosis.

It seems clear that these diseases are chronic inflammatory disorders in which the first manifestation is the accumulation of inflammatory cells in the lung parenchyma. In previous studies of BALF cells from patients with sarcoidosis, several proinflammatory cytokines and other mediators were found to control both the formation and persistence of granulomas (7). Platelet-derived growth factor produced by alveolar macrophages has been postulated to mediate fibrosis in pulmonary sarcoidosis and IPF (12). However, the mechanisms by which inflammatory cells accumulate in the interstitium are not well understood. With respect to this issue, emphasis has recently been placed on the roles of chemoattractants in pulmonary sarcoidosis and IPF (13).

The chemokine family characterized by a conserved four-cysteine residue motif is thought to possess leukocyte-specific chemoattractive properties, which might contribute to selective extravasation of inflammatory cells into inflamed tissues. RANTES, a member of the C-C branch, has been reported to be chemotactic for certain lymphocyte subsets, eosinophils, monocytes, and mast cells, but not for neutrophils (14). We hypothesized that RANTES was associated with the accumulation of these inflammatory cells in the lungs in various interstitial lung diseases. In the present study, we examined the expression of RANTES mRNA and measured RANTES protein in patients with these common but distinct interstitial lung diseases.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Study Population

Bronchoalveolar lavage (BAL) was performed by the use of a flexible fiberoptic bronchoscope in 20 patients with sarcoidosis, 10 patients with CVD-IP, and 10 patients with IPF. Eight healthy volunteers (HV) served as control subjects. All subjects were lifetime never-smokers. The present study was conducted solely in never-smokers because it has been demonstrated that smoking exerts profound effects on the production of certain cytokines (17, 18).

Diagnosis of sarcoidosis was based on compatible clinical features and biopsy evidence of noncaseating epithelioid cell granulomas without positive growth of mycobacteria and fungus. On chest X-ray films, one patient was in Stage 0, 11 patients were in Stage I, five patients in Stage II, and three patients in Stage III. Ten patients had CVD-IP (rheumatoid arthritis in three, systemic lupus erythematosus in three, systemic sclerosis in one, polymyositis in one, Sjögren's syndrome in two). The diagnosis of CVD-IP was based on diffuse interstitial patterns and/or fibrosis on chest roentgenograms, serologic evidence of CVD, and compatible clinical manifestations. Diagnosis of IPF was based on roentgenologic findings showing various intensities of pulmonary fibrosis with compatible impairment of pulmonary function and on the absence of evidence for underlying diseases that could cause pulmonary fibrosis. Lung biopsies obtained by fiberoptic bronchoscopy or thoracoscopy yielded findings compatible with IPF in all patients. Impairment of pulmonary function was generally mild in all disease groups (Table 1). Patients with sarcoidosis had significantly better vital capacity than those with the other diseases (Table 1). One patient with sarcoidosis and four patients with CVD-IP were taking a systemic steroid at the time of BAL. No patient with IPF was treated with inhaled or systemic steroids.

To qualify as a control subject, an individual had to have absence of clinical or radiologic evidence of lung diseases and no abnormality of differential cytologic counts in BALF. All patients and HV gave informed consent for BAL.

Bronchoalveolar Lavage

We performed BAL as previously described (8). Briefly, three 50-ml aliquots of 0.9% NaCl were instilled into the lingular segment in most sarcoid cases or other segments where roentgenologic manifestation of CVD-IP or IPF was predominant. Each aliquot of infused fluid was immediately aspirated into a sterile plastic tube by a gentle suction system and kept at 4°C. Recovered BALF was strained through two layers of gauze and centrifuged to recover cell pellets. The cell-free supernatants were stored at 40°C for further analysis. Cell pellets were washed twice with modified (calcium- and magnesium-free) Hanks' balanced salt solution (HBSS; GIBCO, Grand Island, NY), and resuspended in a known volume of HBSS. Total cell counts were made using a hemocytometer and cell differentials were determined after cytocentrifugation and staining with Diff-Quik (International Reagents Corp., Kobe, Japan). T-cell subsets were determined by staining BALF cells with monoclonal antibodies (fluorescein isothiocyanate-conjugated T3, T4 and T8; Coulter Immunology, Hialeah, FL) and analyzing them with a flow cytometer (Epics C; Coulter Immunology).

