UvA-DARE (Digital Academic Repository) Biomarker discovery for asthma phenotyping: From gene expression to the clinic Wagener, A.H. Publication date 2016 Document Version Final published version Link to publication Citation for published version (APA): Wagener, A. H. (2016). Biomarker discovery for asthma phenotyping: From gene expression to the clinic. General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:25 Sep 2021 CHAPTER 3 Supporting Information File The impact of allergic rhinitis and asthma on human nasal and bronchial epithelial gene expression methods Primary epithelial cell culture Primary cells were obtained by first digesting the biopsies and brushings with collage- nase 4 (Worthington Biochemical Corp., Lakewood, NJ, USA) for 1 hour in Hanks’ bal- anced salt solution (Sigma-Aldrich, Zwijndrecht, The Netherlands). Subsequently cells were washed with Hanks’ balanced salt solution (HBSS) and resuspended in bronchial epithelial growth medium (BEGM) (Invitrogen, Breda, The Netherlands) and seeded in one well of a 6 wells plate. Cells were grown in fully humidified air containing 5% CO2 at 37°C, and culture medium was replaced every other day. Cells were cultured to 80% confluence and were pre-incubated with bronchial epithelial basal medium (BEBM) for 48 hours prior to the removement of supernatant and RNA extraction. For bronchial epithelial cells it took 14 days on average, and for nasal epithelial cells it took 24 days on average to grow to 80% confluence. There was no difference in time of culture between the three subject groups. rnA extraction Total RNA from each sample was extracted using Trizol (Life Technologies Inc., Gaiters- burg, MD, USA) using manufacturer’s protocol, followed by purification by nucleospin RNA II (Machery-Nagel, Düren, Germany). RNA concentration of all samples was measured on the nanodrop ND-1000 (NanoDrop Technologies Inc., Wilmington, DE, USA). The quality of the RNA was checked by using Agilent 2100 bio-analyser (Agilent Technologies, Palo Alto, CA, USA). All RIN scores were ≥9.5. microarray Affymetrix U133+ mP Human Genome U133+ PM Genechip Array (Affymetrix inc., Santa Clara, CA, USA) representing more than 47,000 transcripts and variants, including over 33,000 well- characterized genes, was used in the analysis of the genes. The MicroArray Department (MAD) of the University of Amsterdam, a fully licensed microarray technologies centre for Affymetrix Genechip® platforms, performed the technical handling and the quality control of the microarray experiments. The quality of the images was checked by visual inspection and all raw data passed quality criteria based on borderplots, pseudocolor slide images, RNA degradation plots, box and density plots, RI plots (against a pseudore- ference), correlation and PCA plots. nlP network discovery Network analysis was performed on the same set of genes using NLP Network Discovery (GeneSpring GX12, Agilent Technologies, Amstelveen, The Netherlands) that derives its relations from PubMed. A direct interaction network was built that captures relations Gene Expression of Airway Epithelium 55 based on regulation, connecting the genes entered into the programme. In more detail, the majority of relations in the GeneSpring Interaction database are derived using a Natural Language Processing (NLP) algorithm that runs on published Medline abstracts. NLP is based on a ‘‘deep parsing” method and is driven by an elaborate sentence gram- mar that maximizes accuracy and has control over different aspects of a sentence with- out compromising recall. The NLP system operates on a sentence-by-sentence manner and extracts only those relations that are completely within a sentence. There are four main phases: 1) Entity recognition by consulting entity dictionaries, taking into account that there are variations in how terms appear in literature. 2) Using a set of rules, the syntactic tree structure of the sentence is derived using context free grammar rules for English, breaking up the sentence into its underlying linguistic constituents and capturing the functional roles of different parts of the sentence. 3) A semantic analysis mapping all words of interest to semantic concepts and iden- tifying which entity regulates another entity using the sentence structure imposed by the syntax tree. The relationships captured by the semantic tree are only direct relationships (finds relations that connect the selected entities by the previous NLP system). 