This Article Abstract Full Text (PDF) Online Supplement All Versions of this Article: 2006-0056OCv1 35/6/651    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lonergan, K. M. Articles by Lam, W. L. Search for Related Content PubMed PubMed Citation Articles by Lonergan, K. M. Articles by Lam, W. L.

Identification of Novel Lung Genes in Bronchial Epithelium by Serial Analysis of Gene Expression

Cancer Genetics and Developmental Biology, Department of Cancer Imaging, Canada's Michael Smith Genome Sciences Centre, British Columbia Cancer Research Centre; and Computer Science, University of British Columbia, Vancouver, British Columbia; and Ontario Cancer Institute/Princess Margaret Hospital, Toronto, Ontario, Canada

Correspondence and requests for reprints should be addressed to Kim Lonergan, British Columbia Cancer Research Centre, 675 West 10th Avenue, Vancouver, BC, V5Z 1L3 Canada. E-mail: klonergan{at}bccrc.ca

	Abstract

Top Abstract MATERIALS AND METHODS RESULTS AND DISCUSSION References

 
A description of the transcriptome of human bronchial epithelium should provide a basis for studying lung diseases, including cancer. We have deduced global gene expression profiles of bronchial epithelium and lung parenchyma, based on a vast dataset of nearly two million sequence tags from 21 serial analysis of gene expression (SAGE) libraries from individuals with a history of smoking. Our analysis suggests that the transcriptome of the bronchial epithelium is distinct from that of lung parenchyma and other tissue types. Moreover, our analysis has identified novel bronchial-enriched genes such as MS4A8B, and has demonstrated the use of SAGE for the discovery of novel transcript variants. Significantly, gene expression associated with ciliogenesis is evident in bronchial epithelium, and includes the expression of transcripts specifying axonemal proteins DNAI2, SPAG6, ASP, and FOXJ1 transcription factor. Moreover, expression of potential regulators of ciliogenesis such as MDAC1, NYD-SP29, ARMC3, and ARMC4 were also identified. This study represents a comprehensive delineation of the bronchial and parenchyma transcriptomes, identifying more than 20,000 known and hypothetical genes expressed in the human lung, and constitutes one of the largest human SAGE studies reported to date.

Key Words: bronchial epithelial • ciliogenesis • expression profile • lung parenchyma • SAGE

The bronchial epithelium is a pseudo-stratified structure, consisting of specialized cell types including basal cells, goblet cells, and ciliated columnar cells, and plays an active role in airway defense by protecting the respiratory tract from infection and damage induced by environmental toxins. Moreover, maintenance of tissue architecture and cellular polarity is crucial for proper lung function. Disorders such as cystic fibrosis and primary ciliary dyskinesia originate from impaired ionic transport across the bronchial epithelium and impaired ciliary function, respectively (1–3).

Several large-scale gene expression studies have been published that describe disease states of the lung such as chronic obstructive pulmonary disease (COPD), emphysema, and non–small cell lung cancer (NSCLC), as well as response to microbial exposure in bronchial epithelial primary cell cultures (4–11). In a recent study, 2,382 genes were identified to be consistently expressed in large-airway epithelial cells of healthy never-smokers, as were 97 genes induced by smoking (12). Despite these informative studies, knowledge of gene expression in the bronchial epithelium remains limited.

An improved understanding of the bronchial epithelium transcriptome, specifically that exposed to tobacco smoke and therefore at an increased risk of malignant transformation and other lung pathologies, should serve as a baseline to facilitate an understanding of molecular mechanisms underlying central airway disorders of diverse etiologies. In this study, we have determined the gene expression profile of 19 bronchial epithelial samples from current or former smokers by serial analysis of gene expression (SAGE) (13), constituting one of the largest human expression studies reported to date. Significantly, we were successful in constructing SAGE libraries from human bronchial epithelial cells isolated by endoscopic brushing of the central airways. This was achieved without the need for either cell culturing or linear amplification of RNA. SAGE profile comparisons defined bronchial gene expression relative to that of lung parenchyma, and offered the potential for discovery of alternate transcript variants in known lung genes. Further comparison with profiles derived from various normal human tissues revealed novel bronchial-enriched genes, including those associated with innate defense and ciliogenesis.

	MATERIALS AND METHODS

Top Abstract MATERIALS AND METHODS RESULTS AND DISCUSSION References

 
Specimens
Bronchial epithelial specimens used in this study were obtained from segmental and subsegmental bronchi by brushing with a 3-mm teflon brush with a sheath (Hobbs Medical, Stafford Springs, CT) during autofloresent bronchoscopy. The areas brushed were without abnormalities for moderate dysplasia or worse pathology as determined by combined autoflouresent and white light bronchoscopy using the LIFE-Lung device (Xillix Technologies Inc., Richmond, BC, Canada) (14). The brush (with adhering tissue) was immediately immersed in RNAlater (Ambion, Austin, TX) and stored at –85°C. Cytologic examination indicated that on average, specimens consisted of over 90% bronchial epithelial cells, with the remainder consisting of leukocytes and alveolar macrophages. All individuals were white, ranging in age from 54–72 yr, and were either current or former smokers, with smoking exposure ranging from 30–100 pack-years (Table 1). No individuals were known to have asthma, based on clinical history and lung function testing. This study was approved by the Review of Ethics Board of the Ministry of British Columbia.

SAGE Library Construction and Sequence Processing
Bronchial brushing specimens in RNAlater were diluted 2-fold with PBS, and cells were collected by centrifugation and homogenized in lysis/binding solution (100 mM Tris-HCl, pH 7.5, 500 mM LiCl, 10 mM EDTA, pH 8.0, 1% LiDS, 5 mM dithiothreitol). The resultant lysate was used directly for bronchial epithelial (BE) SAGE library construction.

For the lung parenchyma (LP) libraries, RNA was isolated from eight individuals by guanidium isothiocyanate and phenol/chloroform extraction (15). Each of two libraries was constructed from RNA pooled from four specimens in equal amounts; 19 µg of total RNA were used in constructing each library.

All SAGE libraries were constructed according to the MicroSAGE protocol, using NlaIII as the anchoring enzyme and BsmFI as the tagging enzyme (13; detailed at www.sagenet.org). Reagents, primers and restriction enzymes were purchased from Dynal Biotech (Brown Deer, WI), Integrated DNA Technologies (Toronto, ON, Canada), Fisher Scientific (Nepean, ON, Canada), and New England Biolabs (Pickering, ON, Canada). The I-SAGE kit and Platinum Taq polymerase were purchased from Invitrogen Life Technologies (Burlington, ON, Canada).

