Vascular dysfunction by myofibroblast activation in patients with idiopathic pulmonary fibrosis and prognostic significance.

In this study, we demonstrated the importance of telomerase protein expression and determined the relationships among telomerase, endothelin-1 (ET-1) and myofibroblasts during early and late remodeling of parenchymal and vascular areas in usual interstitial pneumonia (UIP) using 27 surgical lung biopsies from patients with idiopathic pulmonary fibrosis (IPF). Telomerase+, myofibroblasts α-SMA+, smooth muscle cells caldesmon+, endothelium ET-1+ cellularity, and fibrosis severity were evaluated in 30 fields covering normal lung parenchyma, minimal fibrosis (fibroblastic foci), severe (mural) fibrosis, and vascular areas of UIP by the point-counting technique and a semiquantitative score. The impact of these markers was determined in pulmonary functional tests and follow-up until death from IPF. Telomerase and ET-1 expression was significantly increased in normal and vascular areas compared to areas of fibroblast foci. Telomerase and ET-1 expression was inversely correlated with minimal fibrosis in areas of fibroblast foci and directly associated with severe fibrosis in vascular areas. Telomerase activity in minimal fibrosis areas was directly associated with diffusing capacity of the lung for oxygen/alveolar volume and ET-1 expression and indirectly associated with diffusing capacity of the lungs for carbon monoxide and severe fibrosis in vascular areas. Cox proportional hazards regression revealed a low risk of death for females with minimal fibrosis displaying high telomerase and ET-1 expression in normal areas. Vascular dysfunction by telomerase/ET-1 expression was found earlier than vascular remodeling by myofibroblast activation in UIP with impact on IPF evolution, suggesting that strategies aimed at preventing the effect of these mediators may have a greater impact on patient outcome.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a devastating chronic fibrosing interstitial pneumonia of unknown etiology that typically increases in prevalence with advanced age, characterized by excessive collagen deposition and irreversible remodeling of the lung parenchyma (1,2). The histologic pattern associated with IPF is the usual interstitial pneumonia (UIP), characterized by patchy and temporally heterogeneous fibrosis with excessive extracellular matrix (ECM) and honeycomb change, interspersed with normal and collapsed areas (1). Although the etiology of IPF/UIP is not understood, many investigators hypothesize that dysregulated communication between injured epithelial cells and activated mesenchymal cells represents an important factor in its pathogenesis (3). Vascular changes are another event frequently observed in IPF (4,5). In previous studies, we found a direct relationship between progressive vascular occlusion by collagen/elastic fiber deposition and tissue remodeling in surgical lung biopsies from patients with IPF (4,5). These findings probably are the consequence of persistent myofibroblast activation by mesenchymal cells, resulting in a heterogeneous fibroblast phenotype, intermediate between fibroblasts and smooth muscle cells (6). Although heterogeneity in fibroblast phenotype is evident, a subset of differentiated activated fibroblasts is characterized by the expression of alpha-smooth muscle actin (α-SMA) (7). Depending on the precise stimulatory milieu, fibroblasts either transform to myofibroblasts or proliferate, resulting in vascular occlusion or areas of fibroblastic foci, which are thought to be the sites of active ECM, collagen and elastic synthesis, and are considered to be the leading edge of fibrosis and vascular remodeling (8). In this regard, endothelin-1 (ET-1) is one of the more potent inducers of collagen deposition, with its fibrogenic effects being central in wound healing and tissue repair. A major role of ET-1 involves transition of quiescent fibroblasts to an “activated” phenotype, whereby matrix production and wound contraction are important consequences (9). Telomerase is a specialized polymerase that adds telomere repeats to the ends of chromosomes by using its intrinsic RNA component as a template, thereby compensating for the telomere loss that normally occurs with each cell division (10). Telomerase has been shown to be essential for unlimited cell proliferation and has been linked to immortality (11). Bleomycin-induced lung injury and fibrosis is also known to induce telomerase activity in the affected lung tissue and isolated lung fibroblasts (12). The fibroblasts isolated from such lungs undergoing fibrosis show increased intrinsic proliferative capacity and are able to differentiate into α-SMA expressing myofibroblasts, which are also a key source of cytokines with inflammatory and fibrogenic properties (13, 14). An injured lung fibroblast population may contain telomerase expressing cells with an extended life span, which could contribute to the observed increased numbers of lung fibroblasts (12).