Extraction of Total Cellular RNA

Immediately after centrifugation of the resuspended BALF cells, total RNA was extracted by the acid guanidinium thiocyanate-phenol-chloroform method of Chomczynski and Sacchi (19) using Isogen^® (Nippon Gene, Toyama, Japan). The RNA pelleted at the tube bottom was dissolved in RNase-free water. The RNA concentration was then determined by measuring optical density at 260 nm. Blood was obtained by venipuncture with a heparinized syringe, and peripheral blood mononuclear cells (PMNC) were isolated by density gradient centrifugation with Ficoll-Paque Plus^® (Pharmacia, Uppsala, Sweden). We extracted total RNA from PMNC by the same method as for BALF cells.

Semiquantitation of the Expression of RANTES mRNA

To better quantitate the amount of RANTES mRNA, we developed a competitive reverse transcription-polymerase chain-reaction (RT-PCR) method using an internal standard for RANTES mRNA according to the method previously described (20). The RANTES gene spans approximately 7.1 kbp and is composed of three exons of 133, 112, and 1,075 bp and two introns of approximately 1.4 and 4.4 kbp with the position of intron/exon boundaries conserved relative to the other C-C chemokine family members (21).

First, we synthesized a pair of primers to amplify a cDNA strand (target cDNA) reverse-transcribed from RANTES mRNA. One of them (primer A) had the structure for the 5' untranslated region of exon 1 (5'-ACAGGT-ACCATGAAGGTCTC-3'), and the other (primer B) for the 3' end of the translated region of exon 3 (5'-TCCTAGCTCATCTCCAAAGA-3'). They were designed based on the published nucleotide sequence of human RANTES (21). Second, we designed a third primer (primer C, 5'-CACAGCTTTGTCACCCGAAA-3') that had the structure for the 5' end of exon 3 to amplify a DNA strand from the genomic RANTES DNA in conjunction with primer B. Third, we tandemly combined primer A with primer C (primer A+C) so that the DNA strand could be amplified with the same primer pair as that for the target cDNA (primers A and B). The PCR product amplified with primers A+C and B from genomic DNA was then subcloned into a vector of the TA Cloning System^® (Invitrogen, San Diego, CA) according to the manufacturer's instructions. Using T7 RNA polymerase (Stratagene, La Jolla, CA), we then transcribed an artificial RNA strand as an internal standard for RANTES mRNA from the plasmid DNA.

Reverse Transcription of RANTES mRNA and the Internal Standard

A fixed amount (250 ng) of sample RNA and 10-fold serial dilutions of the internal standard were reverse transcribed together. Briefly, 20 µl of reaction mixture contained 10 units of Molony murine leukemia virus reverse transcriptase (Boehringer Mannheim, Mannheim, Germany), enzyme buffer, sample total RNA, the 10-fold serially diluted internal standard, 20 units of RNase inhibitor (Toyobo, Osaka, Japan), 10 pmol of random primer (Takara, Kyoto, Japan), and 1 mM deoxynucleoside triphosphate (Takara). The internal standard was included in the reaction mixture at dilutions ranging from 10⁵ to 10⁹ of 250 ng. The reaction mixture was incubated at 42°C for 60 min, heated at 95°C for 5 min to inactivate the enzyme, and then chilled on ice.

PCR Amplification

A 5-µl aliquot of the reverse-transcription product was amplified in a 50-µl reaction mixture containing 5 µl of 10 × PCR buffer, 200 mM deoxynucleoside triphosphate, 25 pmol each of primers A and B, and 2 units of Taq DNA polymerase (Takara). It was overlaid with mineral oil to prevent evaporation. The reaction was initiated by heat denaturation at 94°C for 1.5 min, annealing at 61°C for 1 min, and extension at 72°C for 1 min. This was repeated for 40 cycles with the final extension process for 10 min to allow completion of synthesis of the amplified product.

PCR products were electrophoresed on 2% gel and visualized by staining with ethidium bromide. The gel electrophoretic profile of such competitive RT-PCR series allowed easy identification of the equivalent point (i.e., equal amounts of competitor RNA and target mRNA) at which the ethidium bromide fluorescence intensities of the short band of the internal standard (133 bp) and the long band of the target mRNA (296 bp) appeared to be visually similar (Figure 1). We arbitrarily defined the amount of expression of RANTES mRNA as 10 + n (U), where n equals the exponent of 10 (Figure 1).

View larger version (49K): [in this window] [in a new window]   Figure 1.   Competitive RT-PCR analysis of RANTES mRNA. cDNA for target mRNA and for serially diluted artificial RNA (the internal standard for RANTES) was synthesized in the same tube, amplified with the same primer pair (A and B), and visualized on 2% gel electrophoresis by staining with ethidium bromide. In this case, the ethidium bromide fluorescence intensities of the short band of the internal standard and the long band of the target mRNA seemed most similar at 106 dilution. The expression of RANTES mRNA was then calculated in arbitrary units by adding 6 to 10 (= 4U) in accord with the definition given in the text.

