ADOA3R as a Therapeutic Target in Experimental Colitis: Proof by Validated High-density Oligonucleotide Microarray Analysis

ADOA3R as a Therapeutic Target in Experimental Colitis:Proof by Validated High-density Oligonucleotide

Microarray Analysis

Jorge Guzman, PhD,*,** Jun Ge Yu, PhD,*,** Zacharias Suntres, PhD,*** Andrey Bozarov, MD,.**Helen Cooke, PhD,.** Najma Javed, MD, PhD,L Herbert Auer, MSc,`** Jeff Palatini, BSc,`**

Hamdy H. Hassanain, PhD,*||P** Arturo J. Cardounel, PhD,P#** Asad Javed, BSc,*,**Iveta Grants, BSc,*,** Jacqueline E. Wunderlich, MD, PhD,** and Fievos L. Christofi, PhD*,**

Abstract: Adenosine A3 receptors (ADOA3Rs) are emerging asnovel purinergic targets for treatment of inflammatory diseases. Ourgoal was to assess the protective effect of the ADOA3R agonistN(6)-(3-iodobenzyl)-adenosine-5-N-methyluronamide (IB-MECA)on gene dysregulation and injury in a rat chronic model of 2,4,6-trinitrobenzene sulfonic acid (TNBS)Y induced colitis. It wasnecessary to develop and validate a microarray technique for testingthe protective effects of purine-based drugs in experimentalinflammatory bowel disease. High-density oligonucleotide micro-array analysis of gene dysregulation was assessed in colons fromnormal, TNBS-treated (7 days), and oral IB-MECAY treated rats(1.5 mg/kg b.i.d.) using a rat RNU34 neural GeneChip of 724 genesand SYBR green polymerase chain reaction. Analysis included clinicalevaluation, weight loss assessment, and electron paramagnetic reso-nance imaging/spin-trap monitoring of free radicals. Remarkablecolitis-induced gene dysregulation occurs in the most exceptionalcluster of 5.4% of the gene pool, revealing 2 modes of colitis-relateddysregulation. Downregulation occurs in membrane transporter,mitogen-activated protein (MAP) kinase, and channel genes. Upregu-lation occurs in chemokine, cytokine/inflammatory, stress, growthfactor, intracellular signaling, receptor, heat shock protein, retinoidmetabolism, neural, remodeling, and redox-sensitive genes. Oral IB-MECA prevented dysregulation in 92% of these genes, histopathology,

gut injury, and weight loss. IB-MECA or adenosine suppressedelevated free radicals in ex vivo inflamed gut. Oral IB-MECA blockedthe colitis-induced upregulation (e20-fold) of Bzrp, P2X1R, P2X4R,P2X7R, P2Y2R, P2Y6R, and A2aR/A2bR but not A1R or A3R genes ordownregulated P2X2R, P2Y1R, and P2Y4R. Real-time SYBR greenpolymerase chain reaction validated gene chip data for both inductionof colitis and treatment with IB-MECA for 990% of genes tested (33of 37 genes). We conclude that our validated high-density oligonu-cleotide microarray analysis is a powerful technique for moleculargene dysregulation studies to assess the beneficial effects of purine-based or other drugs in experimental colitis. ADOA3R is newpotential therapeutic target for inflammatory bowel disease.

Key Words: TNBS colitis, microarray, adenosine A3 receptors,antioxidant effects, IB-MECA, gene dysregulation, purine receptorgene dysregulation

(Inflamm Bowel Dis 2006;12:766Y789)

Many substances have been used to prevent the develop-ment of experimental colitis and inflammatory bowel

disease (IBD), but to date, available treatments for Crohn_sdisease (CD) and ulcerative colitis (UC) are limited. Modelsof IBD used to test new therapeutic agents include chemical,microbial, or polymer induction; genetic models; andspontaneous colitis models. The rat 2,4,6-trinitrobenzene sul-fonic acid (TNBS) model is one of the most frequently usedmodels of experimental colitis to test new therapeutic strat-egies, and it has attributes of both UC and CD.1 Adenosine(ADO) receptors may offer a novel therapeutic target in gutinflammation, IBD,2 and other diseases.3Y6 Endogenous ADO(eADO) is a metabolite of adenosine triphosphate (ATP) thatacts at ADOA1R, ADOA2aR, ADOA2bR, or ADOA3Rs toinfluence mucosal secretion and neural sensory motor reflexes.ATP itself or other related purinergic mediators act at additionalP2YRs and P2XRs to influence gut reflexes and function.7Y9