On average, 10⁵ SAGE tags, excluding linker and duplicate ditags, were sequenced per library (Table 1). For normalization, tag counts were scaled to 10⁶ tags/library and presented as tags per million (TPM). Tag-to-gene mapping was according to the SAGE Genie database (16); cgap.nci.nih.gov/SAGE), with reference to SAGEmap.

Cluster Analysis
To evaluate the degree of similarity of the lung libraries to those generated from multiple tissue types, the 19 BE and the 2 LP libraries from this study were compared with 14 libraries derived from normal lung, brain, colon, breast, and prostate tissue selected from the GEO (Gene Expression Omnibus) data repository at SAGEmap (17). For cluster analysis, the 300 most abundant tags (representing 1/3 of the total tag count) were retained from each library, yielding a merged list of 1,610 unique tags. A correlation coefficient matrix between all pairs of libraries was generated, processed through the statistical software package R (18), and then clustered based on the single-link hierarchical clustering algorithm.

Identification of Bronchial-Enriched Genes
To identify genes preferentially expressed in bronchial epithelium, we compared the 19 BE libraries with those generated from a variety of normal tissue types including brain, breast, colon, prostate, kidney, leukocyte, skin, peritoneum, liver, heart, and spinal cord, all retrieved from the GEO data repository at SAGEmap.

All 19 BE libraries were grouped together, and the mean and SD were computed for each tag. For some tags, the SD may be considerable relative to the mean. Thus, a simple fold change may be misleading. One improvement is to use the standard deviation–adjusted ratio (SD-Adj Ratio), which in this case is defined as (bronchial mean – bronchial SD)/(non-lung mean + non-lung SD). This ratio, by design, is conservative because it always gives a value no greater than the simple fold change. Bronchial-enriched tags were identified according to their SD-Adj Ratio, and the results were sorted in descending order.

RT-PCR Analysis
RNA was isolated from bronchial brushing specimens using Trizol reagent (Invitrogen) and treated with DNase I (Roche Diagnostics, Laval, PQ, Canada) before cDNA synthesis. Forty nanograms of total RNA was converted into cDNA using SuperScript II reverse transcriptase and oligo dT₂₀ primer according to manufacturer's recommendations (Invitrogen). Control reactions were set up in parallel, omitting the reverse transcriptase. For validation of lung-specificity of expression, human MTC Multiple Tissue cDNA Panels I and II (cat# 636742, 636743; Clontech, Mississauga, ON, Canada) were used in addition to bronchial epithelial cDNA. PCR was performed for 30 cycles using Platinum Taq DNA Polymerase (Invitrogen) and gene-specific primers (Alpha DNA, Montreal, PQ, Canada). Primers were selected from sequence close to the SAGE tag and designed to generate 100- to 200-bp amplicons (see Table E2 in the online supplement). PCR products were resolved by agarose gel electrophoresis, and visualized by ethidium bromide staining.

For quantitative RT-PCR, total RNA was converted into cDNA using the High-Capacity cDNA Archive kit (cat# 4322171; Applied Biosystems, Streetsville, ON, Canada), and gene-specific quantitative PCR was performed using TaqMan Universal PCR Master Mix and TaqMan primers (cat# 4326708; Applied Biosystems), according to manufacturer's recommendation. Beta-actin gene expression was used as an endogenous control (primer product code 4352935E). Primer product codes for test genes were as follows: ARMC3, Hs00330456_ml; Blu, Hs00210720_ml; MDAC1, Hs00373644_ml. The reactions were run on an iCycler iQ Real-Time PCR Detection System (Bio-Rad, Mississauga, ON, Canada), and method of analysis was the delta-delta CT.

SFTPB Transcript Variant 2-Short: Cloning and cDNA Sequencing
The 3'-terminal region of SFTPB (surfactant, pulmonary-associated protein B) transcript variant 2-short was identified by differential display (DD), based on a previously described method (19). Briefly, poly (A)⁺ RNA isolated from lung parenchyma was primed and amplified using C-anchored oligo dT with a HindIII site at the 5'-end (5'-TG CCGAAGCTTTTTTTTTTTC-3') and arbitrary primer encoding an EcoRI site (5'-CCGTGAATTCGCTGGGAT-3'). Full-length SFTPB transcript variant 2-short was amplified from a Human Lung Marathon-Ready cDNA library (cat# 7408–1; Clontech), using the Marathon Adaptor Primer 1 (AP1), and an antisense primer designed according to the sequence of the DD product (5'-GCTAAGGCTTGTTTGG CTTTTTGTT-3'). The primary PCR product was reamplified using the same primer, but including an EcoRI site at the 5'-end (5'-CGGAATT CGCTAAGGCTTGTTTGGCTTTTTGTT-3'). The 5'-RACE product ( 1.8 kb) was cloned into NotI/EcoRI-digested pBluescript II KS (+/–) vector (Stratagene, Cedar Creek, Texas).

Northern Hybridization
RNA was extracted from frozen lung parenchyma using Trizol reagent. Three to five micrograms of total RNA were resolved by 2.2 M formaldehyle/1% agarose gel electrophoresis, and transferred to nylon membrane. SFTPB transcript variant 2-short 3'-UTR oligonucleotide probe (5'-TCCTCATGACCTAACCTCATCCCAGT-3') was labeled with -³²P dATP using terminal deoxynucleotidyl transferase (Promega, Fisher Scientific Ltd., Nepean, ON, Canada), and allowed to hybridize at a concentration of 0.1 pmol/ml of hybridization solution (50 mM NaPO₄, pH 7.2, 0.65 M NaCl, 7% SDS, 1% BSA) containing 10 µg/ml poly (A)+ RNA as blocker, at 60°C for 7 h. The probed blot was washed repeatedly in 2x SSC, 0.5% SDS at room temperature, with a final wash in 0.5x SSC, 0.1% SDS at 44°C for 2 min, and exposed to autoradiographic film. The SFTPB coding-region probe (spans nucleotides 1 through 644, GenBank Accession No. DQ317589) was labeled by random priming in the presence of -³²P dATP, and hybridization was at 62°C for 19 h in hybridization solution (as above) containing 0.1 mg /ml salmon sperm DNA as blocker. The probed blot was washed repeatedly in 2x SSC, 1% SDS at room temperature, with a final wash in 0.2x SSC, 0.5% SDS at 55°C for 30 min, and exposed to autoradiographic film.

Tissue Dot Blot Hybridization
Human RNA Master Blot (cat# 7770–1; Clontech) containing poly (A)⁺ RNA from 50 different tissues, was probed with SFTPB transcript variant 2-short 3'-UTR oligonucleotide probe as described above, at 58°C for 7 h. The probed blot was washed repeatedly in 2x SSC, 1% SDS at room temperature, with a final wash in the same solution at 41°C for 2 min, and exposed to autoradiographic film.