In view of these considerations suggesting an interaction among telomerase, endothelium and myofibroblast activation, the aim of the present study was to evaluate 1) whether telomerase, ET-1 and myofibroblast α-SMA are expressed in vascular and parenchymal areas of UIP; 2) to verify if the measures of mediator expression correlate with some measure of fibrosis, and 3) to determine the impact of the resultant parenchymal/vascular remodeling on patient survival. It was hypothesized that in UIP, a devastating chronic fibrosing interstitial pneumonia, the mediators under study exert their effect by influencing parenchyma/vascular remodeling/fibrosis with impact on patient outcome.

Material and Methods

This research was approved by the Institutional Ethics and Scientific Committees (protocol No. 0960/08) and all patients gave written informed consent to participate in the study.

Open lung biopsies

Pulmonary specimens were obtained by surgical lung biopsy from 27 patients with IPF/UIP (14 males and 13 females; mean age 64 ± 1.4 years), according to criteria outlined in the American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of idiopathic interstitial pneumonias (1). Only specimens from patients who fulfilled these consensus criteria were included. Three lung specimens obtained by open surgical biopsy from each patient were collected according to the normal, intermediate and more affected areas in different parts of the lung determined by high-resolution computed tomography. The histologic UIP pattern of IPF was characterized by the patchy subpleural and paraseptal distribution of parenchymal injury. Temporal heterogeneity was seen at low magnification, with alternating areas of normal lung parenchyma with septal fibro-myxoid tissue (fibroblastic foci), honeycomb and vascular thickening.

Baseline characteristics

The pulmonary function tests included vital capacity, forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), FEV1/FVC ratio 100X, total lung capacity (TLC), residual volume (RV), and carbon monoxide transfer factor (diffusing capacity of the lung for carbon monoxide, DLCO). The percentages of TLC, RV, and RV/TLC were measured by the helium dilution method using a Master Screen apparatus (Erich Jaeger GmbH, Germany). The DLCO and DLCO/alveolar volume (VA) were assessed by the single breath holding helium dilution method (15). Lung function results (Table 1) are reported as the percentage of patient lung function compared to predicted values for a standardized healthy population. In all patients, the arterial partial pressure of carbon dioxide (PaCO2) and partial pressure of oxygen (PaO2) were also measured at rest.

Table 1.

Clinical data of the patients.

Variables	Idiopathic pulmonary fibrosis
Age at biopsy (years)	64.5 ± 7.4
Gender (males/females)	13/14
Spirometry
FEV1 (% predicted)	78 ± 4.4
FVC (% predicted)	69.5 ± 3.5
FEV1/FVC	20 ± 8.8
TLC (% predicted)	79 ± 5
RV (% predicted)	102 ± 15.9
TLC/RV (%predicted)	44.2 ± 3.4
DLCO (% predicted)	59.8 ± 8
DLCO/VA (% predicted)	59 ± 11.7
PaCO2	60.2 ± 3.7
PaO2	38.1 ± 1.3

Data are reported as means ± SEM. FEV1 = forced expiratory volume in 1 s; FVC = forced vital capacity; TLC = total lung capacity; RV = residual volume; DLCO = diffusing capacity of the lung for carbon monoxide; VA = alveolar volume; PaCO2 = partial pressure of carbon dioxide; PaO2 = partial pressure of oxygen.