In general, ADO may exert anti-inflammatory effectsby inhibiting synthesis of TH1 cytokines (i.e., interferon-g

ORIGINAL ARTICLE

766 Inflamm Bowel Dis & Volume 12, Number 8, August 2006

Received for publication April 3, 2006; accepted April 24, 2006.From the Departments of *Anesthesiology, †Neuroscience, zCancer Genetics,

and ||Cardiothoracic Surgery, the PDorothy Heart and Lung Institute, andthe #Department of Pharmacology, Ball State University, Muncie, Indiana;**Ohio State University, Columbus; and ***Lakehead University, ThunderBay, Canada.

This work was sponsored by the National Institutes of Health (NIH) grants RO1DK44179 (F.L.C.), DK44179-10S2 (F.L.C.), RO1 DK 57016 (H.J.C.), andDK 37240 (H.J.C.), and the Minority Fellowship to Jorge Guzman, an MD/PhD student in the Medical Scientist Program (grant DK44179-07S1).

Reprints: Fievos L. Christofi, Department of Anesthesiology, College ofMedicine and Public Health, Ohio State University, 226 TzagournisMedical Research Facility, 420 W 12th Avenue, Columbus, OH 43210(e-mail: [email protected])

Copyright * 2006 by Lippincott Williams & Wilkins

Copyr ight © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

[INF-g], tumor necrosis factor-a [TNF-a]) and suppressingneutrophil functions, including degranulation, superoxideproduction, and adhesion to and damage to the endothe-lium.10Y13 Clinical use of ADO, ADOR agonists, or certainmodulators of ADO degradation may be complicated byundesirable pharmacokinetics, side effects, or progressivedesensitization of certain ADORs14,15 that may limit theirusefulness in IBD2 or other diseases.2,9,16Y21 Increasing theconcentration of eADO is one mechanism by which severaldrugs may reduce gut inflammation.20,22Y24 An ADO kinaseinhibitor (used in dextran sulfate [DDS] colitis) or anADOA3R agonist N6-(3-iodobenzyl)-adenosine-5-N-methyl-uronamide (IB-MECA) may be beneficial in murine modelsof colitis,16,17 but the mechanisms involved remain poorlyunderstood. ADOA3Rs in particular have a wide range ofphysiological and disease-related effects,5,9,25,26 with promisefor treatment of a variety of heart disease, uveitis, colorectalcancer, and inflammation.19 IB-MECA is in phase I and IIclinical trials for a chronic inflammatory disease, rheumatoidarthritis, and is apparently without toxicity (www.canfite.com/develop.html).

In the digestive tract, ADOA3Rs seem to play animportant role in the function of the distal colon that is theprimary site of inflammation in rat TNBS colitis. ADOA3Rsare expressed throughout the human, rat, guinea pig, andmouse colon.9,27 A3R immunoreactivity is differentiallyexpressed, with the highest expression being in the distalcolon of the rat. Our initial studies also identified aninhibitory limb of the motor reflex that may be activated byputative A3Rs in rat distal colon.9 ADOA3Rs modulate theneural reflexes involved in the coordinated motility andsecretory responses in rodent colon (F.L.C., unpublishedobservations).

We were interested in probing further the complexmechanisms that may be involved in any beneficial effects ofthe ADOA3R agonist IB-MECA in the rat TNBS colitis modeland whether such treatment can prevent or attenuate alterationsin gene expression and dysregulation profiles. No attempts havebeen made to establish gene expression profiles in experimentalmodels of colitis or IBD such as in TNBS colitis using high-density oligonucleotide microarray analysis. Validation ofmicroarray data is necessary through other approaches, includ-ing real-time quantitative polymerase chain reaction (PCR) andparallel analysis of histopathology and function. A recentmicroarray study in human IBD has compared and contrastedthe patterns of gene expression and dysregulation in UC, CD,and other chronic inflammatory conditions.28

Activation of ADOA3Rs in various cell types leads toan increase in activity of the cellular antioxidant enzymes,29

but it remains unknown whether ADOA3R agonists haveantioxidant effects in a model like TNBS colitis, whichrepresents a suitable model to study oxidative stress.1

Our approach is to study the effects of an ADOA3R

agonist on oxidative stress genes or free radical generationin colitis tissues by electron paramagnetic resonanceimaging (EPRI).

