Host Genomics Plasticity in Response to Ambient Temperature Change: Transcriptional Regulation Induced by Cold Temperature Perception in the Human BEAS-2B Cell Line

Current Pharmacogenomics and Personalized Medicine, 2012, 10, 33-42 33

Original Research Article Host Genomics Plasticity in Response to Ambient Temperature Change: Transcriptional Regulation Induced by Cold Temperature Perception in the Human BEAS-2B Cell Line

Seyeon Park1,*, Sohyun Chun1,2 and Danuh Kim1,2

1Department of Applied Chemistry, Dongduk Women’s University, Seoul, Korea; 2Department of Pharmacology, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Abstract: Pharmacogenomics has long attempted to identify genome-drug interactions and environmental exposures as guideposts for personalized medicine. However, non-drug related environmental factors, too, can interact with the host genome and potentially cause confounding in explaining drug-genome interactions. One such environmental factor that has been bracketed out in the past is the ambient temperature change, and ways in which it can influence genomic plasticity. Indeed, recognition of temperature is an important element of microsensory perception that allows cells to evaluate both their external environment and internal physiological milieu. In this paper, we report the changes in global gene expression three hours after a 30-min cold temperature (10°C) treatment in the human bronchial epithelial cell line BEAS-2B using DNA microarrays. We found 11,276 candidate genes (6,297 with increased, 4,979 with decreased expression) that were differentially expressed after low-temperature treatment of BEAS-2B compared to the untreated control cells (p<0.001). Additionally, up- and down-regulated transcription factor genes were further verified using real- time polymerase chain reaction. We found expression changes in response to cold temperature in transcription factor genes such as ZXDA, ZNF44, ZDHHC13, ZNF423, ZFYVE20, ZNF45, ZC3HAV, ZCCHC5, RUNX1T1, DMRT1, STAT4, EFCAB1 and HSFX1 that can alter the temperature-adaptive responsiveness of the bronchial epithelial cells. In summary, we herein show that a moderately cold temperature induces a genome-wide response in the human bronchial epithelial cell line BEAS-2B, and further discuss its relevance for pharmacogenomics and upstream drug discovery. For example, these observations provide a new crucial putative link between ambient temperature and genomic plasticity that together inform personalized medicine such that future pharmacogenomics biomarker discovery research can better control and account for ever-present dynamic environmental exposures such as ambient temperature. We conclude with the implications of these data in relation to rational drug design for neuropathic and other chronic pain syndromes that are in part moderated by host-ambient temperature interactions. Keywords: Ambient temperature, BEAS-2B cell line, chronic pain syndrome, environmental confounding and biomarker discovery, genome-temperature interaction, genomics plasticity, personalized medicine, transcription factor gene.

1. INTRODUCTION specific genotypes [1]. However, it is unknown how these patterns of environment-induced expression plasticity differ Personalized medicine and pharmacogenomics more between genetically divergent individuals of a biological specifically, have attempted to identify the genome-drug species, especially such as the higher vertebrates. Further- interactions. However, non-drug related environmental more, in contrast to the exposures to hot temperatures, factors, too, can interact with the host genome and potentially populations under natural conditions can often be exposed to cause confounding in explaining drug-genome interactions. longer periods of less extreme or moderate cold temperature One such environmental factor that has been bracketed out in changes. Phenotypic plasticity to temperature plays an past pharmacogenomics research is the ambient temperature important role in the evolution of life in a variable climate change and ways in which it can influence genomic [5] and is widespread among species [6-8]. plasticity and gene regulation. The mRNA levels respond rapidly to variable ambient conditions such as temperature Indeed, the ‘‘genotype-by-environment’’ interaction is change [1]. For instance, this has been shown for yeast [2], the prerequisite for adaptive evolution in a fluctuating bacteria [3], and C. elegans [4] after exposure to heat shock. environment [9]. In fact, it has been shown that more than The genomic plasticity response to such ambient temperature half of the regulatory connections in a gene expression changes can differ depending on the host genetic make-up or network are unique for specific conditions such as cell cycle, spore formation DNA damage and stress response [10]. New insights into genomic plasticity response to ambient

*Address correspondence to this author at the Department of Applied temperature in higher vertebrates are more than essential for Chemistry, Dongduk Women’s University, 23-1 Wolgok-dong, Sungbuk- discovery of biomarkers that stand the test of dynamic ku, Seoul 136-714, Korea; Tel 82-2-940-4514; Fax 82-2-940-4193; changes in environmental exposures. E-mail: [email protected]; [email protected]

