Heme Oxygenase-1, a Potential Biomarker of Chronic Silicosis, Attenuates Silica-induced Lung Injury

Inhalation of crystalline silica for prolonged periods can lead to silicosis, an inflammatory disorder (1, 2). The acute manifestations of this disease include inflammation and apoptosis, with attendant destruction of the lung tissue (3), whereas chronic silicosis is characterized by progressive fibrosis and emphysematous changes (4). Although lung function is severely impaired in patients with advanced disease, airway obstruction is common even in silica-exposed workers with no radiological abnormalities (4, 5). Defining the molecular mechanisms triggered by exposure to silica could facilitate early diagnosis by identifying markers of early disease, and could elucidate novel therapeutic strategies.

Crystalline silica induces the production of reactive oxygen species (ROS) (2), which play a key role in the development of silicosis (6) via the mitogen-activated protein kinase (MAPK) pathway (7). ROS cause direct tissue injury and activation of alveolar macrophages to undergo apoptosis (8), enhance the synthesis of proinflammatory cytokines (9), and augment the induction of inflammatory responses via the MAPK-dependent pathways (10). Silica-induced oxidative DNA damage, reflected by an increase in 8-hydroxydeoxyguanosine (8-OHdG), is associated with an increased risk of carcinogenesis (11). ROS production is down-regulated by elements of the extracellular signal-regulated kinase (ERK) pathway, which trigger increased transcription of antioxidants (12).

One of the antioxidants that protect against ROS-induced airway inflammation is heme oxygenase (HO)-1, an enzyme that degrades heme into bilirubin, Fe^sup 2+^, and carbon monoxide (CO) (13, 14). Lung epithelial cells, smooth muscle cells, macrophages, and endothelial cells can all produce HO-1 (15-17). The protective activity of HO-1 was established in several model systems. For example, adenovirus-mediated gene transfer of HO-1 cDNA suppresses lipopolysaccharide-induced lung injury (18), influenza-induced pneumonia (19), and bleomycin-induced pulmonary fibrosis (20) in murine models. The current work explores whether HO-1 can play a role in attenuating silica-mediated lung injury.

We evaluated HO-1 protein levels in sera from patients with silicosis and in a murine model of that disease. To clarify the role of HO-1, we assessed the effects of hemin and zinc protoporphyrin (ZnPP) (an inducer and an inhibitor of HO-1, respectively) in murine silicosis. Abbreviated results from this study were previously reported in abstract form (21, 22).

METHODS

For a detailed description of the methods, see the online supplement.

Patients

Forty-six male patients (age range, 56-83 yr) were diagnosed as having silicosis on the basis of their occupational history and the International Labor Organization (ILO) International Classification of Radiographs of Pneumoconioses (23). Radiographic opacities were graded on a 12-point scale using the ILO system (score of 1 for 0/- and 12 for 3/+) (24). More details are shown in the online supplement. Twenty-seven male patients with chronic obstructive pulmonary disease (COPD) (age range, 60-87 yr) and 42 healthy male volunteers (age range, 55-80 yr) served as age-matched disease and normal control subjects, respectively. Subjects were defined as smokers if they had smoked at least 1 pack-year, and as ex-smokers if they had stopped smoking for at least 1 yr (25). Serum HO-1 and 8-OHdG were measured by enzyme-linked immunosorbent assay (ELISA). C-reactive protein (CRP) in serum was analyzed by immunonephelometry (26). Spirometry was performed using standard protocols according to the American Thoracic Society recommendations (27). Local HO-1 expression was examined by immunohistochemistry in the lungs from 10 patients with silicosis (8 autopsies and 2 surgical specimens) and from 17 control autopsy cases with no occupational lung diseases. In silicosis samples, HO-1-positive cells were counted microscopically in 50 high-power fields (HPF) (approximately equivalent to 10 mm^sup 2^) containing areas of silicotic fibrosis; in control samples, this was done in 50 randomly selected HPF. The study was approved by the local Institutional Review Board. All patient studies were performed after obtaining written informed consent, except for the studies on autopsy samples, for which the need for informed consent was waived.