Expression of purine-related genes is modulated byinflammatory or other diseases or drug treatments.30,31 There-fore, it was also of interest to establish whether purine receptorgene dysregulation occurs in TNBS colitis for adenosine, P2X,and P2Y gene receptor families. This is critical in theassessment and interpretation of purine-based therapies inexperimental models of IBD. In particular, oral IB-MECA ispresumed to be highly selective for A3Rs in the treatment ofinflammatory diseases such as rheumatoid arthritis in humans,but definitive proof is lacking. It remains possible that in vivoactivity of IB-MECA is not restricted to ADOA3R, and itcould act on other ADO receptors such as A1R, A2aR, orA2bR. The ADOA3R, a potential therapeutic target, may itselfbe subject to dysregulation by inflammation.30,31

Our goal was to assess the protective effect of theADOA3R agonist IB-MECA on gene dysregulation andinjury in a rat chronic model of TNBS-induced colitis. It wasnecessary to develop and validate a microarray technique totest the protective effects of purine-based drugs in exper-imental IBD. We applied high-density oligonucleotide micro-array analysis using a rat RNU34 neurobiology chip of 724genes to identify gene expression patterns of dysregulationand to assess the prophylactic effect of the ADOA3R agonistIB-MECA on gene dysregulation and gut injury. Immune,cytokine, chemokine, inflammatory, epithelial transporter,growth factor, and oxidant/antioxidant genes that are relevantin IBD were examined using the RNU34 chip, and thefindings validated by real-time SYBR green PCR for selectedclusters of genes. Tissue injury and inflammation were con-firmed by histological and clinical evaluations and weight loss.The antioxidant potential of IB-MECA also was assessed exvivo by spin-trap/EPRI of free radicals and oxidant enzymegene expression profiles. Validated gene array analysisrevealed useful cluster information on neural and nonneuralgenes that can be targeted easily in drug treatment studies. Wepresent novel data that oral IB-MECA nearly prevents ratTNBS colitis and gene dysregulation, suggesting thatADOA3R is a new potential therapeutic target in IBD.

MATERIALS AND METHODS

Induction of Inflammation andIB-MECA Treatment

Harlan Sprague-Dawley Lewis rats (Harlan, India-napolis, Ind) were randomized to receive enemas with saline(0.9% NaCl) or TNBS (60 mg in 33% ethanol, 1.0 mL totalvolume). Some animals with TNBS received IB-MECA(1.5 mg/kg b.i.d.) by oral gavage, and age-matched controls

Inflamm Bowel Dis & Volume 12, Number 8, August 2006 ADOA3Rs Prevent Colitis Gene Dysregulation

* 2006 Lippincott Williams & Wilkins 767


TABLE 1. Primers Used for SYBR Green PCR in Colonic Tissues from Age-matched Control and TNBS-induced Colitis Rats