In this study, our hypothesis builds on, and extends upon, anti-TRPA1 antibody (diluted in 1% BSA, 1:100) (Abcam, prior observations on gene-ambient temperature adaptive Cambridge, UK) overnight at 48°C. After the primary responses, and thermosensation specifically. In brief, antibody treatment, the cells were incubated with FITC- thermosensitive afferents express ion channels of the conjugated anti-rabbit IgG (diluted in 1% BSA, 1:500) transient receptor potential (TRP) family that respond at (Santa Cruz Biotechnology, Santa Cruz, CA) for 2 h at room distinct temperature thresholds, thus establishing a molecular temperature. Next, cells were counterstained using 4,6- basis for thermosensation. Transient receptor potential diamidino-2-phenylindole (DAPI; Sigma-Aldrich, St. Louis, melastatin 8 (TRPM8) and transient receptor potential A1 MO). The cover-slips were mounted on glass slides, and (TRPA1) are nonselective cation channels expressed on a fluorescence was detected using a Nikon fluorescence subset of peripheral afferent fibers. TRPA1 is activated by microscope (Nikon, Tokyo, Japan). Images were taken with the pungent ingredients in mustard and cinnamon, but it also a Nikon Eclipse TE2000-U camera (Nikon). has been postulated to mediate perception of noxious cold temperatures [11,12]. In particular, the molecular and 2.3. RNA Isolation and Microarray Analysis biochemical pathways inside cells that regulate cold-induced To investigate the alterations associated with low- responses are largely unknown. In general, cold exposure temperature adaptation after sensing at the molecular level, induces adaptive thermogenesis, which elevates energy differential genome analysis of BEAS-2B cells was expenditure in mammals via mitochondrial uncoupling [13]. performed by microarray analysis 3 h after treatment at 10°C The respiratory epithelium is constantly exposed to the for 30 min. Total RNA was isolated by following protocols external environment, and prolonged inhalation of cold air is of the RNeasy Mini Kit (Qiagen, Hilden, Germany) detrimental to human airways [14]. Physiologically, mild including DNAse digestion. The RNA was checked for cold exposure is known to elevate energy expenditure in integrity and purity on the Agilent 2100 Bioanalyzer mammals, including humans [13]. This process is known as (Agilent Technologies, Böblingen, Germany). Total RNA adaptive thermogenesis. In small animals, adaptive thermo- from each control and low-temperature treatment was used genesis is mainly achieved by mitochondrial uncoupling in for genome-wide gene expression profiling. Hybridization brown adipose tissue and regulated via the sympathetic on Agilent’s human whole genome 4 44K microarrays nervous system [13]. Transcriptional activation after thermo- was performed according to standard procedures supplied sensitive receptor binding could be coupled with changes in by the manufacturer (Agilent Technologies). In brief, RNA intracellular responses of representative determinants that was amplified and labeled using 50 ng of total RNA lead to adaptive thermogenesis and compensatory events due and Agilent’s Low RNA Input Linear Amplification Kit to cooling. Although environmental temperature inevitably PLUS to generate biotin-labeled complementary RNA. The influences an organism, the mechanism of systemic fragmented complementary RNA was hybridized to the gene temperature perception remains largely unknown [15]. chips for 16 h at 45°C. Arrays were washed and stained The present genome-wide gene expression study, using under standard conditions, and scanned on an Agilent DNA the human bronchial epithelial cell line BEAS-2B as a model microarray scanner. Feature Extraction Software was used to for proximal sensor, and mediator of compensatory events in control washing and scanning; to generate DAT, CEL, and the cooling of respiratory tract cells, aimed to identify a EXP files; and to process the raw data for signal calculation broad range of molecular correlates of adaptive responses to and pairwise chip comparison [16]. cold temperature. We herein show that a moderate cold temperature induces a genome-wide response and discuss its 2.4. Real-Time Polymerase Chain Reaction (PCR) biological relevance in relation to neuropathic and other To validate the differential genomic data, real-time PCR chronic pain syndromes that are in part moderated by host- was performed. Total RNA was extracted from cultured cells ambient temperature interactions. by homogenization in TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Synthesis 2. MATERIAL AND METHODS of cDNA was performed as previously described [17]. 2.1. Cell Culture Briefly, 50 μg of total RNA was reverse-transcribed to double-stranded cDNA using an oligo-dT primer. For PCR Human bronchial epithelial (BEAS-2B) cells (ATCC analysis, primer pairs were designed for each gene. The number CRL-9609) were purchased from the American information for each probe was obtained from the Stanford Type Culture Collection (Rockville, MD) and cultured in a Online Universal Resource for Clones and ESTs (SOURCE; bronchial epithelial cell growth medium bullet kit (BEGM http://www.source.stanford.edu), which compiles informa- kit) supplemented with a gentamicin–amphotericin B mix tion from publicly accessible databases, including UniGene, (Clonetics/Lonza Corp., Basel, Switzerland). Culture flasks dbEST, Swiss-Prot, GeneMap99, RHdb, GeneCards and (Corning, Corning, NY) for the cells were pre-coated LocusLink. Each probe was designed to have a melting at 37°C with BEGM fortified with collagen (30 mg/ml), temperature of approximately 60°C. For real-time PCR, 50 fibronectin (10 mg/ml), and bovine serum albumin (BSA; 10 ng of cDNA was mixed with 10 μM of each primer in SYBR mg/ml). Cells were maintained at 