Agents associated with lung inflammation induce similar responses in NCI-H292 lung epithelial cells

Abstract

The aim of this study was to investigate an in vitro lung epithelial model for assessment of potential inhalation toxicity. The selected NCI-H292 lung carcinoma cell line is sensitive to cigarette smoke, responds in a similar manner to primary human lung epithelial cells and produces airway mucins. The following agents associated with inhalation toxicity were tested in the model: cigarette smoke total particulate matter, fly ash, bleomycin, lipopolysaccharide, vanadyl sulphate, diesel exhaust particles and carbon black. Polystyrene, poly-methylmethacrylate and dimethyl sulphoxide were used as negative controls. Response markers were chosen on the basis of reported injurious effects of lung toxicants in humans, and included pro-inflammatory cytokines, matrix metalloprotease-1, the airway mucin MUC5AC and heparin-binding epidermal growth factor-like growth factor. Markers were quantified at the mRNA and/or protein level in control and treated cells. Many of the selected markers were regulated in a similar manner by cigarette smoke and the other toxic substances in the H292 cell model. By comparison, the negative control agents were largely ineffective. We conclude that, with further validation, this assay may form part of a tiered strategy for toxicological assessment of inhaled agents prior to more complex primary cell models and animal inhalation studies.

Introduction

Inhaled toxicants can have acute effects in the lung, e.g. inflammation and oedema, or chronic effects leading to, for instance, lung cancer or chronic obstructive pulmonary disease (COPD) depending on the period of exposure. The potential of test agents to induce lung toxicity is frequently assessed by in vivo acute or sub-chronic rodent inhalation studies although these generally fail to model adequately endpoints such as cancer and chronic bronchitis. The mutagenic potential of inhaled agents such as cigarette smoke can be determined by in vitro genotoxicity tests (e.g. the Ames and micronucleus tests). Also, in vitro cytotoxicity tests can identify hazards that may cause acute lung inflammation. By comparison, there are no validated in vitro tests with endpoints relevant to chronic lung inflammation, over-production of mucus and airway remodelling which are key pathological features of COPD. Therefore, this study investigates the ability of a lung epithelial model to respond to lung toxicants based on a mechanistic approach to COPD.

The A549 type II alveolar cell line is used widely in respiratory research, for example to compare the relative toxicity of inhaled agents (Hetland et al., 2004). However, some have reported that A549 cells are relatively insensitive to cigarette smoke compared to primary lung epithelial cells (Kode et al., 2006). Our own pilot studies indicated that A549 cells were unresponsive to cigarette total particulate matter compared to NCI-H292 mucoepidermoid pulmonary carcinoma cells and so the latter was selected as a model for further work. Cigarette smoke has been shown to induce IL-8, transforming growth factor-α and amphiregulin in both H292 cells and primary human lung epithelial cells indicating some functional similarities (Richter et al., 2002). For such a model to have predictive value it would be expected to respond in a similar manner to a number of inhaled agents that induce lung inflammation in animals and/or humans. Therefore, in addition to cigarette smoke in the form of mainstream smoke total particulate matter (TPM) we also tested a series of other agents associated with lung toxicity, namely fly ash (FA), vanadyl sulphate (VS, which accounts for a significant part of the biological activity of FA), diesel exhaust particles (DEP), carbon black (CB), bleomycin (BM) and lipopolysaccharide (LPS). Microspheres of polystyrene (PS) and poly-methylmethacrylate (PMMA), representing relatively inert materials, plus the TPM vehicle, dimethyl sulphoxide (DMSO), were used as negative controls.

A series of cellular response markers were chosen on the basis of reports associating them with human exposure to lung toxicants and their injurious effects. For example, increased levels of inflammatory mediators are reported in the lungs of smokers and COPD patients (Barnes, 2002, Takizawa et al., 2001). Neutrophilic inflammation, in particular, is a significant element of COPD. Therefore, the neutrophil chemoattractants interleukin (IL)-8 (CXCL8) and growth-related oncogene-α (GRO-α, CXCL1) were included as markers along with pro-inflammatory cytokines, namely IL-6, IL-1β, and IL-13, granulocyte macrophage colony stimulating factor (GM-CSF), eotaxin (CCL11), tumour necrosis factor (TNF-α, TNFSF1A) and regulated on activation, normal T-cell expressed and secreted (RANTES, CCL5). Emphysema, a component of COPD, is characterised by destruction of the alveolar walls. An excess of proteases in the lung, and/or a lack of anti-proteases is hypothesised to be responsible for the tissue damage and remodelling (Parks and Shapiro, 2001, Stockley, 1999). While macrophages are a major source of some matrix-metalloproteases (MMPs), such as MMP-9 and MMP-12, others are produced by epithelial cells and may also contribute to disease processes. In this study we determined the effect of the test agents on MMP-1, which is produced by epithelial cells (Elkington et al., 2005) and has been shown to be up-regulated in the lungs of COPD patients (Imai et al., 2001, Mercer et al., 2004), and monocyte chemotactic protein (MCP-1, CCL2), which is induced in rodent lungs following exposure to lipopolysaccharide (Jansson et al., 2004). Finally, chronic bronchitis is an important clinical endpoint in COPD. The over-production of the secreted, gel-forming mucins MUC5AC and MUC5B by goblet cells and glands, respectively, of the lung epithelium is an underlying cause of the productive cough of chronic bronchitis (Kirkham et al., 2002). Mucin gene expression is reportedly regulated by epidermal growth factor-receptor (EGF-R) signalling (Deshmukh et al., 2005). Hence, we included both MUC5AC and a key EGF-R ligand, heparin-binding epidermal growth factor-like growth factor (HB–EGF) (Higashiyama et al., 1992), in the panel of response markers. Finally, CYP1A1 and CYP1B1 were also used because these cytochrome P450 enzymes metabolise polyaromatic hydrocarbons, aromatic and heterocyclic amines in cigarette smoke and are up-regulated in the lungs of smokers (Kim et al., 2004, Willey et al., 1997).

This study shows that many markers of lung inflammation and disease in smokers respond to cigarette smoke in the H292 cell model. Moreover, this is the first study to compare the effects of a broad range of agents linked to lung toxicity in a single test system and demonstrate a similar mechanism of action.