Upregulation of pirin expression by chronic cigarette smoking is associated with bronchial epithelial cell apoptosis

Healthy non-smokers and healthy chronic smokers were recruited using local print advertisements. The study individuals are part of an ongoing project to assess gene expression in the human airway epithelium in regard to the chronic airway disorders associated with cigarette smoking [22‐24]. The study was approved by the Weill Cornell Medical College Institutional Review Board and written informed consent was obtained from each individual before enrollment in the study. The smokers had an approximate smoking history of 20 pack-yr and were in otherwise good health, with no evidence of respiratory tract infection, chronic bronchitis or lung cancer. Each individual had to complete an initial screening evaluation, which included a history of smoking habits, respiratory tract symptoms, and prior illnesses, a complete physical exam, chest radiograph, and pulmonary function tests. Routine screening blood and urine studies were performed, including urinary levels of nicotine and its derivative cotinine, and serum levels of carboxyhemoglobin to verify reported levels of smoking.

All individuals who met the inclusion and exclusion criteria underwent fiberoptic bronchoscopy with brushing of the 3^rd to 4^th order bronchi as previously described [22]. Smokers were instructed not to smoke the evening before undergoing bronchoscopy. A 1 mm disposable brush (Wiltek Medical, Winston-Salem, NC) advanced through the working channel of the bronchoscope was used to collect the airway epithelial cells by gently gliding the brush back and forth on the airway epithelium 5 to10 times in 10 different locations in the third branching of the bronchi in the right and left lower lobe of each individual. The cells were detached from the brush by flicking it into 5 ml of ice-cold LHC8 medium (GIBCO, Grand Island, NY). An aliquot of 0.5 ml was kept for differential cell count and for cytology; the remainder (4.5 ml) was processed immediately for RNA extraction. Total cell number was determined by counting on a hemocytometer. Differential cell count (epithelial vs inflammatory cells) was assessed on cells prepared by cytocentrifugation (Cytospin 11, Shandon Instruments, Pittsburgh, PA) stained with DiffQuik (Baxter Healthcare, Miami FL).

All analyses were performed using the Affymetrix HuGeneFL chip and associated protocol from Affymetrix (Santa Clara, CA). Total RNA was extracted from brushed cells using TRIzol (Life Technologies, Rockville, MD) followed by RNeasy (Qiagen, Valencia CA) to remove residual DNA, which yielded approximately 2 μg RNA from 10⁶ cells. First strand DNA was synthesized using the T7-(dT)₂₄ primer and converted to double stranded cDNA using Superscript Choice system (Life Technologies). cDNA was purified by phenol chloroform extraction and precipitation, and the size distribution was examined after agarose gel electrophoresis. The cDNA was then used to synthesize biotinylated RNA transcript using the Bioarray HighYield reagents (Enzo, New York, NY). This was purified by RNeasy (Qiagen) and fragmented immediately before use. The labeled cRNA was hybridized to the HuGeneFL GeneChip for 16 hr, and then processed by the fluidics station under the control of Microarray suite software (Affymetrix). The chip was then manually transferred to the scanner for data acquisition.

RNA levels for pirin were measured relative to 18s rRNA by real time quantitative PCR (TaqMan) with fluorescent TaqMan chemistry using the ΔΔCt method (PE Biosystems, Instruction Manual). TaqMan reactions for pirin were optimized and validated to show equal amplication efficacy compared to 18s rRNA using adult human lung RNA (Strategene, La Jolla, CA). Two sets of primers and probes were used, one to measure endogenous RNA (including 3' untranslated end), and one to measure both endogenous and adenovirus-produced pirin mRNA (which spans two exons and would not amplify genomic DNA). The endogenous specific pirin primers were: forward AATGGGTTTGAAAGGGCCA and reverse TCAAGACCTGCTCTTCCGCT, with probe AACCTGGAAATCAAAGATTGGGAACTAGTGGA. The endogenous and adenovirus-produced pirin primers were: forward CACGCTGAGATGCCTTGCT and reverse ACCATCTTCTCTGAGCTCCTCAA with probe CAGCCCATGGCCTACAACTGTGGGTTATA.