37°C in an air-ventilated Green Master Mix (Takara Bio Inc., Shiga, Japan). Real-time and humidified incubator with 5% carbon dioxide. PCR was performed on an ABI PRISM 7700 system (Applied Biosystems, Foster City, CA). All amplifications 2.2. Immunohistochemistry for TRPA1 were run in triplicate. Intron-specific primers were used to The cells were fixed with pre-chilled 1% paraformaldehyde control for genomic contamination. A no-template control for 1 min at room temperature and incubated with rabbit was performed for each primer pair. 18S RNA was selected Ambient Temperature and Host Genomics Plasticity Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 35 as the endogenous control. The conditions were optimized to each group comparison. Specifically, we mainly analyzed the show single peaks in the melting curves. CT was computed statistical mean values between the experimental group and by subtracting the CT (the number of cycles to reach the the baseline group. To find genes whose expression was threshold) for 18S RNA from the CT for each gene. The increased by cold sensation, we defined FC 2. Criterion for expression level of each gene, expressed in units relative to decreased expression was defined as FC 0.5. In addition, 18S RNA, was then taken as 2-CT. These units were the increased/decreased values had to be at least greater normalized to a value of 10,000 for the 18S RNA. PCR than 50% and the p-values of different t-test analysis at primer: zinc finger, X-linked, duplicated A (forward: 5’- least 0.001. All candidate genes were verified by repeating AGTTCTGTAGCACTGTCTATAAC-3’; reverse: 5’-TGA- the data analysis. GCACACATCCATTTGT-3’; amplicon size 116 bp); zinc finger protein 44 (forward: 5’- -3’; reverse: 5’- -3’, amplicon 3. RESULTS size bp); zinc finger DHHC-type containing 13 (forward: 5’- 3.1. Distribution of TRPA1 Receptors in the BEAS-2B CCATCATCATTACATCTTCTTCTTG -3’; reverse: 5’- Cells GGTCCATAGTCCATCTTCCT-3’, amplicon size 124 bp); zinc finger protein 423 (forward: 5’-GCAGAACCAC- We found that the TRPA1 receptors play a role in low- ACGATGAG-3’; reverse: 5’-GACACTTTGATTGGATGT- temperature perception. To this end, cultured BEAS-2B cells AATGTT-3’, amplicon size 124 bp); zinc finger, FYVE stained with goat anti-TRPA1 antibody and FITC-conjugated domain containing 20 (forward: 5’-TGCGGCTACAGA- IgG were used while DAPI was employed to detect the cell GAATGA -3’; reverse: 5’-TTGGTTGGCAGTGACATC - nucleus. As shown in Fig. (1), TRPA1 was detected whose 3’; amplicon size 82 bp); zinc finger protein 45 (forward: 5’- distribution resided within the cell body. GGTGGACTCGTGCCTACG-3’; reverse: 5’-CTGCCTGG- CTCTTTGTGTTA-3’, amplicon size 89 bp); zinc finger CCCH-type antiviral 1 (forward: 5’-GCTGGTTCACAGAA- TTACGA -3’; reverse: 5’-ATGCCTTTGAGTCTTGGAAG -3’, amplicon size 75 bp); zinc finger, CCHC domain containing 5 (forward: 5’-TCCAGCCCTACATCTCTAAA - 3’; reverse: 5’-CTCCGTGACTTTTCCAAGG -3’, amplicon size 125 bp); runt-related transcription factor 1translocated to 1 (forward: 5’-CGAGAAGCACTCCACAATG -3’; reverse: 5’-CGTTAATGTCGTTGGCGTAA -3’, amplicon size 136 bp); doublesex and mab-3 related transcription factor 1 (forward: 5’-TCGGTGCCTTAGGTTGTA -3’; reverse: 5’-AGTGTCTCCATTTACTGTCTG -3’, amplicon size 139 bp); signal transducer and activator of transcription 4 (forward: 5’-GAGCCACAATCTCCTTCAG -3’; reverse: 5’-CACCGTCATTCAGCAGAA -3’, amplicon size 125 bp); EF-hand calcium binding domain 1 (forward: 5’-ATAAGC- GTATCAGAGTGGATTC -3’; reverse: 5’-AACATCTCCT- CCTTTGAAATGA -3’, amplicon size 128 bp); heat shock transcription factor family, X linked 1 (forward: 5’-CCATT- Fig. (1). TRPA1 in BEAS-2B cells. TRPA1 was detected with a ATCCGGGAGCACAT-3’; reverse: 5’-CTCGCCATCTTT- rabbit anti-TRPA1 primary antibody and FITC-conjugated anti- CACTGGAT-3’, amplicon size 80 bp). rabbit IgG (1:500). The cells were counterstained with DAPI. The images were taken at a magnification of 400. 2.5. Data Analysis Initially, data obtained from all four microarray 3.2. Effect of Cold Temperature Treatment on Gene experiments were processed with Agilent's GeneSpring Expression in BEAS-2B Cells software. In pairwise chip comparison analyses, each of the two control cell samples and the two low temperature- The data were analyzed using ANOVA based on stimulated cultures were compared. Subsequently, pairwise comparison of samples after stimulation with low normalization and clustering of signal values were temperature (control versus 10°C treatment) that reached performed using Agilent's GeneSpring software to obtain p- statistical significance at the level of p < 0.001. The data values (absolute data) as well as signal log ratios (SLR). For showed reproducibility with a Pearson correlation coefficient each microarray, all signals were normalized to the 50th of 0.99 (Fig. 2). In the cDNA microarray containing percentile (median). For each of the selected two group 11,276 candidate genes, a total of 6,297 genes for each pair comparisons, the normalized signal was obtained as total were up-regulated and 4,979 genes were down-regulated fold change (FC) per gene. In more detail, for each of the with statistically significant differences. Among these selected two group comparisons (always consisting of one genes, the up- and down-regulated genes within the top baseline group with the pairwise cold-treatment culture), t- 1% according to the rank order of normalized ratios are test analyses on signals between two groups as well as on shown in Tables 1 and 2. SLR values of pairwise comparisons of chips within the Among the differentially expressed genes, we identified same group or between groups were performed as part of genes for biologically important transcription factors such as 36 Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 Park et al.