	RESULTS AND DISCUSSION

Top Abstract MATERIALS AND METHODS RESULTS AND DISCUSSION References

 
Enumeration of Sequence Tags Expressed in Bronchial Epithelium by SAGE
This study describes large-scale gene expression profiling of smoke-damaged bronchial epithelium and lung parenchyma, through generation and analysis of 21 SAGE libraries, sampling nearly 2 million sequence tags (Table 1). Even with the precautionary exclusion of singleton sequence tags (some of which potentially contain sequencing errors), > 80,000 unique tags were identified from the 19 bronchial epithelial (BE) libraries collectively, and > 10,000 unique tags from the 2 lung parenchyma (LP) libraries (pooled from four individuals each). However, only 70% of the unique tags from the BE dataset (55,869/80,183) mapped to a UniGene cluster according to SAGE Genie tag-to-gene mapping. Remarkably, the fact that 24,314 unique tags do not match a UniGene cluster suggests that many genes expressed in bronchial epithelium are not represented in the current databases of expressed sequence tags (ESTs). This interpretation is consistent with findings that the majority of unmatched SAGE tags represent novel transcript variants and/or novel genes (20). Moreover, as multiple SAGE tags frequently map to the same gene, the number of unique tags with UniGene mappings may not necessarily reflect the number of genes expressed. Accordingly, the 55,869 mapped unique SAGE tags converged to 22,822 unique UniGene clusters, presumably reflecting an abundance of transcript variants (alternative splicing and/or alternate poly (A⁺) adenylation site usage), and antisense transcripts (Figure 1). Our tag-to-gene ratio of 2.45:1 is close to that calculated for the entire publicly available SAGE database (2.25:1), which at the time of analysis consisted of 101 human libraries (20).

View larger version (17K): [in this window] [in a new window]   Figure 1. Number of expressed genes detected within the bronchial epithelium by SAGE. Singleton tags are defined as sequence tags having a raw tag count of one within the entire bronchial epithelial SAGE dataset. Tag-to-gene mapping was per SAGE Genie, October 2004. Nonannotated refers to no associated Gene Symbol assigned to the mapping.

Relatedness of Epithelium and Parenchyma Expression Profiles
Cluster analysis indicate that bronchial epithelium and lung parenchyma are distinguishable based upon gene expression profiles. The 19 BE libraries cluster as one clade distinct from both lung parenchyma libraries and select nonlung libraries in GEO, while the two lung parenchyma libraries from this study (LP-1, LP-2) and an additional lung library (Lung_762) from SAGEmap database cluster together (Figure 2).

View larger version (16K): [in this window] [in a new window]   Figure 2. Relatedness of bronchial epithelial SAGE libraries (BE-1 through BE-16) and lung parenchymal SAGE libraries (NLP-1, NLP-2; or LP-1, LP-2 in text). All 21 SAGE libraries generated in this study, along with 14 libraries from the GEO data repository at SAGEmap, were analyzed by cluster analysis using a single-link hierarchical algorithm. In the resultant dendrogram, branch length (height) represents distance. SAGE libraries retrieved for analysis from the GEO data resposiory at SAGEmap include: 676_NT_Brain_M; 677_NT_Breast_LuminarMammaryEpithelium_BerEp4; 685_NT_Prostate_M; 695_NT_Brain_Cerebellum; 728_NT_ColonicEpithelium1; 729_NT_ColonicEpithelium2; 739_NT_Prostate_M; 760_NT_LuminarMammary-EpitheliumAntibodyPurified_F; 761_NT_Cerebel-lum_F; 763_NT_Brain_Pooled_M; 780_NT_Breast_GestationalHyperplasia_F; 781_NT_Breast_Myoepi-thelial_F; 786_NT_Brain_PediatricFrontalCortex_M; 762_Lung.

Repeatability of SAGE
To test the repeatability of the SAGE protocol, we generated duplicate libraries (BE-4A/4B) from a single tissue lysate of a bronchial brushing. According to the clustering analysis, these duplicate libraries group together (Figure 2). Similarly, linear regression scores indicate that duplicate libraries BE-4A/4B are more closely related to one another (R = 0.99) than either is to any other library in the dataset. For reference, the average R value for BE-4A versus the other bronchial libraries (excluding BE-4B, and BE-8A/8B) is 0.9; the average R value for BE-4B versus the other bronchial libraries (excluding BE-4A, and BE-8A/8B) is 0.89.

Reproducibility of Bronchial Brushings
To evaluate the reproducibility of bronchial brushings in terms of gene expression profile, two pairs of libraries (BE-8A/8B and BE-11A/11B) were constructed from brushings attained from the same individuals, taken 1 mo apart. According to cluster analysis, BE-11A and BE-11B group together (Figure 2). Similarly, linear regression data support a strong relatedness between libraries 11A and 11B (R = 0.97; compared with average R values to the other brushing libraries of 0.86 for BE-11A [excluding BE-11B and BE-8A/B] and 0.91 for BE-11B [excluding BE-11A and BE-8A/B]).

Conversely, although BE-8A and BE-8B (libraries originating from the same individual) cluster within the bronchial epithelial clad (Figure 2), linear regression data suggests that these two libraries are distantly related to the other BE libraries (average R = 0.75 and 0.74, respectively), and moreover have a relatively low similarity score to each other (R = 0.69). It is noted that the presence of red blood cells was atypically evident within the lysate used to generate library BE-8A; this is consistent with a relatively high abundance of SAGE tags specifying hemoglobin transcripts in this library. Whether or not this contributes to the disparity observed between libraries BE-8A and BE-8B is not known. Although BE-11A/11B strongly supports the reproducibility of bronchial brushings, BE-8A/8B illustrates that care must be taken at the time of sample acquisition.

Expression Profile of Bronchial Epithelial SAGE Libraries
The complete data for the 19 BE libraries has been deposited in the GEO database under GenBank accession number GSE3707. Tag-to-gene mapping classifications of the 50 most abundant SAGE tags from the average of these libraries are summarized in Figure 3A. Twenty-one of these tags map to nuclear-encoded, nonribosomal transcripts, eight of which show enriched expression in the bronchial epithelium libraries relative to other tissue-specific SAGE libraries (per SAGE Anatomic Viewer, SAGE Genie), and are described in Table 2. At least four of these bronchial-enriched proteins are associated with defense of the bronchial epithelium against susceptibility to infection, protection from cytotoxic effects of proinflammatory reactants, or modulation of inflammatory responses: MUC5B (mucin 5B), a major component of respiratory tract mucus associated with mucociliary transport and clearance (21); LPLUNC1 (long palate, lung and nasal epithelium carcinoma-associated 1), one of seven members belonging to the PLUNC family of proteins postulated to play a role in innate immune defense (22); SCGB1A1 (secretoglobin family 1A, member 1; also known as Clara cell–specific 10-kD protein; uteroglobin), which is the most abundantly expressed transcript in the bronchial epithelium libraries, associated with immunoregulatory and anti-inflammatory activities (23); SLPI (secretory leukocyte proteinase inhibitor), a protease inhibitor associated with protection against proteolytic damage during inflammatory responses; also exhibiting antimicrobial and wound-healing activities (24, 25).