Three separate primary human bronchial epithelial (HBE) cell cultures were isolated from trachea and bronchi of donor lungs and seeded onto collagen-coated semi-permeable membranes (Millipore, Bedford, MA) and grown at the air-liquid interface [25]. The viability of the cells was confirmed before each experiment by measurement of transepithelial resistance [25].

Cigarette smoke extract (CSE) was prepared using a modification of the method used by Wyatt et al [26]. Four research grade cigarettes (2R4F, University of Kentucky) were bubbled into 50 ml of 1:1 DMEM:Ham F12 medium using a vacuum pump apparatus. The CSE was filtered through a 0.22 μm filter to remove particles and bacteria before use. Solutions of 10% and 100% CSE were prepared from this stock. The solution of CSE (15 μl) was applied to the apical surface of the HBE cells and RNA was isolated from the cultures at 2, 24, and 48 hr after CSE exposure using TRIzol (Life Technologies) followed by RNeasy (Qiagen). Samples were obtained in triplicate for each time point. Pirin RNA levels were measured by TaqMan RT-PCR using the primers and probe described above. Pirin expression levels relative to 18s rRNA were assessed. Each data point was generated from triplicate wells for each of the three separate cell lines.

To assess the relationship between up-regulation of pirin and the induction of apoptosis, an adenovirus (Ad) gene transfer vector coding for pirin was used to transfer the human pirin cDNA to human bronchial epithelial cells, and pirin expression and apoptosis were assessed over time. The recombinant Ad vectors AdPirin and AdNull used in this study are E1a^-, partial E1b^-, and partial E3^-, based on the Ad5 genome, with the expression cassette in the E1 position [27‐29]. The AdPirin expression cassette includes the cytomegalovirus early/intermediate enhancer/promoter (CMV), an artificial splice signal, the human pirin cDNA (obtained from A549 cells), and an SV40 stop/poly (A) signal. The AdNull vector is identical to the AdPirin vector, except that it lacks a cDNA in expression cassette [28]. The vectors were propagated, purified, and stored at -70°C [27].

AdPirin-induced apoptosis was assessed in the human airway epithelial BEAS-2B cell line [30]. BEAS-2B cells (ATCC, Rockville, Maryland) were grown on lysine coated coverslips in LHC-9 medium (Biosource International), Camarillo, CA) until they were 50 to 60% confluent. The cells were then infected with AdNull and AdPirin at varying concentrations [10³ and 10⁴ particle units (pu)].

Two assays were used to assess apoptosis: TdT-mediated dUTP nick end labeling (TUNEL) assay and cytoplasmic nucleosome ELISA. For the TUNEL assay, cells were fixed to the coverslips using 4% paraformaldehyde and then permeabilized with 0.2% Triton X-100 in PBS. Cells were equilibrated with equilibration buffer, nucleotide mix, and rTdT enzyme (Promega, Madison WI) for 60 min and then washed. DAPI nuclear counterstain was applied before cells were mounted onto slides and evaluated under fluorescent microscope. The percentage of apoptotic cells per 10× field were manually counted in 10 fields per slide. For the cytoplasmic nucleosome ELISA assay (Cell Death Detection ELISA, Roche, Indianapolis, IN), BEAS-2B cells were lysed with lysis buffer, centrifuged for 10 min at 200 × g to pellet nuclei. The supernatant (20 μl) was added to the immunoreagent containing anti-histone biotin and anti-DNA horseradish peroxidase (HRP). Sample wells were placed on shaker at 300 rpm for 2 hr, 23°C. 2, 2azino-di[3-ethylbenzthiazolin-sulfonate] (ABTS) solution was added and photometric analysis was measured at 405 nm, subtracted from background 490 nm. For each sample, the fluorescent value was normalized to the internal negative control of the experiment to generate an apoptotic index, which reflects the fold change in the number of apoptotic cells for experimental condition compared to control.