A. B. M Cy5-channelsignalintensity 10 100 1000 1e4 1e5 1e4 1000 100 10 0.01 0.1 1 10 100 100 10 1 0.1 0.01 100 1000 1e4 1e5 100 1000 1e4 1e5

Cy3-channel signal intensity A

Fig. (2). Gene expression profile as Scatter plot (A) and MA plot (B). The scatter plot shows the distribution of signal intensities of Cy3 and Cy5 probes. The MA plot shows the relationship between average signal intensity of each channel and the R/G ratio. R = Cy5 signal – background; G = Cy3 signal – background; M = R/G; A = (R+G)/2. These plots collectively show that the experiment meets the condition of the binding isotherm for data analysis.

Table 1. Up-regulated genes within the top 1% (normalized ratios).

Accession No. Gene Normalized Ratio

NM_001774 CD37 molecule 333.21417

NM_005117 Fibroblast growth factor 19 180.40018

NM_178860 Seizure related 6 homolog (mouse) 139.83392

AK027082 Unnamed protein product; Homo sapiens cDNA: FLJ23429 fis, clone HRC10578. 116.02334

NM_012199 Eukaryotic translation initiation factor 2C, 1 89.32227

KIAA0226 79.33165

Human tenascin-X (XB) mRNA, RACE clone N1, 77.33808

NM_003636 Potassium voltage-gated channel, shaker-related subfamily, beta member 2 72.27391

CR615589 Hypothetical protein BC004941 65.74664

NM_173465 Collagen, type XXIII, alpha 1 54.638077

AF018283 Runt-related transcription factor 1; translocated to, 1 (cyclin D-related) 52.800835

NM_004673 Angiopoietin-like 1 51.49603

NM_016079 Vacuolar protein sorting 24 homolog (S. cerevisiae) 51.17402

NM_007156 Zinc finger, X-linked, duplicated A 51.08436

BC032246 Zinc finger protein 44 50.8672

NM_138342 Hypothetical protein BC008326 48.847996

Q9CXZ8 (Q9CXZ8) Mus musculus 13 days embryo liver cDNA, RIKEN full-length enriched library, clone:2510048A01 product:sudD, suppressor of bimD6 homolog (Aspergillus nidulans), full insert sequence. 45.0847 (Fragment), partial (52%) [THC2343602]

NM_178864 Neuronal PAS domain protein 4 44.408825

Ambient Temperature and Host Genomics Plasticity Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 37

Table 1. contd….

Accession No. Gene Normalized Ratio.

NM_001012361 WD repeat domain 31 42.68389

NM_021637 Transmembrane protein 35 41.830597

NM_130830 Leucine rich repeat containing 15 41.42196

AK025420 Hypothetical protein FLJ21767 40.97265

H.sapiens mRNA HTPCRX03 for olfactory receptor. [X64981] 40.75834

NM_021951 Doublesex and mab-3 related transcription factor 1 40.250618

NM_003578 Sterol O-acyltransferase 2 39.35871

AB064188 immunoglobulin lambda light chain VLJ region {Homo sapiens;}, partial (90%) [THC2428956] 38.13184

AW136683 Immunoglobulin J polypeptide, linker protein for immunoglobulin alpha and mu polypeptides 37.472565

NM_015162 Acyl-CoA synthetase bubblegum family member 1 35.47174

NM_004789 LIM homeobox 2 35.394

NM_153204 Homo sapiens chromosome 21 open reading frame 90 (C21orf90), mRNA [NM_153204] 35.18354

NM_006574 Chondroitin sulfate proteoglycan 5 (neuroglycan C) 35.0884

NM_025257 Solute carrier family 44, member 4 34.368736

NM_006424 Solute carrier family 34 (sodium phosphate), member 2 34.063385

Y11718 Sperm associated antigen 10 33.94468

NM_198060 Nebulin-related anchoring protein 33.387276

NM_031911 C1q and tumor necrosis factor related protein 7 33.27971

AK023328 CDNA FLJ13266 fis, clone OVARC1000960 32.460636

NM_205848 Synaptotagmin VI 32.279503

BX118153 Transcribed locus 31.316563

Homo sapiens mRNA; cDNA DKFZp762M126 (from clone DKFZp762M126). [AL834437] 30.902111

NM_023013 PRAME family member 1 30.286865

AK126415 FLJ44451 fis 30.002148

NM_002071 Guanine nucleotide binding protein (G protein), alpha activating activity polypeptide, olfactory type 29.60639

AK098307 SPEG complex locus 29.46181

BQ441611 Transcribed locus 28.686844

NM_018203 Kelch domain containing 8A 27.98057

AF086413 MCF.2 cell line derived transforming sequence-like 27.778017

NM_003151 Signal transducer and activator of transcription 4 26.862337

Hypothetical protein BC007652 26.156412

AK094219 CDNA FLJ36900 fis, clone BRACE2001954 26.12895

NM_173669 Homo sapiens hypothetical protein FLJ34047 (FLJ34047), mRNA [NM_173669] 25.530155

NM_138574 Hepatoma derived growth factor-like 1 25.484568

Q9VYV2 (Q9VYV2) CG1841-PA, isoform A Cg1841 isoform b) (HL07962p), partial (5%) [THC2373555] 25.092865

NM_001140 Arachidonate 15-lipoxygenase 25.072807

AL832535 Hypothetical protein LOC157627 25.06273 38 Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 Park et al.