View larger version (32K): [in this window] [in a new window]   Figure 3. Pie chart depicting tag-to-gene mapping classifications of the 50 most abundant, unique tags from (A) the average of the 19 bronchial epithelial SAGE libraries and (B) the average of the two lung parenchymal SAGE libraries. Data in A correspond with that presented in Table E3; data in B correspond with that presented in Table E4. Tag-to-gene mapping was per SAGE Genie, August 2005, with reference to SAGEmap. Repetitive tags map with equally high reliabilities to multiple transcripts, which presumably contribute to the cumulative tag counts.

Expression Profile of Lung Parenchyma SAGE Libraries and Novel Transcript Discovery
The complete data for the two LP libraries has been deposited in the GEO database under GenBank accession number GSE3708. Tag-to-gene mapping classifications of the 50 most abundant SAGE tags from the average of the two libraries are described in Figure 3B. Twenty-five of these tags map to nuclear-encoded, nonribosomal proteins and are described in Table 2. Surfactant-associated protein (SFTP) gene tags, including those mapping to SFTPA2, SFTPB, and SFTPC, are prominent within this dataset. Surfactant is an extracellular phospholipid–protein complex that plays an essential role in normal respiration by lowering surface tension at air–liquid interfaces in the alveoli, and also plays an important role in innate immune defense within the lung (26, 27). Notably, tags mapping to genes associated with humoral immune response are also prominent within the LP dataset.

Detailed investigation into tag-to-gene mapping has resulted in discovery of a novel transcript variant in lung parenchyma. The most abundant SAGE tag identified in the parenchyma libraries for SFTPB has an internal localization within the 3'-UTR of transcript variant 2 (tag position 3 spanning nt 1703–1716, GenBank Accession No. NM_198843) and consequently has a low tag-to-gene mapping reliability score of 54%. This, in combination with the finding that the most reliable SAGE tag mapping to transcript variant 2 (tag position 1 spanning nt 2378–2391; 92% reliability) is not prominent within the LP libraries, prompted us to further investigate possible transcript variants of SFTPB within lung parenchyma. Using differential display, we identified a transcript from lung parenchyma that terminates within the 3'-UTR of SFTPB transcript variant 2; more specifically, just downstream of the low-reliability SAGE tag described above. We refer to this transcript as SFTPB transcript variant 2-short. Significantly, a potential poly (A)⁺ addition signal can be identified just upstream of the experimentally determined 3'-terminus of transcript variant 2-short (Figure 4A). Northern hybridization of normal lung RNA, using a probe specific to the 3'-UTR of transcript variant 2, detects 2 transcripts measuring 2.6 and 1.9 kb in length. Rehybridization of the same blot to a SFTPB coding region probe suggests that the 1.9-kb species represents SFTPB transcript variant 2-short. Although the 2.6-kb species is similar in size to that predicted for full-length transcript variant 2, the absence of detectable hybridization to the SFTPB coding region–specific probe, leaves the exact identity of this species unresolved (Figure 4B). In addition, the SFTPB coding region probe also detects a second relatively abundant species within the 1.5- to 2-kb size range, which may correspond to SFTPB transcript variant 1 (GenBank Accession No. NM_000542), gene-specific tags for which are also prominent within our SAGE database. In accordance with surfactant gene expression, SFTPB transcript variant 2-short shows tissue-specific expression in lung (Figure 4C). It is noted that the 3'-terminus of cDNA clone from library NCI_CGAP_D10 (GenBank Accession No. CA439044), generated from lung tissue RNA primed with oligo (dT), matches that of SFTPB transcript variant 2-short reported here. These data demonstrate the utility of SAGE for the identification of novel transcript variants, even for well-studied genes such as the SFTPs.

View larger version (49K): [in this window] [in a new window]   Figure 4. Expression analysis of SFTPB transcript variant 2-short. (A) cDNA sequence of SFTPB transcript variant 2-short. Nucleotide positions of the 3'-UTR, the SAGE tag, the putative poly (A)+-addition signal, and the positions of probes used for hybridizations, are indicated. It is noted that the entire sequence presented here is contained within GenBank Accession No. NM_198843 (SFTPB transcript variant 2), spanning nts 18–1772, but with several nucleotide differences identified (GenBank Accession No. DQ317589). Linear representation comparing SFTPB transcript variant 2 (NM_198843) and SFTPB transcript variant 2-short (DQ317589) is shown below. SAGE tag position refers to the location of the NlaIII site relative to the 3'-terminus of the given transcript, as defined by SAGEGenie nomenclature. (B) Northern hybridization of SFTPB transcript variant 2-short in lung. Two hybridizing species are detected in normal lung parenchyma by the 3'-UTR probe (see above), measuring roughly 2.6 and 1.9 kb in length (lanes 1 and 2, solid arrows). Hybridization of the same blot to a probe specific to the coding region of SFTPB (see A above), detects two species within the 1.5- to 2-kb size-range, but without detection of the 2.6-kb species detected by the 3'-UTR oligonucleotide probe (lanes 3 and 4, open arrows). Migration positions for the 28S and 18S ribosomal RNAs are indicated by the open arrowheads on the left. (C) Tissue dot blot illustrating expression of SFTPB transcript variant 2-short specific to lung (F2) and fetal lung (G7). Oligonucleotide 3'-UTR was used as hybridization probe; thus hybridizing signals reflect expression of two species (as shown in B).