A ³²P-labeled pirin specific DNA probe was synthesized using strip EZ labeling kit (Roche, Indianapolis, IN) from a pirin cDNA template amplified from human genomic DNA. RNA electrophoresis was performed in 1% agarose gel followed by transfer to nitrocellulose membrane and UV crosslinking. The membrane was then hybridized with DNA probe and exposed to X ray film for 1 hr.

The microarray analysis was carried out in a data set previously reported using a total of 44 Affymetrix HuGene FL microarrays to assess left and right samples from 22 individuals, including 9 non-smokers and 13 smokers [22, 24]. These 44 microarrays passed quality control as assessed by the GeneSpring software (Silicon Genetics, Redwood City, CA). The smokers and non-smokers were comparable with respect to yield and percentage of non-epithelial cells. To eliminate genes not expressed in airway epithelium, or expressed at low levels, those genes that were called absent by the Microarray Suite software (Affymetrix) in all of the 44 microarrays were discarded before further analysis. The number of genes remaining (i.e., expressed on at least 1 of the 44 microarrays) was 4,512. Using this subset of genes, non-parametric statistical methods (GeneSpring software) were used to identify genes which were expressed at a higher or lower level in a significant number of smokers vs non-smokers. Of the 4,512 genes that were expressed, there were a total of 85 probesets that were significantly (p < 0.05) up-regulated and 13 probesets down-regulated in smokers compared to non-smokers. The 98 probesets were functionally annotated by manual review of public databases (e.g. Medline, Locuslink) into categories that described their cellular processes. Of these, 7 of the up-regulated genes were identified to be associated with apoptosis, including pirin, retinoic acid receptor responder 2, prostrate differentiation factor, insulin-like growth factor binding protein 5, bone morphogenic protein 7, carcinoembryonic antigen-related cell adhesion molecule 6, and S100 calcium-binding protein A10 (Table 1). Of these, only two (retinoic acid receptor responder 2 and insulin-like growth factor binding protein 5) had any known association with cigarette smoke exposure. All of the genes identified were involved in signal transduction and transcription factors. Pirin, which has been shown to be induced during stress to cause cell death [21, 31], was selected for further study because it had the highest fold change (3.12 smokers/non-smokers) amongst genes in this category.

The microarray data demonstrated (n = 18 samples from n = 9 non-smokers, n = 26 samples from n = 13 smokers) a significant up-regulation of pirin in the airway epithelium in the smokers (p = 0.002; Table 1, Figure 1A). Pirin expression levels are reproducible in samples from the right and left lungs of individuals, with no significant difference in pirin expression levels in the right lung versus the left lung (p > 0.15). To confirm the microarray data showing overexpression of pirin in the airway epithelium of smokers, pirin expression levels were assessed by an independent method using a subset (n = 6 samples from n = 3 non-smokers, n = 18 samples from n = 9 smokers) of the RNA samples studied by microarray analysis, for which there was adequate amount of RNA available. TaqMan RT-PCR confirmed that expression levels were significantly different between the two groups, reinforcing the validity of the observation with the microarray analysis that pirin mRNA levels are markedly elevated in the airway epithelium of smokers compared to non-smokers (p < 0.01; Figure 1B).

To further establish that cigarette smoke up-regulates pirin expression in human bronchial epithelium, primary human bronchial epithelial cells were exposed to cigarette smoke extract in vitro. Human bronchial epithelial cells were used, as they most closely mimic airway epithelial cells in their natural environment in vivo [25]. TaqMan PCR with pirin RNA specific primers was used to quantify the amount of mRNA produced in the cells. Forty-eight hours after exposure to either 10% or 100% CSE, there was a respective 1.4-fold increase in pirin RNA levels in the cells exposed to cigarette smoke extract compared to the control group cultured in media (p < 0.03, 10% and 100% compared to controls; Figure 2).