Table 2. Down-regulated genes within the top 1% (normalized ratios).

Accession No. Gene Normalized Ratio

NM_002135 Nuclear receptor subfamily 4, group A, member 1 0.01

NM_018569 Chromosome 4 open reading frame 16 0.01

AB051444 KIAA1657 protein 0.01

AK022348 Spermatogenesis associated 5-like 1 0.01

NM_021643 Tribbles homolog 2 (Drosophila) 0.012346361

NM_205767 Hypothetical protein P117 0.01347182

AK023049 DEP domain containing 2 0.013801249

NM_018086 Fidgetin 0.015192012

NM_031469 SH3 domain binding glutamic acid-rich protein like 2 0.015197019

NM_019043 Amyloid beta (A4) precursor protein-binding, family B amily B, member 1 interacting protein 0.015849194

BX097529 Transcribed locus 0.016685026

NM_000406 Gonadotropin-releasing hormone receptor 0.01694819

NM_000399 Early growth response 2 (Krox-20 homolog, Drosophila) 0.017471364

NM_199367 Spastic paraplegia7(pure and complicated autosomal recessive) recessive) 0.019291563

BC000825 Hypothetical protein LOC339524 0.019605318

Homo sapiens mRNA; cDNA DKFZp434I0714 ZFZp434I0714). [AL137273] 0.020745458

NM_033179 Olfactory receptor, family 51, subfamily B, member 4 0.020929407

NM_022350 Leukocyte-derived arginine aminopeptidase 0.021294603

AK094853 CDNA FLJ37534 fis, clone BRCAN2014788 0.022425825

AK001697 RIO kinase 2 (yeast) 0.022459578

NM_000420 Kell blood group, metallo-endopeptidase 0.022932287

NM_001879 Mannan-binding lectin serine peptidase 1 (C4/C2) 0.02435484

NM_001004303 Chromosome 1 open reading frame 168 0.02446145

NM_031889 Enamelin 0.02487484

NM_020215 Chromosome 14 open reading frame 132 0.02618164

AF119909 Predicted protein of HQ2958; Homo sapiens PRO2958 mRNA, complete cds. 0.026959565

NM_147199 MAS-related GPR, member X1 0.027367838

NM_031935 Hemicentin 1 0.030843474

NM_138812 DPY30 domain containing 1 0.03226938

NM_001346 Diacylglycerol kinase, gamma 90kDa 0.03257056

AL136578 Transmembrane protein 108 0.032871775

NM_002055 Glial fibrillary acidic protein 0.03350383

AF130093 Clone FLB9835 PRO2660 0.034032073

NM_003014 Secreted frizzled-related protein 4 0.034611426

NM_182758 WD repeat domain 72 0.03504397

AK001357 CDNA FLJ10495 fis, clone NT2RP2000297similar to ZF ZROTEIN 184 0.03555011 Ambient Temperature and Host Genomics Plasticity Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 39

Table 2. contd….

Accession No. Gene Normalized Ratio

NM_004371 Coatomer protein complex, subunit alpha 0.03571115

NM_004024 Activating transcription factor 3 0.036004264

NM_000590 Interleukin 9 0.036265675

NM_003425 Zinc finger protein 45 0.037050832

NM_024625 Zinc finger CCCH-type, antiviral 1 0.038655728

Homo sapiens mRNA for KIAA0518 protein, partial cds. [AB011090] 0.03887354

NM_003053 Solute carrier family 18 (vesicular monoamine), member 1 0.03916563

NM_000091 Collagen, type IV, alpha 3 (Goodpasture antigen) 0.04071001

AF090906 Rho guanine nucleotide exchange factor (GEF) 12 0.040881846

NM_002562 Purinergic receptor P2X, ligand-gated ion channel, 7 0.0422606

NM_003154 Statherin 0.042319905

NM_015914 Thioredoxin domain containing 11 0.043061946

NM_001481 Growth arrest-specific 8 0.04362062

NM_003670 Basic helix-loop-helix domain containing, class B, 2 0.044151098

NM_020726 Neurolysin (metallopeptidase M3 family) 0.044226438

NM_000093 Collagen, type V, alpha 1 0.044264805

NM_005060 RAR-related orphan receptor C 0.044775575

zinc finger proteins ZXDA, ZNF44, ZDHHC13, ZNF313, ZNF313, ZNF423, ZFYVE20, ZNF45, ZC3HAV, and ZNF423, ZFYVE20, ZNF45, ZC3HAV and ZCCHC5. These ZCCHC5) and other transcription factors (RUNX1T1, transcription factor genes are known to be involved in DMRT1, STAT4, EFCAB1, and HSFX1). The FC in immune cell growth and proliferation, cell signal trans- expression of these genes, normalized to 16S rRNA, ranged duction, and transport. Other differentially regulated genes from 7.7 (for HSFX1) to approximately 36 (for ZNF44) were classified by function as shown in Table 3. (p<0.05) compared to levels observed in untreated cells. The expression levels of these genes in the untreated and cold 3.3. Verification of Differentially Expressed Genes temperature-treated cells were consistent with the microarray We further evaluated, by quantitative RT-PCR, the data, although the values were lower than those obtained in mRNA levels of a subset of genes, including zinc finger the arrays (Fig. 3). transcription factor genes (ZXDA, ZNF44, ZDHHC13,

Table 3. Number of differentially expressed genes identified by function among the top 10% of up- and down-regulated genes with normalized ratios.