Immunohistochemistry

A standard peroxidase technique was used, with Harris's hematoxylin as the counterstain, to identify telomerase+ epithelial cells, ET-1+ endothelium and α-SMA+ myofibroblasts and caldesmon+ smooth muscle cells in normal tissue, fibroblastic foci and vascular areas of UIP. Since the α-SMA antibody stains myofibroblasts as well as smooth muscle cells, which are a regular component of arteries, the caldesmon (C21) antibody sc-58700 was specifically used to characterize these cells. All antibodies used were biotinylated rabbit polyclonal antibodies. Anti-telomerase, anti-ET-1, anti-αSMA and anti-caldesmon polyclonal antibodies (Santa Cruz Biotechnology, Inc., USA) were incubated with tissue sections at 1:100 dilution. The Max Polymer Novolink amplification kit (Leica, Newcastle Inc., UK) was used for signal amplification and 3,3′-diaminobenzidine tetrachloride (0.25 mg dissolved in 1 mL 0.02% hydrogen peroxide) was used as a precipitating substrate for signal detection. The specificity of primary antibodies was confirmed by appropriate reagent controls (omitting the primary antibody or substituting non-immune serum for the primary antibody in the staining protocol), which revealed no staining.

Histomorphometry

Regional differences between samples from the same patient in the three lung specimens obtained by open surgical biopsy in each patient were evaluated for telomerase+ epithelium, ET-1+ endothelium, α-SMA+ myofibroblasts, caldesmon+ smooth muscle cells, and fibrosis severity in 30 fields by the point-counting technique (16) and a semiquantitative score (4). As previously described, the histologic pattern of UIP was characterized by alternating areas of normal lung parenchyma, minimal fibrosis (fibroblastic foci), severe (mural) fibrosis, honeycombing, and vascular thickening. At 400X magnification, 30 fields were chosen from across multiple sites accounting/adjusting for cellularity or as number of positively staining cells per mesenchymal tissue. In other words, normal lung parenchyma, minimal fibrosis (fibroblastic foci), severe (mural) fibrosis, and vascular areas present in UIP in each field were determined according to the number of points hitting mesenchymal tissue, as a proportion of the total grid area. The number of positive cells within the normal lung parenchyma, minimal fibrosis (fibroblastic foci), severe (mural) fibrosis, and vascular areas were then counted. The immunostaining cellularity was determined as the number of positive cells in each field divided by the normal lung parenchyma, minimal fibrosis (fibroblastic foci), severe (mural) fibrosis, and vascular areas. The final results are reported as means ± SEM of 3 lung specimens from each patient in 30 random, non-coincident microscopic fields.

The severity of fibrosis was assessed by semiquantitative analysis in UIP areas of 1) mild interstitial thickening by fibroblastic foci (minimal fibrosis, Figure 1A and B), and 2) severe mural-organizing fibrosis with honeycombing and foci of actively proliferating fibroblasts and myofibroblasts (severe fibrosis, Figure 1C-F).

Figure 1.

Degree of fibrosis in the pattern of usual interstitial pneumonia (UIP). UIP minimal fibrosis is characterized by interstitial thickening (arrows) and by fibroblastic foci (FF) (A and B) and UIP severe fibrosis is characterized by severe mural-organizing fibrosis (arrows) (C) with foci of actively proliferating fibroblasts and myofibroblasts (FF) (D) and honeycombing (stars) (E). Vascular (VAS) representation is observed in the middle of pulmonary fibrosis (F). Hematoxylin and eosin.

Interobserver comparisons were performed for 20% of the slides by 2 observers (E.R.P. and V.L.C.). The coefficient of variation for the interobserver error of cell count was <5%.

Statistical analysis

Data are reported as means ± SEM with 95% confidence intervals. ANOVA with the Bonferroni test were used to analyze the relationship between continuous variables and the residuals were examined to ensure that their distribution was approximately normal. The relationship between the immunostaining cellularity was evaluated by Pearson's correlation. Actuarial survival was estimated using the Kaplan-Meier method and compared using the log-rank test. Cox proportional hazards regression was used to ascertain the individual contribution of factors associated with survival and to compare adjusted survival between groups. The statistical program used was SPSS 18.0 (SPSS Inc., USA). The level of significance was set at 5% in all tests.