Murine Silicosis Model

Six-week-old male BALB/c mice were used for all animal experiments. All studies were approved by the Institutional Animal Care and Use Committee at Yokohama City University. Under anesthesia with ketamine (80 mg/kg; Sigma-Aldrich, St. Louis, MO) and xylazine (10 mg/kg; Sigma), 100 mg/kg of sterilized crystalline silica (Min-U-Sil-5, U.S. Silica, Berkeley Springs, WV) in 100

Statistics

Values were expressed as mean ± SEM. In the mouse studies, only nonparametric methods were used for statistical analysis. Comparisons between multiple independent groups were made by Kruskal Wallis test followed by Mann-Whitney U test. In the human studies, comparisons between study groups were made by using one-way ANOVA followed by post hoc analysis with the Bonferroni test. The relationships between serum HO-1 and other laboratory data were examined by using linear regression analysis with the method of least squares. Probability values of less than 0.05 were considered statistically significant.

RESULTS

Serum HO-1 Levels Are Increased in Patients with Mild Silicosis

The characteristics of the 46 patients with silicosis, 27 patients with COPD, and 42 normal control subjects are shown in Tables 1 and 2. The patients with silicosis and COPD had longer smoking histories than did control subjects, and the majority of patients were ex-smokers. Vital capacity (VC) and/or forced expiratory volume in one second (FEV^sub 1^) were < 80% of predicted values (30) in 21/46 patients with silicosis, a significant reduction when compared with control subjects. Patients with COPD had reduced FEV^sub 1^ but not VC when compared with control subjects but not patients with silicosis. The silicosis group included patients with disease of varying radiologic severity (28 simple silicosis and 18 complicated silicosis) (Table 2). Of the 46 patients, 43 had an ILO profusion of = 1/0 (with or without large opacities) and four had large opacities only. Some of the patients with silicosis or COPD were being treated with theophylline (32.6% and 40.7%, respectively), short-acting ß-agonists (28.3% and 25.9%, respectively), anticholinergics (26.1% and 37.0%, respectively), and/or inhaled corticosteroids (8.7% and 22.2%, respectively).

Serum CRP levels were significantly higher in patients with silicosis than those with COPD or control subjects (Table 1). Serum HO-1 levels were also significantly higher in patients with silicosis than the other groups (Table 1). Patients with silicosis with only mildly impaired respiratory function showed the highest serum HO-1 levels. Thus, increased HO-1 levels correlated positively with FEV^sub 1^ (r = +0.390, p = 0.008) and VC (r = +0.296, p = 0.046) (Table 3). Other factors examined, including age, duration of silica exposure, smoking history, or serum CRP, did not correlate with serum HO-1 levels. Among patients with silicosis, occupational history did not correlate with serum HO-1 levels. In contrast to patients with silicosis, lung function parameters and HO-1 levels did not correlate in patients with COPD (data not shown).

Serum 8-OHdG levels were significantly higher in silicosis patients than those with COPD or healthy control subjects (Table 1). Furthermore, a significant negative correlation was found between serum HO-1 and 8-OHdG levels (r = -0.311, p = 0.035) (Table 3).

Localization of HO-1 Protein in Silicosis Lungs

Since elevated serum HO-1 was preferentially found in patients with silicosis, we examined HO-1 expression in the lung lesions. Surgical specimens from 2 patients with silicosis and with lung cancer, 8 autopsied subjects with silicosis, and 17 control subjects were studied (Table 4). HO-1 protein was detected by immunohistochemical staining in bronchial epithelial cells (Figure 1) and macrophages (Figure 2) of granulomatous tissue from areas free of tumor. Silica particles were consistently associated with lesions containing HO-1-expressing cells (Figure 1). The number of HO-1-expressing cells was significantly higher in patients with silicosis than in control subjects (HO-1-positive cell number/cm^sup 2^; 409 ± 72 versus 39 ± 12, p < 0.05) (Table 4).