Concentration of BALF

BALF was concentrated using Centricon-3^® (Amicon, Beverly, MA) with a molecular weight cutoff of 3 kD according to the recommendation of the manufacturer. Briefly, 2 ml of BALF was added to a Centricon-3^® tube and then centrifuged at 5,000 × g for 100 min. After centrifugation, the retained volume of BALF was 80 to 250 µl, which corresponded to concentrations from 8- to 25-fold. The volume of the concentrated BALF was adjusted to 250 µl to keep the concentration factor constant (8-fold).

Measurement of RANTES in BALF and Serum

The concentration of RANTES protein was determined by specific enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN). The sensitivity of detection was 31.2 pg/ml. All samples were measured using the same assay systems.

Statistical Analysis

Data are expressed as median and ranges. Results were tested by the Mann-Whitney U test for differences between groups. To examine correlations, Spearman's rank correlation coefficients were calculated. Differences with a P value of less than 0.05 were considered significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Bronchoalveolar Lavage

Table 1 shows results of pulmonary function tests and BALF cellular profiles for HV and patients in whom we could examine both the expression of RANTES mRNA and RANTES protein levels in BALF. There was no statistically significant difference in the recovery rates of the four groups. The BALF profile of sarcoid patients was characterized by the accumulation of lymphocytes with CD4 predominance, as has been well documented (22) (Table 1). In patients with CVD-IP, the percentages of lymphocytes and neutrophils were significantly higher than those in HV. In patients with IPF, eosinophils and neutrophils were the inflammatory cells that significantly increased in BALF.

Expression of RANTES mRNA by Total BALF Cells and PMNC

With the use of a sensitive and semiquantitative RT-PCR method we could detect the expression of RANTES mRNA by freshly isolated BALF cells and PMNC from most HV, thus demonstrating constitutive expression of RANTES mRNA by resident BALF cells and resting PMNC (Figures 2a and 2b). The expression of RANTES mRNA in BALF cells from patients with sarcoidosis, CVD-IP, and IPF were significantly increased compared with HV (Figure 2a). Meanwhile, the expression of RANTES mRNA by PMNC was similar in all four groups (Figure 2b).

View larger version (13K): [in this window] [in a new window]   Figure 2.   Expression of RANTES mRNA by total BALF cells (a) and peripheral blood mononuclear cells (b). Definitions for abbreviations: HV = healthy volunteers; Sar = pulmonary sarcoidosis; CVD-IP = interstitial pneumonia associated with collagen vascular disease; IPF = idiopathic pulmonary fibrosis. Medians are indicated by bold lines.

Correlation between the Expression of RANTES mRNA and BALF Findings

There were significant positive correlations between the expression of RANTES mRNA by BALF cells and the percentage of lymphocytes (Figure 3a), the absolute number of lymphocytes, and the absolute number of CD4⁺ lymphocytes in patients with sarcoidosis. A similar correlation between the expression of RANTES mRNA and the proportion of BALF lymphocytes was observed for patients with CVD-IP (Figure 3b).

View larger version (14K): [in this window] [in a new window]   Figure 3.   Relationship between the expression of RANTES mRNA by total BALF cells and the percentage of lymphocytes in BALF in patients with pulmonary sarcoidosis (a) and CVD-IP (b).

Levels of RANTES Protein in BALF and Serum

Patients with sarcoidosis, CVD-IP, and IPF had a significantly higher level of RANTES protein in BALF than HV (Figure 4a), whereas serum RANTES levels in the three disease groups did not differ significantly from those in HV (Figure 4b).

View larger version (15K): [in this window] [in a new window]   Figure 4.   Levels of RANTES protein in BALF (a) and serum (b). For definitions of abbreviations, see Figure 2. Medians are indicated by bold lines.

Correlation between the Expression of RANTES mRNA and Levels of RANTES Protein

There was a loose but significant positive correlation between the expression of mRNA of RANTES by BALF cells and levels of RANTES protein in BALF in overall study subjects, including both HV and all patients (rs = 0.43, P < 0.001, n = 48). This correlation was not observed for the individual study group because of the small number of subjects in each group.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The major findings in this study were the more spontaneous presence of RANTES mRNA in unstimulated BALF cells and the higher protein levels in BALF from patients with pulmonary sarcoidosis, CVD-IP, and IPF than in that from HV. A few studies have recently reported increased mRNA expression of RANTES or elevated RANTES protein in asthmatic patients (23, 24). However, to our knowledge, the significance of RANTES has not yet been evaluated for interstitial lung diseases.