Number of Genes Affected

Function/Pathway Total Up-Regulated Down Regulated

Transcriptional factors 39 19 20

Channel and carrier protein 28 19 9

Cell signaling 39 21 18

Immune protein 17 12 5

Receptor and growth factor 81 45 36

40 Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 Park et al.

50 Control 40 * * * + Treatment 30

20 * * * * * 10 * *

-10 * -20 * * -30 ZXDA ZNF44 ZDHHC13 ZNF423 ZFYVE20 ZNF45 ZC3HAV ZCCHC5 RUNX1T1 DMRT1 STAT4EFCAB1HSFX1

Gene Names

Fig. (3). RT-PCR analysis of mRNA levels. Each mRNA level was normalized to the 18S RNA level. Each data point represents a mean of three individual values and standard deviations. Quantitative analysis was performed, and results were expressed as activity relative to the untreated control group. Asterisks indicate statistically significant differences between treatment and untreated control conditions (*p<0.01). ZXDA, zinc finger, X-linked, duplicated A; ZNF44, zinc finger protein 44; ZDHHC13, zinc finger DHHC-type containing 13; ZNF423, zinc finger protein 423; ZFYVE20, zinc finger, FYVE domain containing 20; ZNF45, zinc finger protein 45; ZC3HAV, zinc finger CCCH-type antiviral 1; ZCCHC5, zinc finger, CCHC domain containing 5; RUNX1T1, runt-related transcription factor 1 translocated to 1; DMRT1, doublesex and mab-3 related transcription factor 1; STAT4, signal transducer and activator of transcription 4; EFCAB1, EF-hand calcium binding domain 1; HSFX1, heat shock transcription factor family, X linked 1.

4. DISCUSSION at 10°C. We focused on transcription factor-related genes in this experiment because the effect of changes in transcription Pharmacogenomics has thus far been framed rather factor genes could be amplified by other downstream target narrowly, particularly in the context of discerning the genes changes. Selected transcription factor genes were genome-drug and genome-environment interactions that confirmed by real-time PCR. The changes observed in gene crucially underpin person-to-person and population variability expression 3 h after a 30-min treatment at 10°C support a in drug efficacy and safety. These past discussions have possible link between transcriptional regulation and specific tended to neglect the role that is played by dynamic modulation of the innate immune response or adaptive genomics plasticity that can be induced by ambient response. So far, the zinc finger proteins ZXDA and ZNF44 temperature changes. It is in this context that the present have been identified as being involved in MHC transcription study, using an in vitro cell culture model, makes in humans [19-21]. The DMRT1 gene was found in a cluster an important contribution to help broaden the extant with two other members of the gene family, having in frameworks on genome-environment interactions in the field common a zinc finger-like DNA-binding motif [22]. of pharmacogenomics, such that a broader range of variables are considered to account for genome-drug-environment We underscore that the epithelial cells of the trachea and interactions. This broader frame is a key step toward under- bronchi are constantly exposed to inspired air [13]. Human standing why individuals respond differently to drugs, airways thus exhibit a complex pattern of responses when toxins, pathogens, nutrients and other environmental exposed to cold air, including altered cytokine profiles and influences, a key focus of the CPPM and its readership in an increase in the number of inflammatory cells in the lungs postgenomics personalized medicine. Moreover, because all [13]. Hence, recognition of temperature is an important step species are subject to environmental change, both at for cells to evaluate their external environment and adapt individual and evolutionary time scales, the presented work by changing their internal status. Physiologically, mild here calls for studying the plasticity of gene regulation in cold exposure is known to elevate energy expenditure humans in vivo to enhance understanding of the ambient in mammals, including humans [13]. Exposure to cold air temperature related environmental forces. The latter shape often leads to asthma-like symptoms and is commonly accompanied by chronic airway inflammation [13]. Human not only evolutionary adaptation but also drug pharma- bronchial epithelial cells are capable of synthesizing and cokinetics and pharmacodynamics [18]. releasing a multitude of biochemical mediators, including We identified expression of TRPA1 in the human cytokines and chemokines, upon stimulation by cold [23-25]. bronchial epithelial cell line BEAS-2B and used this cell However, the precise and immediate mechanism by which line to obtain a differential genome profile before and cold air promotes an inflammatory response is not well after moderate changes in the ambient temperature. More understood. In this regard, it is noteworthy that we identified importantly, we observed differential broad gene expression herein transcription factor genes with either increased or changes in the bronchial epithelial cell line BEAS-2B treated decreased differential expression as candidates of interest in Ambient Temperature and Host Genomics Plasticity Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 41 the context of cold adaptation. Our results suggest early persistent and chronic pain syndromes. Moreover, the activation of genes within 3 h of exposure to cold findings establish a new crucial putative link between temperature. Based on these results, these candidate genes ambient temperature and genomic plasticity that together might be responsible for the subsequent physiological inform personalized medicine such that future responses. pharmacogenomics biomarker discovery research can better control and account for ever-present dynamic environmental 5. CONCLUSIONS AND FUTURE OUTLOOK exposures such as ambient temperature.