Gene Forward Sequence Reverse Sequence

Primers used to confirm gene dysregulation demonstrated with microarray

Ania4 GGGAAATACGTCGTCTGAGGCCATC TTATTACGTGGTTTGGCTAAGCTGC

Selp TGGACACCACATCTGACTGCAGACT CTCTCTGTGGACAGGCGACATACAA

Cebpb CCTGCACCGCGCACCGGGTTTCGGG AACATCAACAGCAACAACCCCGCAG

Cxcl2 AGACGCAGTCAGATGGCTTTCAAGG TGCCTTACAGAGAAGACATAGAATG

Bzrp TGGTGGACCTCATGCTTGTCAGTGG GCTGCATTCCTGATGGCTAGGTGTC

Rbp1 GGCGACCACATGATCATCCGCACGC GCTTGCAGGTCACACCCTCTGCTCT

Scya2 TTGATATTACACTATTCCCTTCCAT CAATACCTTGGACTCTCAAACACAA

C3 ACCTCATGTGGGGCCTCTCCTCCGA GGTCAGTTGGGGCAGCCGAAAACCA

Vim GAGAAACTAACCTGGAGTCACTTCC AAACTGCAGTAAAGGCACTTGAAAG

Rat b-actin GGTCATCACTATCGGCAATGAGCGG CTTGATCTTCATGGTGCTAGGAGCC

Atp1b1 TGGACACTGAAATCCGCATTGAGTG ATTAAGTGTAGGTCCCATACGTATG

Atp2b1 TTATTTCAACCATTCCAACCAGCCG TCTGTTCAGGCCTCTAAACCACAAG

Atp12a GCTAGACTATGATTCCTCAGGCCTT AGTCACACATGTCACACATGAATAC

Ccl3 CCACATCCAGGGACTCTTCACTTGA GGTATAGACAAAGCATCCATTGTAC

Egr1 ATACTGCTCGACTGTAACTCTCACA GTCCAGCCATCGGCAAGGTGTGTCA

Hmox1 AGCTCTATCGTGCTCGCATGAACAC GTCTGGGATGAACTAGTGCTGATCT

Hspa1a GGCTGCTGGTGCACGATTCTTATCA ACTGTACACAGGGTGGCAGTGCTGA

Igf1 CCTACAAAGTCAGCTCGTTCCATCC CTCAAGCAGCAAAGGATCTTGCGGT

Il1b AGCTGAAAGCTCTCCACCTCAATGG TGGGATCCACACTCTCCAGCTGCAG

LOC57301 AAGCCGTGCGTCTGATATTTGCTAT TTTGCCAGCGGTATAGCCACATGCC

LOC60665 GTTTGCTTAACCTTAGCTCCAAGAA TGACCAGTGCAAGTGCATTCCGCTT

Mapk6 GTTGGCCGCAGTAACCAATGAGACT ACCATAGTGCCAAACGGGAAGAAGG

Nos2 TGAACCACCCGACTGAAGCACTTTG TCATCAAGGAATTATACACGGAAGG

Pdgfra TGACAAGCTGTATCACTGCCTTTGT ACATGAACACGGGTATCTGGAAGCC

Plat GATCGTGTCTCAACAGTGAAATACA TACACTGTGTACAAAGGGATATGCC

Rbp2 AGGAAACACCCTGGTGTGTGTGCAG GGGCTCTGTCAGGAACCAGGGAGTC

Sc4mol CAGAATATGCACATCCCTTGGAAAC ATGAAGTTCATGTGGTGGAAATCGT

Slc1a3 GACTTCTGGGAATTTGTTCAGTGGT TAATCTACAGCACACGTTCTTACAT

Sod2 AGAACAGTACATACTTGGTGTGAGC TTTCCTAACTGCAGTAGAACAGGAT

Tac1 GGTGACCTCCTCAGACAGAAGTAGA GCTAGGATACAAATAGAGTCAGATA

Tf TTAGCACTGTGCTGACTGCCCAGAA CTAGGAGTCGTGAGGTTGAACACTT

ApoE GCGATCGCGCCCAGGCTTTGAGTGA GGTAGCCACAGAGGCCTGTATCTTC

Atp1a1 GAGACAGCTCTTGCTGCTTTCCTGT TGCAGTGTGTGGCACAATGTTCCAC

C4a TTATCCCCGAGTGGAGTATGGCTTC TCTCCCTTGAGGTCACTGGTGACTC

Hspb1 GGATGAACATGGCTACATCTCTCGG TGAAGGCTTCTACTTGGCTCCAGAC

Igfbp5 ATCAGCCTCTACCAAGGAGCCTAGG TGTGAAGTGGTGAAAGTCTATACTC

Ncam1 TGACACAGCGGCTTCACCAGAGCAT AACGGAGTGTCCTAGTTTGGCTGTG

Nrp GTGACTGACTTCATGGTTCATCCTA ACGTAGCGATGCAAAAGCGGACATT

Stat3 ATGGATCTGACCTCGGAGTGTGCTA CATCACGGACTCAAGAGGTGCCCTC

Purinoceptors

Adora1 AGCCTGGATGTCTTCCTTGTATGGA TAGACATAGGGACCTCCTTGAGAAC

Adora2a GAGCTGCTGGCCTAGAAGTGGCACT CTAAGTGCATGGTAGCTTGGGCTAG

AGCTGCTGGCCTAGAAGTGGCACTT ACTAAGTGCATGGTAGCTTGGGCTA

Adora2b AGCTGCTCTTACTTACTGTGTGGAT TGAGAGAGAAGCAAACTCTAGTACC

Adora3 GAGCTTCTCTCATTCAATTCTGTGG CCTAGGGATCCTTCAACGCAGGTTC

P2rx1 GACATCATCCCTACTATGACTACTA CATGTTCTCCTGCAGGCCCAGAGTG

P2rx2 GAGGCTCGACCCCAAGTATGACCCT ATATGCTGGCCAAGTGTGTCCACCA

P2rx3 TAGTATCCAGAGACCTGGCCAGGCT ACACACTTACTGAAGGCCCAGAAGA

Guzman et al Inflamm Bowel Dis & Volume 12, Number 8, August 2006

768 * 2006 Lippincott Williams & Wilkins