On the other hand, many of the most highly expressed genes differ when comparing bronchial epithelium with lung parenchyma. Tags enriched in bronchial epithelium relative to most major tissue types (including lung parenchyma) include those mapping to MSMB (also highly expressed in prostate), MUC5B, LPLUNC1, AGR2 (anterior gradient homolog 2, also highly expressed in stomach), TFF3 (trefoil factor 3, a mucosal peptide also highly expressed in thymus and colon), CAPS (calcyphosine), CGI-38 (compararaive gene identification-38), TUBB2 (tubulin, 2), and SLPI (Table 2). The most abundant tag in the bronchial dataset maps to SCGB1A1, and is also detected in the parenchyma libraries, albeit at 10-fold lower abundance. Conversely, tags mapping to transcripts encoding surfactant-associated proteins and NAPSA (napsin A aspartic peptidase), a protease involved in post-translational processing of the pro-SFTPB precursor (28), are enriched in lung parenchyma relative to most tissue types including bronchial epithelium. In addition, tags mapping to a number of transcripts including IGHG1 (immunoglobulin heavy constant 1), SPARC (osteonectin), RNASE1 (ribonuclease, RNase A family 1, also highly expressed in pancreas), EGR1 (early growth response 1), APOC1 (apolipoprotein C-I, also highly expressed in liver), TMSB4X (thymosin, 4, X chromosome), and FTL (ferritin, light polypeptide) are highly represented in lung parenchyma and unrelated tissue types relative to the bronchial epithelium (Table 2). These observations illustrate that, despite similarities in expression profiles between bronchial epithelium and lung parenchyma, significant differences exist reflecting regional distinctions in cellular composition and biological function. These data, taken in conjunction with the cluster and linear regression analysis data, stress the importance of using matching tissue types when analyzing expression profiles.

Identification of Bronchial-Enriched Genes
Genes whose expression is enriched in bronchial epithelium relative to other tissue types were identified by first comparing our data with normal nonlung libraries in the SAGEmap database, and second by validating tissue specificity of expression for select genes by RT-PCR experimentation. Through this approach, we have discovered the expression pattern of genes previously unknown to be expressed in bronchial epithelium. Tag-to-gene mapping classifications of the top 100 bronchial-enriched tags are summarized in Figure 5. A description of the top 30 tags with 70% or greater mapping reliabilities to defined transcripts is presented in Table 3.

View larger version (27K): [in this window] [in a new window]   Figure 5. Tag-to-gene mapping classifications of the top ranking 100 bronchial-enriched SAGE tags. Data here correspond with that presented in Table E5. Tag-to-gene mapping was per SAGE Genie, August 2005. In addition to the nonlung libraries used in the cluster analysis, the following libraries retrieved from the GEO data repository at SAGEmap were included for identification of bronchial-enriched genes: 708_NT_Kidney_F; 709_NT_Leukocyte_F; 727_NT_Skin_Primary-Mesothelioma; 738_NT_Peritoneum_Mesothelial; 785_NT_Liver_M; 1499_NT_Heart_M; 2386_NT_SpinalCord.

KCNE1 (potassium voltage-gated channel, Isk-related family, member 1) is a member of the KCNE family of accessory protein subunits, and in complex with the pore-forming channel protein KCNQ1, is involved in the regulation of potassium (K+) channel activity in the heart (30). Expression profiling reveals that KCNQ1 is expressed in many human tissues in addition to heart, highlighting the relevance of voltage-gated K+ channels for normal physiology of many tissues, including lung (31). Enriched expression of KCNE1 in BE SAGE libraries reported here suggests that KCNQ1/KCNE1 complexes play a significant role in K+ conductance within the bronchial epithelium.

ABCA13 (ATP binding cassette gene, subfamily A, member 13) is a recently identified member of the ABC transporter superfamily of proteins. Highest expression levels in human tissue is found in trachea, testis, and bone marrow (32). The data reported here suggest that ABCA13 is predominantly expressed in the bronchial epithelium, with lower levels of expression observed in testis, pancreas, and lung parenchyma.

Expression of MS4A8B (membrane-spanning 4-domains, subfamily A, member 8B) has not previously been reported in bronchus, and appears to be relatively specific to bronchus. MS4A8B is a member of the MS4A family of transmembrane proteins structurally related to and including the cell surface hematopoietic proteins CD20, the high-affinity IgE receptor chain, and HTm4 (hematopoietic cell 4 transmembrane protein). These proteins have been proposed to function as ligand-gated ion channels with signal transduction activity (33). Multiple members of the MS4A gene family (including member 8B) are clustered within an 600-kb region on chromosomal region 11q12, one of multiple genetic loci (11q12-q13) linked to asthma development (34). Considering the highly enriched expression in bronchial epithelium, and the chromosomal location, it is suggested that MS4A8B may play an important role in respiratory function.

Discovery of Genes Associated with Ciliary Function in Bronchial Epithelium
Unexpectedly, many of the novel bronchial-enriched genes identified by library comparisons were also found, according to the RT-PCR validation, to be prominently expressed in testis (Table 3; Figure 6). This reflects the absence of a testis library in the SAGEmap database at the time of our analysis; hence those genes predominantly expressed in bronchus and testis were included in our collection of bronchial-enriched tags. Coincidentally, these two tissues share a common structural feature: the axoneme, which is instrumental to flagellar-mediated sperm motility in testis and cilia-mediated mucociliary clearance in lung, thus accounting for many shared transcripts. For example, DNAI2 (axonemal dynein intermediate polypeptide 2) belongs to a family of dynein polypeptides localized to ciliary and flagellar axonemes and functions as a component of a multi-subunit motor complex in association with microtubules to facilitate ciliary/flagellar motility (35). In contrast to axonemal dyneins, expression of cytoplasmic dynein polypeptides is evident within both bronchus and lung parenchyma, consistent with functional expectation. Other examples of genes preferentially expressed in bronchial epithelium and testis with known roles in flagellar/ciliary activity include: SPAG6 (sperm-associated antigen 6), encoding an axonemal component of sperm flagella (36, 37); ASP (AKAP-associated sperm protein), encoding a protein which binds to the A-kinase anchoring protein 110 from sperm flagella (38) and FOXJ1 (forkhead transcription factor J1), required for developmental stages of ciliogenesis (39). These findings concur with the fact that over 200 potential ciliary axonemal proteins were detected in human bronchial epithelial cells using a proteomic approach (40). Furthermore, the abundance of adenylate kinase 7 gene-specific tags in the bronchial epithelium libraries is also consistent with ciliary function, as adenylate kinase activity has been associated with axonemes in protozoa and green algae (41–43).

View larger version (56K): [in this window] [in a new window]   Figure 6. RT-PCR verification of bronchial-enriched expression. Select genes presented in Table 3 were evaluated experimentally for bronchial-enriched expression. Gene-specific PCR products generated from cDNA representing 17 tissue types (15 nonlung) are presented above. Amplicon length was typically 100–200 bp. RT-PCR from -actin (ACTB)–specific primers was used as a loading control. Minus-RT controls using bronchial epithelial cDNA as template were negative for PCR product (not shown). These data are summarized in Table 3. To verify differential expression between bronchial epithelium and lung parenchyma, three genes were selected for real-time quantitative RT-PCR analysis (indicated by asterisks). Differential expression for all three genes was confirmed at a P value of < 0.001 by Mann-Whitney U Test (Table E6). Brc, bronchus; H, heart; Bn, brain; Pl, placenta; Lg, lung; Lv, liver; M, muscle; K, kidney; Pc, pancreas; Sp, spleen; Ty, thymus; Pr, prostate; T, testies; Ov, ovary; Int, intestine; C, colon; Lk, leukocyte.