To test the hypothesis that up-regulation of pirin expression is linked to increased apoptosis in epithelial cells, an adenovirus vector expressing human pirin cDNA was used to modify the human bronchial epithelial BEAS-2B cell line to express high levels of pirin RNA. Northern analysis demonstrated AdPirin-specific up-regulation of the 1.1 kb pirin mRNA (Figure 3A). TaqMan assessment of the pirin mRNA levels showed 79- and 538-fold change in expression at 10³ and 10⁴ particle units of AdPirin per cell, respectively (p < 0.01; Figure 3B). The expression of pirin was time independent, with expression up-regulated >100-fold for AdPirin at 24 to 72 hr (Figure 3C). Using TdT-mediated dUTP nick end labeling (TUNEL) to assess apoptosis, there was an approximately 5-fold increase in the number of TUNEL positive BEAS-2B cells exposed to 10⁴ AdPirin compared to cells exposed to AdNull for 24 hr (p < 0.01; Figure 4A–D).

Confirmation of the TUNEL assay results were made using an ELISA against cytoplasmic nucleosomes. BEAS-2B cells were exposed to varying concentrations of AdPirin, AdNull and cigarette smoke extract (CSE) and evaluated for apoptosis and pirin RNA level. This assay, which uses an increase in fluorescent signal compared to the naive negative control to generate an apoptotic index, demonstrated a 19.3-fold increase for cells exposed to 10⁴ AdPirin (p < 0.01), 2.1-fold increase for cells exposed to 10³ AdPirin (p < 0.01) and 7.9 fold increase for cells exposed to 50% CSE (p < 0.01, Figure 5A). There was no significant increase in apoptotic cells compared to naive negative control for all other conditions. In this experiment, pirin RNA levels were increased 2.3 (p < 0.01) and 1.7-fold (p < 0.03) for BEAS-2B cells exposed to 10% and 50% CSE, respectively. BEAS-2B cells exposed to10³ and 10⁴ AdPirin demonstrated a 2.0 (p < 0.04) and 133.7-fold (p < 0.01) increase in pirin RNA expression level, respectively (Figure 5B).

One of the main functions of the airway epithelium is to provide a physical barrier to protect the underlying tissue from the harmful effects of cigarette smoke [1, 2, 4, 5]. The inflammatory response seen in COPD is initiated when the airway epithelium is disrupted by chronic exposure to cigarette smoke [4]. This breach in epithelial integrity triggers the innate immune response which mobilizes neutrophils, eosinophils, macrophages, natural killer cells and mast cells into the subepithelial space [2, 9]. When the stress of cigarette smoke is exerted over years, the persistent inflammation in the central airways leads to the pathologic finding of chronic bronchitis. As the inflammation extends to the small airways (between 4^th to 14^th generation bronchi), inflammatory exudates can accumulate in the lumen of the bronchi resulting in airflow obstruction [9].

Epithelial permeability, as measured by ^99mTC-DTPA lung clearance in human subjects, has been shown to be increased in chronic smokers compared to non-smokers [7, 32]. Studies using animal models have shown that epithelial cell permeability can increase in response to cigarette smoke with 30 min and 6 hr of exposure [8, 11]. This is followed by an increased number of neutrophils and mast cells in the airway wall 6 hr after cigarette smoke exposure. The mechanism for increased permeability is unclear, but prior studies have shown that damage to the epithelial cell itself, and to a lesser extent, disruption in the tight junctions between epithelial cells are partly responsible [10, 11]. Consistent with the bronchial epithelial cell studies, in vitro cellular studies have shown that alveolar epithelial cell viability decreases after exposure to cigarette smoke in a time and concentration dependent manner [35].

The mechanism of cell death for the epithelium in response to cigarette smoke may involve some necrosis, but apoptosis is likely the dominant mechanism [18]. Mouse models of smoking induced airway disease have demonstrated that apoptosis can occur in the airways in response to cigarette smoke [16, 17]. Due to the efficient nature of the apoptotic pathway, allowing for rapid clearance of dying cells, it can be technically difficult to assess for apoptosis from in vivo studies of the airway in humans.