Results

Telomerase activation

Immunochemical staining of epithelium telomerase+ in normal (Figure 2B), minimal fibrosis (Figure 2C), severe fibrosis (Figure 2D), and vascular (Figure 2F) areas of UIP, appears as brownish cells. Epithelium telomerase+ immunostaining was more exuberant in normal and vascular areas than in fibrosing areas of the histologic pattern of UIP than in control areas (Figure 2A and E).

Figure 2.

Telomerase expression in control areas (A and E) and in a normal area (B), in an area of minimal fibrosis (arrows) (C), of fibroblastic foci (FF) of severe fibrosis (D) and in vascular (F) areas of usual interstitial pneumonia (UIP), when stained by immunohistochemistry. Numerous cells are observed in the different areas of UIP pattern when compared with the control areas.

Quantitative analysis confirmed the morphologic distribution of telomerase+ epithelium expression (Table 2). Telomerase+ epithelium was significantly overexpressed in normal and vascular areas than in UIP fibrosing and control areas.

Table 2.

Histomorphometric markers and cellularity in non-fibrosing lung tissue (control) and different areas of fibrosing lung (usual interstitial pneumonia, UIP).

Lung tissue	Telomerase	α-SMA	Caldesmon	Endothelin-1
Non-fibrosing lung (control)
Septal areas	0.78 ± 0.42	0	0	0.01 ± 0.02
Vascular areas	0.01 ± 0.02	9.57 ± 2.90	5.2 ± 1.15	0.0003 ± 0.0003
Fibrosing lung (UIP)
Normal lung parenchyma	17.8 ± 6.7+	0	0	17.2 ± 6.4#
Minimal fibrosis	7.4 ± 2.1	8.2 ± 3.7	0.002 ± 0.0005§	5.7 ± 2.2
Severe (mural) fibrosis	6.4 ± 2.5	12.0 ± 6.6	0.06 ± 0.001§	7.2 ± 2.9
Vascular areas	18.2 ± 7.2+	51.7 ± 10.4*	0.45 ± 0.02§	18.0 ± 4.3#

Data are reported as means ± SEM percent of markers for 3 lung specimens obtained by open surgical biopsy from each patient (10 random, non-coincident microscopic fields were analyzed in each specimen). All lung specimens were analyzed for epithelial, myofibroblast, endothelial and smooth muscle cells from normal lung parenchyma, and minimal fibrosis (fibroblastic foci), severe (mural) fibrosis and vascular areas of UIP for percentage of telomerase, α-SMA, caldesmon, and endothelin-1 fractional areas.

Telomerase expression values in UIP normal lung parenchyma and vascular areas differed significantly from minimal (fibroblastic foci) and severe (mural) fibrosis (P < 0.01).

α-SMA expression values in UIP vascular areas were significantly different from normal lung parenchyma and fibrosis areas (P < 0.01).

Caldesmon expression values in UIP fibrosis and vascular areas differed significantly from normal lung parenchyma (P < 0.001). Endothelin-1 expression values in normal lung parenchyma and vascular areas differed significantly from fibrotic and control areas (P < 0.01). All markers in UIP were significantly increased compared to non-fibrosing control lungs (P < 0.0001). ANOVA with the Bonferroni test were used for statistical analysis.

Myofibroblasts and smooth muscle cell replication

α-SMA+ myofibroblasts in normal (Figure 3A), fibrosing (Figure 3C) and vascular (Figure 3E) areas of UIP and control areas (Figure 3G), when stained by immunohistochemistry, appear as brownish cells. α-SMA+ myofibroblasts were more prominent than smooth muscle cells caldesmon+ in vascular (Figure 3F) than fibrosing (Figure 3D) and control (Figure 3H) areas. Normal areas did not present α-SMA+ myofibroblasts or caldesmon+ smooth muscle cells (Figure 3B).

Figure 3.