Recognition of temperature is a critical element of ABBREVIATIONS sensory perception both at a cellular and whole organism level. With regard to cold sensation, Story et al. first AMPA = 2-Amino-3-(5-methyl-3-oxo-1,2- oxazol-4- suggested that the TRP-like channel TRPA1 (or ANKTM1) yl)propanoic acid mediates cold-responsiveness in a cold-sensitive, menthol- BEGM = Bronchial epithelial cell growth medium insensitive population of sensory neurons when they bullet reported that noxious cold temperatures activated the mouse ortholog of this ion channel [11]. In vertebrates, the DAPI = 4,6-Diamino-2-phenylondole somatosensory system can discriminate discrete changes in DMRT1 = Doublesex and mab-3 related transcription ambient temperature, which activate nerve endings of factor 1 primary afferent fibers. These thermosensitive nerves can be further segregated into those that detect either innocuous or EFCAB1 = EF-hand calcium binding domain 1 noxious (painful) temperatures, the latter neurons being nociceptors [11]. Interestingly, once temperatures approach ER = Endoplasmic reticulum 15°C, the perception of cold pain is evoked with qualities GABA = Gamma amino butyric acid described as burning, aching, and pricking [26]. The abnormal thermal-sensation is associated with neuropathic HSFX1 = Heat shock transcription factor family, X pain. Among the more common and important clinical signs linked 1 in neuropathic pain disorders are positive sensations: NMDA = N-methyl-D-aspartate stimulus-evoked hypersensitivities, such as allodynia to innocuous stimulation (e.g., light touch and cold), and PCR = Polymerase chain reaction hyperalgesia to noxious stimulation (e.g., pinprick). Various RUNX1T1 = Runt-related transcription factor forms of hyperalgesia increase with repetitive stimulation 1translocated to 1 (windup-like pain) [27]. Paradoxically, these hypersensitivities can occur in areas in which the patient also SR protein = Arginine/serine rich protein complains of and demonstrates loss of sensation [28]. As STAT4 = Signal transducer and activator of with symptoms, the spread of allodynia and hyperalgesia transcription 4 outside the original site of injury is common and may extend to homologous sites in the opposite limb [28]. Neuropathic TRPA1 = Transient receptor potential A1 pain occurs when there is damage to the nervous system. TRPM8 = Transient receptor potential melastatin 8 These past observations together lead us to suggest that neuropathic pain state may be induced by genetic changes. ZCCHC5 = Zinc finger, CCHC domain containing 5 This is notable for global public health because nearly more ZC3HAV = Zinc finger CCCH-type antiviral 1 than one-third of the world’s population suffers from persistent or recurrent pain [29]. Diabetic neuropathy is the ZDHHC13 = Zinc finger DHHC-type containing 13 most common complication of diabetes. More than 50% of ZFYVE20 = Zinc finger, FYVE domain containing 20 diabetic patients are affected by diabetic neuropathy leading to sensory loss. Yet the available drug targets for neuropathy ZNF423 = Zinc finger protein 423 and chronic pain syndromes remain limited [29]. The most ZNF44 = Zinc finger protein 44 promising targets under study to date include sodium channel blockers, calcium channel blockers, drugs affecting ZNF45 = Zinc finger protein 45 the GABAergic system, and N-methyl-D-aspartate (NMDA) and 2-amino-3-(5-methyl-3-oxo-1,2- oxazol-4-yl) propanoic ZXDA = Zinc finger, X-linked, duplicated A acid (AMPA) antagonists. CONFLICT OF INTERESTS Against this overarching background, this genome- None declared/applicable. wide study contributes to our understanding of cold thermoreceptor-dependent molecular changes while in part ACKNOWLEDGEMENTS addressing cold-adaptive responses in an in vitro model. The presented results therefore provide a comprehensive list of This research was supported by a Korea Science and new molecular leads, in addition to and beyond TRPA1, for Engineering Foundation (KOSEF) grant funded by the future analyses of gene signatures of cold temperature Korean Government (Program for Basic Science: 2010- adaptation to identify novel drug targets for personalized 0000580 and 2011-0011014). We thank the two anonymous drug design, particularly those that may be well suited for peer reviewers for constructive critique and comments that 42 Current Pharmacogenomics and Personalized Medicine, 2012, Vol. 10, No. 1 Park et al. improved the discussion in the manuscript. All authors a TRPM8 variant by cold temperatures. Am J Physiol Lung Cell made a significant contribution to conception and design, Mol Physiol 2008; 295: L194-L200. [14] Wijers SLJ, Schrauwen P, Saris WHM, et al. Human skeletal acquisition of data, or analysis and interpretation of data; muscle mitochondrial