Several in vitro studies have shown that apoptosis can occur in response to cigarette smoke extract (CSE) in a variety of cell types, including airway epithelial cells, alveolar macrophages, endothelial cells, alveolar epithelial cells, and fibroblasts [13‐15]. In contrast, Wickenden et al [19] found that CSE induced necrosis of a variety of cell types, while inhibiting the activation of the caspase system and preventing apoptosis. This discrepancy may be explained by the observation that at low concentrations, CSE induces apoptosis, while at higher concentrations, necrosis is induced [18]. Liu et al [20] reported that CSE caused DNA damage and initiated DNA repair processes, but not apoptosis, in bronchial epithelial (BEAS-2B) cells. The reason their conclusions differ from ours is likely due to differences in the concentration of cigarette smoke extract used. The maximum concentration that Liu [20] et al used was 10% CSE generated from 1 cigarette bubbled into sterile water. Similarly, in our experiment we did not observe apoptosis in BEAS-2B cells exposed to 10% CSE, but did observe apoptosis in cells exposed to 50% CSE, generated from 4 cigarettes bubbled into LHC-9 media. Taking all of the in vitro studies into consideration, one potential explanation is that at low concentrations, CSE causes DNA damage that is reparable until a threshhold concentration is reached, after which the cell initiates apoptosis. Very high levels of CSE likely cause overt necrosis.

Pirin was first identified by Wendler et al [33], where it was found to interact with nuclear factor 1/CCAAT box transcription factor (NF1/CTF1). Pirin was found to be exclusively nuclear, expressed in most tissues and highly conserved among mammals. The N-terminus shares significant homologies with proteins from plants, fungi and prokaryotes.

Pirin was later found to bind to the oncoprotein B-cell lymphoma 3-encoded (Bcl-3), a member of the IκB family[36]. Bcl-3 interacts with the anti-apoptotic protein NF-κB, by binding to the p50 and p52 [36]. Recently, the crystal structure of pirin has been elucidated and was found to have a single Fe²⁺ ion in the highly conserved N-terminal region [34]. This metal ion may be necessary for its interaction with Bcl-3 and NF-κB. Bcl-3 can either inhibit or enhance NF-κB DNA binding and transcription depending on its binding to other transcription co-factors.

Pirin stabilizes the formation of quaternary complexes between Bcl-3, NF-κB, and its DNA target. While it is not clear exactly how pirin influences the interaction between Bcl-3 and NF-κB, based on our results and others, it seems possible that pirin directs NF-κB DNA binding towards a pro-apoptotic response. Elucidating this mechanism will require further investigation to understanding the signaling that results from pirin expression. This may not be possible through gene expression studies alone, since several apoptotic pathways are induced by mechanisms other than up-regulation of mRNA levels of pro-apoptotic genes. This might explain why there was no observed upregulation of traditional apoptotic or inflammatory pathway genes, e.g., caspases and Bcl-2, in our microarray data obtained from human airway epithelium. The lack of up-regulation of pro-inflammatory genes in the airway epithelium of smokers is also supported by the transcriptome analysis of Spira et al [37]. One possible explanation for this observation is that many of these inflammatory mediators may be released primarily by non-epithelial cells.

Evidence for pirin involvement in apoptosis comes from experiments in tomatoes [21]. When tomato cells in suspension are treated with the topoisomerase inhibitor, campothecin, a potent initiator of cell death, le-pirin, an ortholog to human pirin, mRNA levels are dramatically increased. In the cyanobacterium Synechocystis, pirin orthologs are highly induced under conditions of severe salt stress [28].

We originally noted pirin to be up-regulated in microarrays of phenotypically normal individuals that smoke [24], a finding also made by Spira et al [37], in their study of genes up-regulated in the airway epithelium of smokers. Interestingly, Spira et al [37] observed that the pirin levels returned to normal for ex-smokers, supporting observation in the present study that pirin is influenced by acute exposure to cigarette smoke.