received saline/vehicle (dimethylsulfoxide) starting 12 h afterinduction on day 1. To induce colitis in the distal colon, ratswere anesthetized with isofluorane (induced at 4%), andTNBS was delivered into the lumen of the colon through apolyethylene catheter inserted rectally 6 cm proximal to theanus.32 Animals were maintained in a control environmentfor 7 days after TNBS, saline, IB-MECA, or other treatments.On day 7, animals were killed, and tissue samples werecollected for analysis. The severity of colitis was assessed byweight loss and reduction in food intake over 7 days,histology, and clinical score. Each colon was removed andopened, and damage was scored as described by Petersonet al.33 All animal handling and histological, biochemical,and clinical assessments were performed as treatment-blindedassessments. Criteria of macroscopic scoring of colonicdamage were as follows: ulceration, normal appearance(score = 1); focal hyperemia, no ulcers (score = 2); ulcerationwithout hyperemia or bowel wall thickening (score = 3);ulceration with inflammation at 1 side (score = 4); 92 sites ofulceration and inflammation (score = 5); and major sites ofdamage extending 91 cm along colon (score = 6). When anarea of damage extended 92 cm along the length of the colon,the score was increased by 1 for each additional centimeterof involvement (score = 7Y11). The criteria for adhesionswere as follows: no adhesions (score = 0), minor adhesions(score = 1), and major adhesions (score = 2). For diarrhea, thescoring was as follows: no (score = 0) and yes (score = 2).The total score for each animal was calculated by adding upthe scores for each criterion. Whole-thickness tissue samplesor microdissected mucosa, submucosa, or longitudinalmuscleYmyenteric plexus8 were immediately frozen in liquidnitrogen for later analysis by microarray or SYBR greenPCR. Animal protocols were approved by the institutionalLaboratory Animal Care and Use Committee at Ohio StateUniversity.

RNA Collection and AmplificationFrozen tissue samples were collected and total cellular

RNA was isolated from each sample with the TRIzol RNAisolation reagent. The total RNAwas then cleaned with QiagenRNeasy Mini columns (catalog ID 74104). Samples were thenrun on RNA 6000 Nano Chips (Agilent, catalog ID 5065-4476)and analyzed for integrity with an Agilent Bioanalyzer 2100system and Degradometer software.34 cDNA was synthesizedwith the following T7-(dT)24 primer: 5-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-(dT)24-3. The dscDNA was cleaned with Qiagen cDNA spin columns(Affymetrix, catalog ID 900371). After cleanup, in vitro tran-scription labeling was performed with the Enzo BioArray Kit(Enzo, catalog ID 42655). The biotinylated cRNAwas purifiedwith Qiagen cRNA spin columns (Affymetrix, catalog ID900371). The cRNA was then fragmented and hybridized toRNU34 GeneChip Arrays (Affymetrix, catalog ID 900283),according to the manufacturer_s recommendations.8 The RNU34array contains 1322 transcripts representing, among others,cell surface receptors, cytokines, chemokines, growth factors,oxidant/antioxidant genes, neural genes, and oncogenes34 (http://www.affymetrix.com/support/technical/manual/expression_manual.affx, http://www.affymetrix.com/support/technical/datasheets/rgu34_arrays_datasheet.pdf).