A significant proportion of tags enriched in bronchial epithelium map to undefined transcripts, including chromosomal open reading frames and hypothetical proteins (Figure 5). We have further investigated the expression of six such transcripts (Table 3). Chromosomal open reading frames C9orf117 and C6orf118, and hypothetical proteins DKFZp434I099 and FLJ32884 were all found to be preferentially expressed in bronchus and testis, while expression of hypothetical protein MGC48998 appeared to be specific to bronchus, and that of hypothetical protein FLJ40919 was found to be highly enriched in bronchus, with minimal expression detected in heart. Sequence similarity search results support a role in ciliogenesis for C9orf117 and FLJ32884.

Interestingly, tags specifying proteins assigned either an established (e.g., DNAI2) or a potential (e.g., ARMC3) role in ciliogenesis are frequently detected at notable levels in ependymoma SAGE libraries in SAGEmap. And since ependymoma constitutes a cancer originating within a ciliated region of the brain, ciliary proteins could potentially serve as markers to detect clonal expansion originating from this cell type.

Correlation of Gene Expression in the Bronchial Epithelium with Smoking Status
We determined genes differentially expressed between current and former smokers in our bronchial epithelium SAGE dataset, which was comprised of 5 current and 11 former smokers. Three hundred forty-nine tags showed at least a 3-fold difference, of which 149 tags were higher in the current smoker category (Table E7) and 200 tags were higher in the former smoker category (Table E8). Despite the small sample size in this comparison, many of the reported smoke-induced gene expression changes were captured in our analysis (12, 46).

Classical phase I and phase II xenobiotic metabolizing enzymes known to be induced by smoking such as subfamilies A and B of cytochrome P450, family 1 (CYP1A1, CYP1B1), and glutathione S-transferase A2 (GSTA2), as well as antioxidants including glutathione peroxidase 2 (GPX2), thioredoxin (TXN), and sulfiredoxin 1 homolog (SRXN1) (47), were among those showing the highest differential expression in our current-smoker dataset. In addition, tags mapping to transcripts encoding oxidoreductases (associated with redox balance) including various members of the aldo-keto reductase family of proteins (AKR1B10, AKR1C2, and AKR1C3), carbonyl reductase 1 (CBR1), alcohol dehydrogenase 7 (ADH7), aldehyde dehydrogenase 3 family, member A1 (ALDH3A1), and NAD(P)H dehydrogenase, quinone 1 (NQO1), were also detected at higher levels in the current smoker SAGE dataset relative to the former smoker dataset. Carbonyl reductase 1 activity mediates inactivation of tobacco-derived carcinogens (48); expression of NQO1 has been shown in a previous study to be induced by acrolein, a component of cigarette smoke (49).

CONCLUSIONS
In this study, we have deduced the transcriptome of smoke-damaged bronchial epithelium by analyzing 1,866,725 sequence tags from 19 SAGE libraries, representing one of the largest human SAGE studies reported to date. We have detected the expression of at least 22,822 genes in the bronchial epithelium and identified 24,314 sequence tags without matches to known UniGene Clusters, cautioning our current understanding of the transcriptome.

Our analysis emphasizes the distinctiveness of the bronchial epithelium from the lung parenchyma at the gene expression level (Table 2, Figure 6). Abundantly expressed genes from the bronchial epithelium dataset are frequently associated with innate defense and protection of the central airways, while those from the parenchyma dataset are frequently associated with respiration and humoral immune response.

In addition, we have identified genes preferentially expressed in bronchial epithelium, some of which were previously unknown to be expressed in lung. Many of these genes are also prominently expressed in testis, where they are associated with flagella-mediated sperm motility, and likely play a role in mucociliary clearance in the lung. It is noted that the majority of tags most highly enriched in bronchial epithelium (63%) map to undefined transcripts, including chromosomal open reading frames and cDNAs. Further investigation of these transcripts will potentially identify additional genes associated with ciliogenesis, and other bronchial-specialized functions. Furthermore, correlation of bronchial epithelium SAGE profiles to smoking status identified a list of 349 differentially expressed gene tags. The detection of genes known to be deregulated by tobacco smoke in this gene list suggests the potential biological relevance of the genes previously unassociated with smoking.

The expression data of smoke-damaged bronchial epithelium generated in this study is available as a public resource serving as a baseline for the benefit of future expression studies pertaining to the bronchial epithelium and lung function. Improvements in tag-to-gene mapping strategies, in conjunction with this comprehensive dataset, will continue to further our understanding of the bronchial epithelial transcriptome and molecular biology of the upper respiratory tract, potentially bringing us closer to the ultimate goal of enhanced understanding and improved management of lung pathologies, most notably those associated with dysfunctional cilia.

	Acknowledgments

 
The authors thank Jin-Hee Kim, Shaminder Sandhu, Sandra Henderson, Andrea Pusic, Sukhinder Atkar-Khattra, George Yang, and Jeff Stott for their expert assistance.

	Footnotes

 
This work was supported by funds from Genome Canada/Genome British Columbia and the National Cancer Institute of Canada.

The following data have been deposited at GEO: bronchial epithelial and lung parenchyma series of SAGE libraries (GSE3754), profile of bronchial epithelial (GSE3707), profile of lung parenchyma (GSE3708), sequence of SFTPB transcript variant 2-short (DQ317589).

This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1165/rcmb.2006-0056OC on June 29, 2006

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.

Received in original form February 5, 2006

Accepted in final form May 27, 2006

	References

Top Abstract MATERIALS AND METHODS RESULTS AND DISCUSSION References

 