4) Semantic interference: the semantic tree captures specific concepts from a sentence after which GeneSpring has to make inferences across these semantic concepts, using agents in one relationship to fill the missing holes in other relationships. GeneSpring extracts relations by searching the resulting semantic network for rela- tion nodes that contain all the required arguments. A signature is created for each relation depending upon the participants, their roles, and their mechanisms, and references to the relation are added. The relationships captured by the semantic tree are only direct relationships, which find relations that connect the selected entities by the NLP system that was previously explained. The relation represents molecular interactions between the entities and is characterized by a set of participating entities. We used Regulation as relation, which is the most basic relation type. In GeneSpring an entity A ‘‘regulates” another entity B, if A has some influ- ence on B. The participant entities in the Regulation relation are ‘‘regulator”, ‘‘target”, and ‘‘modulator”. Each participant either has a positive, negative, or unknown effect. Depending on the source of the relation information, each relation is assigned a Rela- tion score. This property indicates a confidence matrix on the quality of relations in the Interaction Database in GeneSpring. NLP-derived relations are graded on a scale of 1-9, the best being 9 and the weakest being 1. The score properties are internally calculated 56 Chapter 3 based on the number of references and the syntax of the sentences. We used 9 as Rela- tion score for our network discovery. real-time polymerase chain reaction and analysis Quantitative real-time PCR was used to validate the differential expression of selected genes. We chose a set of genes that represent a complete range of fold change values, capturing genes that were either higher expressed in the upper or lower airways. PCR was performed on Bio-Rad CFX96 real-time PCR detection system (Bio-Rad, Veenendaal, The Netherlands). SYBR® Green primer sequences for IL13-Rα2, EREG, PDE4D, IL1-β, IP-10, TIMP2, IL8, β –actin and GAPDH were obtained from Sigma-Aldrich (Sigma-Aldrich, Zwi- jndrecht, The Netherlands). The following primers were used: IL13-Rα2; sense: TGC-TCA- GAT-GAC-GGA-ATT-TGG, antisense: TGG-TAG-CCA-GAA-ACG-TAG-CAA-AG, EREG; sense: ATC-CTG-GCA-TGT-GCT-AGG-GT, antisense: GTG-CTC-CAG-AGG-TCA-GCC-AT, PDE4D; sense: GGC-CTC-CAA-CAA-GTT-TAA-AA, antisense: ACC-AGA-CAA-CTC-TGC-TAT-TCT, IL1-β; sense: GGA-TAT-GGA-GCA-ACA-AGT-GG, antisense: ATG-TAC-CAG-TTG-GGG-AAC- TG, IP-10; sense: TGA-AAT-TAT-TCC-TGC-AAG-CCA-AT, antisense: CAG-ACA-TCT-CTT- CTC-ACC-CTT-CTT-T, TIMP2; sense: ATA-AGC-AGG-CCT-CCA-ACG-C, antisense: GAG- CTG-GAC-CAG-TCG-AAA-CC, IL8; sense: CCA-CAC-TGC-GCC-AAC-ACA-GAA-ATT-ATT-G, antisense: GCC-CTC-TTC-AAA-AAC-TTC-TCC-ACA-ACC-C, β –actin; sense: TGA-GCG-CGG- CTA-CAG-CTT, antisense: TCC-TTA-ATG-TCA-CGC-ACG-ATT-T, GAPDH; sense: GAA-GGT- GAA-GGT-CGG-AGT-C, antisense: GAA-GAT-GGT-GAT-GGG-ATT-TC. For ATF3 and DUSP1 we used TaqMan® gene expression assays from Applied Biosystems (Nieuwerkerk a/d IJssel, The Netherlands) with the following assay IDs: ATF3; HS00231069_M1, DUSP1; HS006102757_G1. Correlations between fold changes (FC) within the microarray data and the real-time PCR data were determined using Pearson’s correlation. resUlts validation of microarray data The results of this microarray experiment were validated by independent real time PCR on the same starting material used for the microarray analysis. We first determined the expression of the housekeeping genes (ACTB and GAPDH) which was not affected by type of tissue (upper or lower airway) or condition (healthy, asthma or rhinitis). A random selection of 9 genes was used that showed a significant different expression between upper and lower airways in at least one of the subject groups. Table S1 shows the ratios calculated from the microarray data and the real-time PCR-derived expression. Statistical analysis revealed a high level of correspondence (R=0.92, P<0.0001), pointing Gene Expression of Airway Epithelium 57 towards a correlation of 85% (R2=0.85) between the microarray data and the real-time PCR (Figure S1). differential gene expression in nasal and bronchial epithelium There were substantial differences in gene expression between the epithelia from up- per and lower airways of healthy individuals. These differences were smaller in patients with allergic rhinitis and even smaller in those with concomitant allergic asthma. Using a cut-off of adjusted p < 0.05 we identified 2705 out of 41976 probe sets that were statistically differentially expressed between healthy nasal and healthy bronchial epi- thelium.