Alpha-smooth muscle actin (α-SMA) (A, C, E, G) and caldesmon (B, D, F, H) expression in normal (A, B), fibroblastic foci (FF) of fibrosing (C, D) and vascular (E, F) areas of usual interstitial pneumonia (UIP) pattern, and vascular control areas (G, H) stained by immunohistochemistry. α-SMA+/myofibroblasts are more prominent than caldesmon+ smooth muscle cells in vascular areas than in fibrosing and control areas. Normal areas do not show α-SMA+ myofibroblasts or caldesmon+ smooth muscle cells.

Quantitative analysis confirmed the morphologic distribution of α-SMA+ myofibroblasts and caldesmon+ smooth muscle cells (Table 2). Interestingly, α-SMA+ myofibroblasts were significantly more increased than caldesmon+ smooth muscle cells in vascular areas of UIP (P < 0.001).

Endothelium activation

ET-1+ endothelium in normal (Figure 4B), minimal (Figure 4C) and severe (Figure 4D) fibrosing, as well as vascular (Figure 4F) areas of UIP, when stained by immunohistochemistry, appears as brownish cells. Numerous cells in normal and vascular areas expressed ET-1 when compared to fibrosing areas of the UIP histologic pattern and control areas (Figure 4A and E).

Figure 4.

Endothelial ET-1+ expression in control areas (A, E) and in normal (B), minimal (C, arrows) and severe (D) fibrosing, showing a fibroblastic foci (FF) as well as vascular (F) areas of usual interstitial pneumonia (UIP) pattern when stained by immunohistochemistry. Numerous cells in normal and vascular areas expressed endothelin-1 when compared to fibrosing areas of the UIP and control areas. Endothelin-1 immunohistochemistry.

Quantitative analysis confirmed the morphologic distribution of ET-1 expression that was significantly increased in normal and vascular areas of UIP when compared to fibrosing areas (Table 2).

Association between mediators

A direct association was found between telomerase and ET-1 expression in vascular areas of UIP (r = 0.45; P = 0.03). An inverse association was found between telomerase and ET-1 expression in normal and minimal fibrosis areas of UIP (r = -0.42; P = 0.03). Telomerase and ET-1 expression were positively associated with vascular areas (r = 0.54; P = 0.004) and negatively associated with minimal fibrosis and normal areas (r = -0.44; P = 0.02) of UIP.

Association between mediators and fibrosis severity

Telomerase and ET-1 expression were inversely correlated with minimal fibrosis in normal and fibroblast foci areas (r = -0.38; P = 0.04 and r = -0.39; P = 0.03, respectively) and directly associated with severe fibrosis in vascular areas (r = 0.41; P = 0.03 and r = 0.54; P = 0.004) of UIP.

Association between mediators and functional tests

A direct association was found between diffusing capacity of the lung for oxygen/alveolar volume (DLCO/VA) and telomerase activity (r = 0.49; P = 0.01) in minimal fibrosis areas. An indirect association was found between DLCO and severe fibrosis in vascular areas displaying high ET-1 expression (r = -0.74; P = 0.02).

Survival analysis

The median follow-up was 42.70 months. Ten patients were still alive and 17 died from causes related to IPF. ROC curve analysis showed that the optimum cut-off point of expression was 14.55% for telomerase and 12.54% for ET-1. These cut-off points resulted in a significant difference in Kaplan-Meier curves. The 5-year survival rate of patients with minimal fibrosis in vascular areas displaying ET-1 ≤12.54% was 57.49 vs 31.13% in the group with minimal fibrosis displaying higher ET-1 levels (P < 0.01). The 5-year survival rate of patients with minimal fibrosis in vascular areas displaying telomerase ≤14.55% was 59.55 vs 28.88% in the group with minimal fibrosis and higher telomerase levels (P < 0.01).

Several combinations of clinical and morphological variables were analyzed to generate a mathematical model with an impact on IPF patient survival. The results of the best combination, determined by Cox model analysis, appear in Table 3. The following variables showed a significant impact on patient survival: gender, age, fibrosis severity, telomerase in UIP normal lung parenchyma, telomerase in minimal (fibroblastic foci) fibrosis, ET-1 in normal lung parenchyma, and minimal (fibroblast foci) fibrosis. Multivariate analyses showed a low risk of death for females, UIP normal lung parenchyma or minimal (fibroblastic foci) fibrosis displaying high telomerase and ET-1 expression in UIP normal lung parenchyma areas.