HO-1 Expression in Murine Silicosis

A murine model of silicosis was studied to clarify the role of HO-1. Disease was induced by administering crystalline silica intratracheally to BALB/c mice. This resulted in acute leukocyte infiltration followed by granuloma formation similar to that observed in human silicosis (Figure E1 in the online supplement). Immunohistochemical analysis showed that HO-1 expression had increased in bronchial epithelial cells within 1 d of silica exposure (Figure 3). HO-1-expressing cells were also found in granulomatous tissues during the chronic phase of disease (Figure 4). Consistent with results from the human lung biopsies, HO-1-expressing cells were found close to silica particles deposited in the lungs throughout the clinical course of disease in mice (Figure 4). Semiquantitative analysis of HO-1 protein by immunoblotting indicated that HO-1 expression in the lungs gradually increased as disease progressed (Figures 5A and 5B).

Plasma 8-OHdG concentrations gradually increased through 12 wk, and then decreased to the baseline by 24 wk (Figure 5C). This contrasts with the persistent elevation in HO-1 expression late in the disease process (Figures 5A and 5B).

Super-induction of HO-1 Attenuates Acute Silicosis in Mice

Consistent with the acute inflammatory responses caused by silica in the lower respiratory tract (31), neutrophils were the dominant cells in BALF during the first 3 d after silica exposure; later macrophages and lymphocytes became dominant (Figure 6).

To evaluate the impact of HO-1 on the development of silicosis, mice were treated with hemin, a potent HO-1 inducer, or ZnPP, an inhibitor of HO. In normal mice, a single administration with hemin induced sustained HO-1 expression for at least 7 d, whereas ZnPP had no effect (Figure E2). In mice with silicosis, HO-1 expression in the lung was enhanced by hemin and suppressed by ZnPP (Figures 7A and 7B). Infiltrating cell numbers were decreased by hemin but increased by ZnPP (Figure 6A). Hemin pretreatment reduced neutrophil infiltration and the subsequent accumulation of macrophages, whereas ZnPP pretreatment enhanced the influx of neutrophils and the subsequent accumulation of lymphocytes (Figures 6B-6D). Histopathologic analysis showed that cellular infiltration into alveolar spaces was suppressed by hemin but augmented by ZnPP (Figure E1A). Thus, hemin suppressed acute inflammation after silica exposure, whereas lesion development was accelerated by ZnPP.

The silica-induced increases in plasma 8-OHdG levels were markedly increased by pretreatment with ZnPP and suppressed by hemin (Figure 7C). There was an inverse correlation between lung HO-1 and plasma 8-OHdG levels (Figures 7A-7C). This observation is consistent with the finding that serum HO-1 levels were not elevated in patients with silicosis with high serum levels of 8-OHdG.

DISCUSSION

More than 10,000 patients with silicosis are currently followed up in Japan, although the incidence has been gradually decreasing due to efforts to prevent further silica exposure (32). Unfortunately, silicosis continues to occur in many developing countries (due to a paucity of preventive medicine) and in some advanced countries (especially in the fields of chemical and civil engineering, where workers are exposed to silica-containing dusts) (33). Although this disease can be prevented by avoiding silica exposure, equally important is the development of therapy for those with pre-existing disease.

This work is the first to demonstrate that HO-1 is synthesized in the lungs of patients with silicosis, thus contributing to a significant elevation of serum HO-1 levels in patients. Our studies suggest that the lung is a major source of circulating HO-1, since silicotic nodules contain abundant HO-1-expressing cells. In contrast, peripheral blood mononuclear cells (PBMC) are unlikely to be the source of serum HO-1, both because HO-1 expression was not elevated in PBMC from patients with silicosis and HO-1 protein levels in serum did not correlate with protein production by PBMC (r = +0.185, p = 0.301) (unpublished observations). If serum HO-1 derives primarily from lung lesions, it could provide a novel biomarker for evaluating the severity of silicosis.

Although HO-1 expression in the lung is increased in diseases characterized by increased oxidative stress (34, 35), changes in serum HO-1 levels have not been reported in pulmonary disorders. We previously demonstrated that serum HO-1 levels were elevated in patients with adult-onset Still's disease and hemophagocytic syndrome, two systemic inflammatory disorders characterized by hyper-ferritinemia and macrophage activation, and that serum HO-1 levels correlate closely with disease activity (36).