It seems reasonable to link increased expression of RANTES at both the mRNA and protein level with the accumulation of lymphocytes in sarcoidosis and CVD-IP, because BALF lymphocytosis was observed for these diseases. This notion was supported by the finding that the expression of RANTES mRNA by BALF cells was positively correlated with the BALF findings that reflected the accumulation of lymphocytes. Patients with CVD-IP also had an increased proportion of BALF neutrophils; however, RANTES does not act on them. These findings suggested that heightened expression of RANTES mRNA and subsequent translation of its protein were at least partially responsible for the accumulation of lymphocytes in these distinct interstitial lung diseases, acting in concert with other chemokines, including monocyte chemotactic protein-1 (MCP-1) (13, 25).

Recently, RANTES has been reported to play a more important role in T-cell activation in the presence of interleukin-2 in vitro than other chemokines, such as macrophage inflammatory protein 1 (MIP-1), MIP-1, MCP-1, and platelet factor-4 (PF-4) (26). Thus, RANTES may be involved in both the accumulation and activation of lung T cells. It is thus possible that RANTES contributes to the formation of an amplification loop of local inflammation by lymphocytes at the site of disease, and hence, to the subsequent fibrosis leading to impairment of lung function.

The mechanisms of predominant accumulation of CD4⁺ lymphocytes in the interstitium of the sarcoid lung are not well understood. Two mechanisms are hypothesized: selective recruitment from the bloodstream and local proliferation. It was initially reported that RANTES selectively recruited human CD4⁺ lymphocytes in vitro (14), which prompted us to initiate the present investigation. Actually, the increased expression of RANTES mRNA by BALF cells was associated with more abundant recovery of CD4⁺ lymphocytes in BALF from patients with pulmonary sarcoidosis (Table 1). However, it was later shown that RANTES induces the migration of both CD4⁺ and CD8⁺ lymphocytes (27). In fact, in the present study, the total number of CD8⁺ lymphocytes in BALF was significantly increased in patients with pulmonary sarcoidosis (Table 1). Accordingly, it is also possible that RANTES is involved in the recruitment and activation of both CD4⁺ and CD8⁺ lymphocytes in pulmonary sarcoidosis.

Patients with IPF also had increased expression of RANTES mRNA by BALF cells and elevated levels of RANTES protein in BALF, like other disease groups. However, the implication of this finding seems to differ from those in sarcoidosis and CVD-IP, because the percentage of BALF lymphocytes was not significantly increased in IPF. Instead, patients with IPF had significantly greater eosinophil and neutrophil accumulation in the lung than did HV. In view of the lack of chemotactic effect by RANTES on neutrophils, it is conceivable that RANTES may, in part, be associated with the accumulation of eosinophils in IPF. The role of RANTES in human lung diseases was initially suggested in bronchial asthma, in which RANTES is localized in bronchial epithelium and inflammatory cells (28). This was later confirmed by Berkman and colleagues (23). Teran and associates recently reported increased levels of RANTES protein in BALF at the site challenged with an allergen (24). They found a significant correlation between the concentration of RANTES and the number of eosinophils. Thus, RANTES may induce the selective distribution of eosinophils to both the airways and lung interstitium.

RANTES is produced by, and acts as potent chemoattractant for, overlapping but distinct classes of cells (14, 15, 26, 29). Its perceived signals may result in chemotaxis, or activation, or both, depending on the concentration of RANTES present and the interaction with other chemokines or surface receptors (35). It is therefore most likely that RANTES is not a determinant of the final profile of inflammatory cells in the lung, but one of the chemotactic cytokines that favor cellular inflammation in interstitial lung diseases. Analysis of the correlation between several chemokines and BALF findings and/or experiments on blocking of chemotactic activities in BALF will be needed to evaluate the real significance of RANTES in comparison with other chemokines.

To semiquantitate RANTES mRNA most accurately, we modified the method of Wang and coworkers (36) and synthesized control RNA that had the same 5'- and 3'- ends and partial composition of the target mRNA, which was therefore smaller than, and could be discriminated from, the target. The internal standard could be amplified with the same primer pair and thus, presumably, with efficiency similar to that for the endogenous target sequences. This method should be widely applicable for measuring the expression of other cytokine mRNAs.

In conclusion, we have demonstrated the adaptability of a competitive RT-PCR technique to semiquantitate the expression of cytokine genes in vivo. We found for the first time that RANTES was expressed in a heightened manner at both the mRNA and the protein levels in the lungs of patients with sarcoidosis, CVD-IP, and IPF. This was accompanied by increased recovery of lymphocytes or eosinophils in BALF, suggesting that RANTES is involved in the mechanism by which these inflammatory cells accumulate in the lung.