1. Boucher RC. New concepts of the pathogenesis of cystic fibrosis lung disease. Eur Respir J 2004;23:146–158.[Abstract/Free Full Text]
2. Chodhari R, Mitchison HM, Meeks M. Cilia, primary ciliary dyskinesia and molecular genetics. Paediatr Respir Rev 2004;5:69–76.[CrossRef][Medline]
3. Mall M, Grubb BR, Harkema JR, O'Neal WK, Boucher RC. Increased airway epithelial Na+ absorption produces cystic fibrosis-like lung disease in mice. Nat Med 2004;10:487–493.[CrossRef][Medline]
4. Fujii T, Dracheva T, Player A, Chacko S, Clifford R, Strausberg RL, Buetow K, Azumi N, Travis WD, Jen J. A preliminary transcriptome map of non-small cell lung cancer. Cancer Res 2002;62:3340–3346.[Abstract/Free Full Text]
5. Hibi K, Liu Q, Beaudry GA, Madden SL, Westra WH, Wehage SL, Yang SC, Heitmiller RF, Bertelsen AH, Sidransky D, et al. Serial analysis of gene expression in non-small cell lung cancer. Cancer Res 1998;58:5690–5694.[Abstract/Free Full Text]
6. Nacht M, Dracheva T, Gao Y, Fujii T, Chen Y, Player A, Akmaev V, Cook B, Dufault M, Zhang M, et al. Molecular characteristics of non-small cell lung cancer. Proc Natl Acad Sci USA 2001;98:15203–15208.[Abstract/Free Full Text]
7. Ning W, Li CJ, Kaminski N, Feghali-Bostwick CA, Alber SM, Di YP, Otterbein SL, Song R, Hayashi S, Zhou Z, et al. Comprehensive gene expression profiles reveal pathways related to the pathogenesis of chronic obstructive pulmonary disease. Proc Natl Acad Sci USA 2004; 101:14895–14900.[Abstract/Free Full Text]
8. Powell CA, Spira A, Derti A, DeLisi C, Liu G, Borczuk A, Busch S, Sahasrabudhe S, Chen Y, Sugarbaker D, et al. Gene expression in lung adenocarcinomas of smokers and nonsmokers. Am J Respir Cell Mol Biol 2003;29:157–162.[Abstract/Free Full Text]
9. Spira A, Beane J, Pinto-Plata V, Kadar A, Liu G, Shah V, Celli B, Brody JS. Gene expression profiling of human lung tissue from smokers with severe emphysema. Am J Respir Cell Mol Biol 2004;31:601–610.[Abstract/Free Full Text]
10. Vos JB, van Sterkenburg MA, Rabe KF, Schalkwijk J, Hiemstra PS, Datson NA. Transcriptional response of bronchial epithelial cells to Pseudomonas aeruginosa: identification of early mediators of host defense. Physiol Genomics 2005;21:324–336.[Abstract/Free Full Text]
11. Vos JB, Datson NA, van Kampen AH, Luyf AC, Verhoosel RM, Zeeuwen PL, Olthuis D, Rabe KF, Schalkwijk J, Hiemstra PS. A molecular signature of epithelial host defense: comparative gene expression analysis of cultured bronchial epithelial cells and keratinocytes. BMC Genomics 2006;7:9.[CrossRef][Medline]
12. Spira A, Beane J, Shah V, Liu G, Schembri F, Yang X, Palma J, Brody JS. Effects of cigarette smoke on the human airway epithelial cell transcriptome. Proc Natl Acad Sci USA 2004;101:10143–10148.[Abstract/Free Full Text]
13. Velculescu VE, Zhang L, Vogelstein B, Kinzler KW. Serial analysis of gene expression. Science 1995;270:484–487.[Abstract/Free Full Text]
14. Lam S, Kennedy T, Unger M, Miller YE, Gelmont D, Rusch V, Gipe B, Howard D, LeRiche JC, Coldman A, et al. Localization of bronchial intraepithelial neoplastic lesions by fluorescence bronchoscopy. Chest 1998;113:696–702.[Abstract/Free Full Text]
15. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate phenol-chloroform extraction. Anal Biochem 1987;162:156–159.[Medline]
16. Boon K, Osorio EC, Greenhut SF, Schaefer CF, Shoemaker J, Polyak K, Morin PJ, Buetow KH, Strausberg RL, De Souza SJ, et al. An anatomy of normal and malignant gene expression. Proc Natl Acad Sci USA 2002;99:11287–11292.[Abstract/Free Full Text]
17. Barrett T, Suzek TO, Troup DB, Wilhite SE, Ngau WC, Ledoux P, Rudnev D, Lash AE, Fujibuchi W, Edgar R. NCBI GEO: mining millions of expression profiles–database and tools. Nucleic Acids Res 2005;33:D562–D566.[Abstract/Free Full Text]
18. Ihaka R, Gentleman RR. A language for data analysis and graphics. J Comput Graph Statist 1996;5:299–314.[CrossRef]
19. Liang P, Pardee AB. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 1992;257:967–971.[Abstract/Free Full Text]
20. Chen J, Sun M, Lee S, Zhou G, Rowley JD, Wang SM. Identifying novel transcripts and novel genes in the human genome by using novel SAGE tags. Proc Natl Acad Sci USA 2002;99:12257–12262.[Abstract/Free Full Text]
21. Thornton DJ, Sheehan JK. From mucins to mucus: toward a more coherent understanding of this essential barrier. Proc Am Thorac Soc 2004;1:54–61.[Abstract/Free Full Text]
22. Bingle CD, Craven CJ. PLUNC: a novel family of candidate host defence proteins expressed in the upper airways and nasopharynx. Hum Mol Genet 2002;11:937–943.[Abstract/Free Full Text]
23. Singh G, Katyal SL. Clara cell proteins. Ann N Y Acad Sci 2000;923:43–58.[Abstract/Free Full Text]
24. Ashcroft GS, Lei K, Jin W, Longenecker G, Kulkarni AB, Greenwell-Wild T, Hale-Donze H, McGrady G, Song XY, Wahl SM. Secretory leukocyte protease inhibitor mediates non-redundant functions necessary for normal wound healing. Nat Med 2000;6:1147–1153.[CrossRef][Medline]
25. Sallenave JM. Antimicrobial activity of antiproteinases. Biochem Soc Trans 2002;30:111–115.[CrossRef][Medline]
26. Johansson J, Curstedt T. Molecular structures and interactions of pulmonary surfactant components. Eur J Biochem 1997;244:675–693.[Medline]
27. Wright JR. Immunoregulatory functions of surfactant proteins. Nat Rev Immunol 2005;5:58–68.[CrossRef][Medline]
28. Brasch F, Ochs M, Kahne T, Guttentag S, Schauer-Vukasinovic V, Derrick M, Johnen G, Kapp N, Muller KM, Richter J, et al. Involvement of napsin A in the C- and N-terminal processing of surfactant protein B in type-II pneumocytes of the human lung. J Biol Chem 2003;278:49006–49014.[Abstract/Free Full Text]
29. Peri A, Cordella-Miele E, Miele L, Mukherjee AB. Tissue-specific expression of the gene coding for human Clara cell 10-kD protein, a phospholipase A2-inhibitory protein. J Clin Invest 1993;92:2099–2109.[Medline]
30. Melman YF, Um SY, Krumerman A, Kagan A, McDonald TV. KCNE1 binds to the KCNQ1 pore to regulate potassium channel activity. Neuron 2004;42:927–937.[CrossRef][Medline]
31. Yang WP, Levesque PC, Little WA, Conder ML, Shalaby FY, Blanar MA. KvLQT1, a voltage-gated potassium channel responsible for human cardiac arrhythmias. Proc Natl Acad Sci USA 1997;94:4017–4021.[Abstract/Free Full Text]
32. Prades C, Arnould I, Annilo T, Shulenin S, Chen ZQ, Orosco L, Triunfol M, Devaud C, Maintoux-Larois C, Lafargue C, et al. The human ATP binding cassette gene ABCA13, located on chromosome 7p12.3, encodes a 5058 amino acid protein with an extracellular domain encoded in part by a 4.8-kb conserved exon. Cytogenet Genome Res 2002;98:160–168.[CrossRef][Medline]
33. Liang Y, Tedder TF. Identification of a CD20-, FcepsilonRIbeta-, and HTm4-related gene family: sixteen new MS4A family members expressed in human and mouse. Genomics 2001;72:119–127.[CrossRef][Medline]
34. Adra CN, Mao XQ, Kawada H, Gao PS, Korzycka B, Donate JL, Shaldon SR, Coull P, Dubowitz M, Enomoto T, et al. Chromosome 11q13 and atopic asthma. Clin Genet 1999;55:431–437.[CrossRef][Medline]
35. Inaba K. Molecular architecture of the sperm flagella: molecules for motility and signaling. Zoolog Sci 2003;20:1043–1056.[CrossRef][Medline]
36. Neilson LI, Schneider PA, Van Deerlin PG, Kiriakidou M, Driscoll DA, Pellegrini MC, Millinder S, Yamamoto KK, French CK, Strauss JF III. cDNA cloning and characterization of a human sperm antigen (SPAG6) with homology to the product of the Chlamydomonas PF16 locus. Genomics 1999;60:272–280.[CrossRef][Medline]
37. Sapiro R, Kostetskii I, Olds-Clarke P, Gerton GL, Radice GL, Strauss IJ. Male infertility, impaired sperm motility, and hydrocephalus in mice deficient in sperm-associated antigen 6. Mol Cell Biol 2002;22: 6298–6305.[Abstract/Free Full Text]
38. Carr DW, Fujita A, Stentz CL, Liberty GA, Olson GE, Narumiya S. Identification of sperm-specific proteins that interact with A-kinase anchoring proteins in a manner similar to the type II regulatory subunit of PKA. J Biol Chem 2001;276:17332–17338.[Abstract/Free Full Text]
39. You Y, Huang T, Richer EJ, Schmidt JE, Zabner J, Borok Z, Brody SL. Role of f-box factor foxj1 in differentiation of ciliated airway epithelial cells. Am J Physiol Lung Cell Mol Physiol 2004;286:L650–L657.[Abstract/Free Full Text]
40. Ostrowski LE, Blackburn K, Radde KM, Moyer MB, Schlatzer DM, Moseley A, Boucher RC. A proteomic analysis of human cilia: identification of novel components. Mol Cell Proteomics 2002;1:451–465.[Abstract/Free Full Text]
41. Nakamura K, Iitsuka K, Fujii T. Adenylate kinase is tightly bound to axonemes of Tetrahymena cilia. Comp Biochem Physiol B Biochem Mol Biol 1999;124:195–199.[CrossRef][Medline]
42. Pullen TJ, Ginger ML, Gaskell SJ, Gull K. Protein targeting of an unusual, evolutionarily conserved adenylate kinase to a eukaryotic flagellum. Mol Biol Cell 2004;15:3257–3265.[Abstract/Free Full Text]
43. Zhang H, Mitchell DR. Cpc1, a Chlamydomonas central pair protein with an adenylate kinase domain. J Cell Sci 2004;117:4179–4188.[Abstract/Free Full Text]
44. Yang P, Sale WS. The Mr 140,000 intermediate chain of Chlamydomonas flagellar inner arm dynein is a WD-repeat protein implicated in dynein arm anchoring. Mol Biol Cell 1998;9:3335–3349.[Abstract/Free Full Text]
45. Zhang Z, Futamura M, Vikis HG, Wang M, Li J, Wang Y, Guan KL, You M. Positional cloning of the major quantitative trait locus underlying lung tumor susceptibility in mice. Proc Natl Acad Sci USA 2003;100:12642–12647.[Abstract/Free Full Text]
46. Hackett NR, Heguy A, Harvey BG, O'Connor TP, Luettich K, Flieder DB, Kaplan R, Crystal RG. Variability of antioxidant-related gene expression in the airway epithelium of cigarette smokers. Am J Respir Cell Mol Biol 2003;29:331–343.[Abstract/Free Full Text]
47. Jonsson TJ, Murray MS, Johnson LC, Poole LB, Lowther WT. Structural basis for the retroreduction of inactivated peroxiredoxins by human sulfiredoxin. Biochemistry 2005;44:8634–8642.[CrossRef][Medline]
48. Finckh C, Atalla A, Nagel G, Stinner B, Maser E. Expression and NNK reducing activities of carbonyl reductase and 11beta-hydroxysteroid dehydrogenase type 1 in human lung. Chem Biol Interact 2001;130–132: 761–773.
49. Tirumalai R, Rajesh Kumar T, Mai KH, Biswal S. Acrolein causes transcriptional induction of phase II genes by activation of Nrf2 in human lung type II epithelial (A549) cells. Toxicol Lett 2002;132:27–36.[CrossRef][Medline]