Table 3.

Cox proportional hazards regression to ascertain the individual contribution of clinical and morphological factors associated with survival and to compare adjusted survival between groups [-2 log likelihood = 28.90; chi-square = 38.26; P < 0.0001].

	β	SE	Wald test	P value	Exp (β)	95%CI for Exp (β)
						Lower	Upper
Gender	-3.74	1.48	6.38	0.01	41.94	2.31	761.02
Age (years)	1.02	0.34	8.94	0.003	2.78	1.42	5.43
UIP fibrosis			9.26	0.01
UIP severe (mural) fibrosis	-9.70	3.26	8.87	0.003	0.00	0.00	0.03
UIP minimal (fibroblastic foci) fibrosis	-12.04	3.98	9.16	0.002	0.00	0.00	0.01
Telomerase in UIP minimal fibrosis	-1.85	0.97	3.60	0.05	6.35	0.94	42.78
Telomerase in UIP normal lung parenchyma	6.05	2.35	6.62	0.01	0.002	0.00	0.23
Endothelin-1 in UIP normal lung parenchyma	-5.13	1.81	8.00	0.005	0.006	0.00	0.20
Endothelin-1 in UIP minimal (fibroblast foci) fibrosis	5.72	2.13	7.24	0.007	0.003	0.00	0.21

UIP = usual interstitial pneumonia; β = beta coefficient; SE = standard error; Exp (β) = exponential beta; CI = confidence interval.

Discussion

We investigated the participation of telomerase and the factors responsible for its activation or suppression in the fibrotic process found in pulmonary specimens obtained by surgical lung biopsy from patients with IPF. Specifically, we investigated telomerase and ET-1 expression in alveolar epithelium, endothelium, and myofibroblast cells present in normal lung parenchyma, minimal (fibroblast foci) fibrosis and vascular areas of UIP to verify the impact of these markers on the remodeling/fibrosis process. In the present study, telomerase and ET-1 were significantly increased in normal lung parenchyma and vascular areas compared to fibroblast foci areas of UIP. Unexpectedly, myofibroblast α-SMA+ expression was more prominent than smooth muscle caldesmon+ cells in vessels. Interestingly, we found a positive association between telomerase and ET-1 in UIP vascular areas, whereas a negative association was found between telomerase and ET-1 in normal lung parenchyma. A positive association was found between telomerase and myofibroblast α-SMA in UIP vascular areas. In addition, the expression of these mediators correlated with the measurement of fibrosis degree, suggesting that they exert their effect by influencing remodeling/fibrosis in UIP.

The current study revealed that telomerase increases myofibroblast activation in vascular areas of UIP. In fact, telomerase can be expressed by different types of cells and its presence indicates not only a persistent activity to maintain a supply of activated cells (e.g., fibroblasts), but is also important in the regulation of cell proliferation (17, 18), apoptosis (19,20), and cell differentiation (21,22). Although telomerase activity is a nonspecific signal of cell proliferation involving different cell types besides fibroblasts, this last assumption does not disagree with our results, which show an influence of telomerase expression on vessels and prognosis. Our findings agree with the study of Nozaki et al. (12) who demonstrated that bleomycin-induced lung injury and fibrosis are also known to induce telomerase activity in the affected lung tissue and isolated lung fibroblasts. The activity, however, is expressed only during the period of active fibrosis, and appears to be localized primarily in fibroblasts rather than myofibroblasts (23). The activity also declines with increased passaging in vitro, which is associated with increased differentiation into myofibroblasts. These studies suggest that the increased enzyme activity could be secondary to higher cell division rates and therefore the increase in telomerase activity would be secondary to the inflammatory and fibrotic process instead of its cause. The factors responsible for regulating telomerase activity in these tissues and cells are mostly unknown. More recent data from our group have shown that abnormal expression of telomerase/apoptosis limits type II alveolar epithelial cell replication in the early remodeling of UIP (24). In fact, we found that epithelial telomerase expression in normal lung parenchyma and fibroblastic foci areas was lower than in normal septal areas, probably reflecting the loss and shortening of telomeres activity. It is tempting to speculate about the existence of a specific checkpoint beyond which loss of telomerase activity due to injurious stimuli and/or inherited mutations may lead to loss of epithelial cells and promote genomic instability. Activation of telomerase after the onset of genomic instability, coupled with a repeatedly injurious microenvironment, may affect the physiological differentiation of normal alveolar epithelial progenitors towards a transformed mesenchymal phenotype, resulting in abnormal reepithelization in normal parenchymal areas and contributing to the fibrogenic process characterized by fibroblastic foci. Telomerase would inhibit the senescence of many normal cells, such as epithelial and mesenchymal cells, and most probably would increase the incidence of tumors and fibrosis in the patient. Thus, carcinogenesis and fibrogenesis may be similar processes respectively involving epithelial and mesenchymal (fibroblasts) neoplastic cells.