Real-time SYBR Green PCRForward and reverse primers used in real-time SYBR

green PCR are listed in Table 1. SYBR green PCR wasperformed by taking 1 mg total RNA and reverse transcribingwith Superscript II (Invitrogen, Palo Alto, Calif) andoligo(dT) priming in 20 mL. Then, 0.5-mL reverse tran-scription reaction was used per PCR reaction. Real-time PCRwas performed in triplicate for each sample with the iQ-Cycler (Biorad, Hercules, Calif). Amplification products weredetected with SYBR Green Supermix (Biorad). Expression

Purinoceptors

P2rx4 GCAAGACGTTCTTCCACCCTATACA TCCATACGCTCACACTGTATAAGCC

P2rx5 GACATCCAGGAGACACTTAGCTTCG CAGCAAGAGCTGAACTGCACAAGTC

P2rx7 TAATGCCTCAGCCTAGTGCCTTTGG CTGCTGCTCCAGAGGGCTCAAGTTC

P2ry1 GCAGCTTCCACTGCCAAAGGCTAAT ATTGTAAAGCTTCAAGATCTGGCAG

P2ry2 AGCAGCTCAGTCAGGTGTCAGTTCA TCAGGTGGCGTTGCCTTAGATACGA

P2ry4 ATAGCTGTCTTGATCCAGTGCTCTA AGCAGCAGGGTTACAATCGATCTCC

P2ry6 TAGGTCCTGGAATAGCACTGCAAAT AAAGTCTTGGCAAATGGATGGGAAT

Glutathione peroxidase

Gpx1 AAGTACATCATTTGGTCCCCGGTGT ACAGCAGGGCTTCTATATCGGGTTC

Gpx2 AAGTAGTCCTCTTCAGAGCTCAGTG AGGGTTTAGGAAGATGCCATCATTC

For detailed descriptions of abbreviations, see Figure 1.

TABLE 1. (Continued)

Gene Forward Sequence Reverse Sequence










levels were estimated relative to the expression levels of thehousekeeping gene b-actin (ACTB). Differential geneexpression was calculated from mean values of expressionfor technical replicates and biological triplicates.

Microarray AnalysisGene expression levels are estimated from GeneChip

perfect match (PM) probe intensities by means of anenhanced version of the Li-Wong PM-only algorithm (http://genomebiology.com/2001/2/8/RESEARCH/0032). Resultswere expressed as fold-change value (baseline = 1) or log2fold-change values (baseline = 0). The fold-change is reportedas the ratio of TNBS to control or TNBS plus IB-MECA tocontrol. Results were then assessed with the PM-only modelusing software developed by Kornacker.34 The enhancedalgorithm (1) scales all PM and MM probe intensities tominimize between-array differences in the scaled MM probeintensity distributions; (2) applies between-array regressionanalysis to the scaled PM probe intensities to estimate PM-specific sensitivities, excluding any PM probes that fail toshow significantly positive sensitivities; (3) estimates geneexpression levels by regressing scaled PM probe intensities onestimated PM probe sensitivities within each probe set,excluding any PM probes that show significant nonmono-specificity; and (4) tests a probe-level general linear modelwithin each probe set to estimate the probability values forbetween-array differential gene expression35 (http://www.pnas.org/cgi/content/full/98/26/15044). The estimated proba-bility values can be several orders of magnitude lower than0.05 (P G 0.000001) as required by the Bonferroni correction,which applies when simultaneously testing thousands of genesfor significant differential expression. Coexpressed genes areidentified by means of a custom Bagglomerative partitioning[algorithm that successively partitions gene expression profilesinto a hierarchy of typical and exceptional subgroups (KTree).Graphs of selected KTree nodes are created using Eisen_sTreeView software, following the usual color scheme thatrepresents log-fold changes as red (positive), green (negative),or black (zero), with color saturation representing at least a2-fold change in gene expression, unless otherwise noted.