This article has been cited by other articles:

				B. Chhin, J. T. Pham, L. El Zein, K. Kaiser, O. Merrot, and P. Bouvagnet Identification of transcripts overexpressed during airway epithelium differentiation Eur. Respir. J., July 1, 2008; 32(1): 121 - 128. [Abstract] [Full Text] [PDF]

				F. A. Barnes, L. Bingle, and C. D. Bingle Pulmonary Genomics, Proteomics, and PLUNCs Am. J. Respir. Cell Mol. Biol., April 1, 2008; 38(4): 377 - 379. [Full Text] [PDF]

				A. J. Ross, L. A. Dailey, L. E. Brighton, and R. B. Devlin Transcriptional Profiling of Mucociliary Differentiation in Human Airway Epithelial Cells Am. J. Respir. Cell Mol. Biol., August 1, 2007; 37(2): 169 - 185. [Abstract] [Full Text] [PDF]

				A. Heguy, B.-G. Harvey, P. L. Leopold, I. Dolgalev, T. Raman, and R. G. Crystal Responses of the human airway epithelium transcriptome to in vivo injury Physiol Genomics, April 24, 2007; 29(2): 139 - 148. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Online Supplement All Versions of this Article: 2006-0056OCv1 35/6/651    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lonergan, K. M. Articles by Lam, W. L. Search for Related Content PubMed PubMed Citation