We also found that telomerase and ET-1 expression were positively associated with vascular tissue and negatively associated with normal lung parenchyma and fibroblast focus areas of UIP. ET-1 can also be synthesized by endothelial and epithelial cells, vascular smooth muscle cells, airway epithelium, granulocytes, and macrophages (25). Moreover, increased concentrations and biosynthesis of ET-1 were demonstrated in the lungs of patients with pulmonary hypertension and systemic sclerosis (26). Furthermore, preproendothelin-1 messenger RNA is elevated in airway epithelial cells, proliferating type II pneumocytes, endothelial cells, and inflammatory cells present in IPF (27,28). The increased cell population expressing ET-1 and telomerase may indicate a continuous stimulation and transformation of these cells. In fact, we found a more evident expression in epithelial cells of normal alveolar areas, reflecting not only the stress to which these cells were submitted but also a probable attempt to maintain a large population to avoid alveolar collapse, as previously demonstrated by our group (24). We also found that many epithelial cells and myofibroblasts present in fibroblast focus areas expressed ET-1 and telomerase showing a high activity of these areas. Actually, ET-1 plays a role in several stages of wound healing: stimulating fibroblast differentiation and collagen metabolism, promoting the wound healing response by the production of collagens IV, V and VII, involvement in the reconstitution of the basement membrane and appropriate wound closure and other unknown effects on different types of cells (29,30).

Our study has clinical consequences. In order to establish the relevance of these findings to the evolution of the patients, the mediators under study were evaluated as a function of fibrosis severity and physiological testing. We found a direct association between DLCO/VA and minimal fibrosis areas displaying high telomerase activity and between DLCO and severe fibrosis areas of UIP displaying high ET-1 expression. Although hypoxia is universal in IPF patients, decreases in DLCO caused by a defect in gas transfer certainly present in the pulmonary parenchyma of UIP, also contribute to hypoxemia. These functional changes observed are probably explained in part by the vascular involvement. The expression of these mediators probably represents another step in the wound healing process in IPF. We believe that high expression of these markers is crucial to maintain the activity of parenchymal injury by the regulation of compromised vascular tone. Many lines of evidence suggest that telomerase activation under chronic hypoxia may result in enhanced and sustained smooth muscle proliferation and vessel wall remodeling, findings characteristic of proliferative vascular disorders, such as pulmonary hypertension (31). On the other hand, the continuous ET-1 stimulus from the endothelium and smooth muscle of vascular cells activates the proliferation and the remodeling process of these vessels. Clearly, in our study poor patient survival was associated with high vascular expression of ET-1 and telomerase, suggesting that a major damage to vessels leads to more parenchymal alterations during the development of pulmonary fibrosis caused by maintained hypoxia.

In conclusion, telomerase activity increases endothelial and myofibroblast activation in early remodeling/fibrosis of IPF with impact on evolution, suggesting that strategies aimed at preventing the effect of these mediators may have a greater impact on patient outcome.