uncoupling is associated with cold induced drafting the article or revising it critically for important adaptive thermogenesis. PLos One 2008; 3: e1777. intellectual content; and approved the final version to be [15] Sugi T, Nishida Y, Mori Y. Regulation of behavioral plasticity by published. systemic temperature signaling in Caenorhabditis elegans. Nat Neurosci 2011; 14: 984-92. [16] Menssen A, Edinger G, Grun JR, et al. SiPaGene: A new REFERENCES repository for instant online retrieval, sharing and meta-analyses of [1] Li Y, Breitling R, Jansen RC. Generalizing genetical genomics: GeneChip expression data. BMC Genomics 2009; 10: 98. getting added value from environmental perturbation. Trends Genet [17] Yang SH, Kim JS, Oh TJ, et al. Genome-scale analysis of 2008; 24: 518-24. resveratrol-induced gene expression profile in human ovarian [2] Xue Y, Haas SA, Brino L, et al. A DNA microarray for fission cancer cells using a cDNA microarray. Int J Oncol 2003; 22: yeast: Minimal changes in global gene expression after temperature 741-50. shift. Yeast 2004; 21: 25-39. [18] Li Y, Álvarez OA, Gutteling EW, et al. Mapping Determinants of [3] Kraus PR, Boily MJ, Giles SS, et al. Identification of Cryptococcus Gene Expression Plasticity by Genetical Genomics in C. elegans. neoformans temperature-regulated genes with a genomic-DNA PLoS Genet, 2006; 2(12): e222. microarray. Eukaryot Cell 2004; 3: 1249-60. [19] Al-Kandari W, Koneni R, Navalgund V, et al. The zinc finger [4] GuhaThakurta D, Palomar L, Stormo GD, et al. Identification of a proteins ZXDA and ZXDC form a complex that binds CIITA and novel cis-regulatory element involved in the heat shock response in regulates MHC II gene transcription. J Mol Biol 2007; 369(5): Caenorhabditis elegans using microarray gene expression and 1175-87. computational methods. Genome Res 2002; 12: 701-12. [20] Brandenberger R, Wei H, Zhang S, et al. Transcriptome [5] Roff DA. Life history evolution. Sunderland (Massachusetts): characterization elucidates signaling networks that control human Sinauer Associates. 2002; pp. 527. ES cell growth and differentiation. Nat Biotechnol 2004; 22(6): [6] Crews D, Bergeron JM, Bull JJ, et al. Temperature-dependent sex 707-16. determination in reptiles: Proximate mechanisms, ultimate outcomes, [21] Lehner B, Semple JI, Brown SE, et al. Analysis of a high- and practical applications. Dev Genet 1994; 15: 297-312. throughput yeast two-hybrid system and its use to predict the [7] Roskam JC, Brakefield PM. A comparison of temperature-induced function of intracellular proteins encoded within the human MHC polyphenism in African Bicyclus butterflies from a seasonal class III region. Genomics 2004; 83(1): 153-67. [22] Raymond CS, Shamu CE, Shen MM, et al. Evidence for savannahrainforest ecotone. Evolution 1996; 50: 2360-72 [8] Landry CR, Oh J, Hartl DL, et al. Genome-wide scan reveals evolutionary conservation of sex-determining genes. Nature 1998; that genetic variation for transcriptional plasticity in yeast is 391(6668): 691-5. biased towards multi-copy and dispensable genes. Gene 2006; 366: [23] Campbell AM. Bronchial epithelial cells in asthma. Allergy 1997; 343-51. 52: 483-9. [9] Levins R. Toward a population biology, still. In: Rama S., Singh [24] Knight DA, Holgate ST. The airway epithelium: structural and MKUeditor. The evolution of population biology, Cambridge functional properties in health and disease. Respirology 2003; 8: (United Kingdom): Cambridge University Press. 2004; pp. 21-48. 432-46. [10] Luscombe NM, Babu MM, Yu H, et al. Genomic analysis of [25] Levine SJ. Bronchial epithelial cell-cytokine interactions in airway regulatory network dynamics reveals large topological changes. inflammation. J Investig Med 1995; 43: 241-9. Nature 2004; 431: 308-12. [26] Morin C, Bushnell MC. Temporal and qualitative properties of cold [11] Story GM, Peier AM, Reeve AJ, et al. ANKTM1, a TRP-like pain and heat pain: a psychophysical study. Pain 1998; 74: 67-73. channel expressed innociceptive neurons, is activated by cold [27] Jensen TS, Baron R. Translation of symptoms and signs into temperatures. Cell 2003; 112: 819-29. mechanisms in neuropathic pain. Pain 2003; 102(1-2): 1-8 [12] McKemy DD. How cold is it? TRPM8 and TRPA1 in the [28] Horowitz SH. The diagnostic workup of patients with neuropathic molecular logic of cold sensation. Mol Pain 2005; 1: 16. pain. Anesthesiology Clin 2007; 25(4): 699-708. [13] Sabnis AS, Reilly CA, Veranth JM, et al. Increased transcription of [29] Jensen MP, Chordroff MJ, Dworkin RH. The impact of neuropathic cytokine genes in human lung epithelial cells through activation of pain on health-related quality of life. Neurology 2007; 68: 1178-82.

Received: November 17, 2011 Revised: January 02, 2012 Accepted: January 05, 2012