Our present results indicate that serum HO-1 levels correlate significantly with FEV^sub 1^ and VC in patients with silicosis. A previous necropsy study revealed that pathologic changes in the lung parenchyma and large airways were associated with impairment of lung function in patients with silicosis (37). Although involvement of ROS and increased circulating carbon monoxide levels are present in patients with COPD (38, 39), serum HO-1 levels are not elevated in those patients (whose lung function is comparable to those with silicosis). Rather, data from human lung samples and from experimental silicosis in mice indicates that silica exposure directly and persistently induces HO-1 expression in granulomatous lung tissue.

The present study examined HO-1 expression in human lung samples. Considerable care was taken to ensure that studies involving these samples were free from artifacts (potentially introduced due to the terminal stage of illness, postmortem phenomena, and/or extrapulmonary disease). Several groups have shown that HO-1 can be accurately detected in autopsy material (40-42). Moreover, there were no differences between surgically resected versus autopsied lung tissue with respect to levels of HO-1 expression (HO-1-positive cell number/cm^sup 2^; 414 ± 204 versus 407 ± 88), suggesting that the terminal stage of disease did not affect expression of this protein. Finally, lungs from patients with various extrapulmonary diseases expressed HO-1 at levels far below those found in silicotic lungs.

There is accumulating evidence that HO-1 helps protect the host from progression of certain inflammatory diseases (43). In the current study, super-induction of HO-1 by administration of hemin suppressed silica-induced leukocyte infiltration in the lung, whereas treatment with ZnPP augmented it. When hemin was administered to normal mice (without silica exposure), HO-1 expression was induced in the lung, and persisted for at least 7 d (Figure E2). ZnPP did not alter HO-1 expression level in the absence of silica exposure, but did suppress HO-1 expression, leukocyte infiltration, and subsequent pathologic changes in mice with silicosis (Figure 7). This discrepancy suggests that ZnPP does not directly act on HO-1 expression but instead suppresses a mediator induced by silica that stimulates HO-1 expression.

ROS play a critical role in the development of silicosis (2). In addition to causing direct tissue toxicity, ROS damage nuclear and mitochondrial DNA, leading to increased generation of 8-OHdG (44). In the murine silicosis model, plasma 8-OHdG levels were elevated from 1-12 wk after particle exposure and then returned to baseline by 24 wk. In contrast, HO-1 expression was persistently increased in lungs of mice with silicosis. The dissociation between lung HO-1 expression and plasma levels of 8-OHdG may have several explanations. First, because ROS activity reflects the balance between oxidative stress and antioxidant systems (45), it is likely that antioxidant activity exceeds silica-induced ROS after 12 wk in this model. Second, by 12 wk, some of the silica particles may have left the lung via the lymphatics (accumulation of silica in the draining lymph nodes was confirmed histologically). The negative relationship between 8-OHdG and HO-1 was particularly evident in mice treated with hemin or ZnPP. Hemin enhanced the expression of HO-1 while reducing 8-OHdG levels compared with controls, whereas ZnPP treatment had the opposite effects. Indeed, serum 8-OHdG levels were increased in patients with silicosis having normal serum HO-1 levels. Previous studies suggest that ROS are generated via the Fenton mechanism after silica exposure (2), and that HO-1 acts as an important scavenger of ROS (46, 47), probably via the heme degradation products bilirubin and CO (43).

In summary, the present study shows that HO-1 is persistently expressed in the lung lesions of patients with silicosis. This appears to reflect the induction of ROS by silica, leading to an elevation in serum HO-1 levels. The increased HO-1 can protect the host by suppressing silica-induced ROS activity. Thus, upregulation of HO-1 may represent a novel strategy for the treatment of silicosis.

Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Acknowledgment: The authors thank Dennis M. Klinman (Center for Biologies Evaluation and Research, FDA) for helpful discussions and Toshihiro Shibata, CT (Division of Pathology, Rosai Hospital for Silicosis) for technical help. The authors also thank the nursing and laboratory staff of Rosai Hospital for Silicosis.

© 2006 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.