Electron Paramagnetic Resonance Spectroscopyand Spin Trapping in Ex Vivo Bowel Segments

Spin-trapping measurements of reactive oxygen species(ROS) generation were made with a Bruker ER 300 spectrom-eter. To detect O2

jI and OHI, 50 mmol/L of the spin trap

DMPO was used. The distal colon was removed, cut openalong the mesenteric border, and pinned out (with $20%stretch), and 1-cm square segments of whole-thickness speci-mens were incubated in Krebs buffer (in millimoles per liter:NaCl 120, KCl 6.0, MgCl2 1.2, NaH2PO4 1.35, NaHCO3

14.4, CaCl2 2.5, glucose 12.7) maintained for 20 min at 37-Cand pH 7.4 by gassing with 95% O2/5% CO2. The spin trapwas added, and ROS was monitored for 10 min by EPRI. Theincubation buffer was loaded into a quart flat cell and measuredat X-band in a TM-110 cavity. Additional control groupsincluded superoxide dismutase (SOD), catalase, and flavopro-tein to verify the nature of the signals observed. Quantificationof oxygen radical signals was performed by comparing thedouble integral of the observed signal with that of a knownconcentration of the 2,2,6,6,-tetramethylpiperidinoxy free rad-ical in solution.36 IB-MECA or other adenosinergic com-pounds were preincubated for 15 min with tissue beforeassessing their effect on ROS generation.

RESULTSA total of 30 rats were used in our studies. A large data set

is included in the analysis for gene dysregulation in TNBScolitis for bothmicroarray and SYBR green PCR. Data of >90%of genes showing significant dysregulation are included in theresults, but detailed descriptions and analyses of genes arelimited to those that are known to be implicated in human IBDand those that display remarkable upregulation or down-regulation with TNBS or sensitivity to IB-MECA treatment.

Microarray Analysis of the Effect of IB-MECA onColitis-related Strong Gene Dysregulation(Up- or Downregulation)

A rat neurobiochip (RNU34) with a gene microarray for$1322 gene transcripts representing 724 genes was usedsuccessfully to identify various genes that were upregulated ordownregulated by TNBS/ethanol-induced colitis in the rat. Onthe basis of our exclusion criteria for analysis, TNBS inductioncaused significant dysregulation in 198 genes representing 27%of the gene pool in RNU34, including neural, immune,proinflammatory, cytokine, chemokine, growth factor, anti-oxidant/oxidant, receptor, epithelial membrane ATPase trans-porter, voltage-gated channel, or other cell signaling genes.Figure 1 is a histogram representation of the changes in genesthat displayed significant colitis-related dysregulation in TNBScolitis. These genes were clustered according to their reported

FIGURE 1. Functional analysis of colitis-related gene dysregulation profiles in rat TNBS-induced colitis using the RNU34 neurobiochipwith 724 genes. Microarray analysis revealed that significant upregulation or downregulation occurred in 198 genes representing awide variety of different functional classes of genes that are implicated in IBD. Alterations occurred in (A) genes representing oxidant/antioxidant, cytokine, inflammation, chemokine, complement, platelet, protease, growth factor, heat shock protein, channels, andsignal transduction/kinase genes, as well as (B) membrane transporter, receptor, neural, early-onset/transcription factor/oncogene,retinoid metabolism, and adhesion/chemotaxis gene clusters. (The most dysregulated genes in colitis are displayed in Fig. 2.)







functions. Significant differences in gene expression betweenage-matched vehicle controls and TNBS colitis occurred witha value of P = 0.000001 or greater. Figures 1, 2, and 4represent genes with the most exceptional level of dysregula-tion or illustrate a unique pattern of dysregulation with orwithout treatment with IB-MECA.

Gene association profiles of microclusters were ob-tained with EASE analysis (http://david.niaid.nih.gov/david).The EASE-corrected probability value (SD Sidak) identifiedsignificant functional groups for colitis-related dysregulation.EASE identified 3 associated downregulated genes involvedin ATPase activity coupled to transmembrane movement ofions. For upregulation, response to biotic stimulus andresponse to stress revealed an association for 20 genes. Geneassociation profiles were also obtained with GeneMergeversion 1.1 software and revealed upregulated genes involvedin the inflammatory response and retinoid metabolism ordownregulated genes involved in amino acid metabolism,potassium ion transport, and sodium ion transport. Data aresummarized in Tables 2 and 3.

Figure 2 represents the most exceptional cluster of 49genes/probe sets and 39 different genes that make up 5.7% ofthe gene pool in RNU34 and illustrate 2 distinct modes ofcolitis-related dysregulation. The first mode produced strongcolitis-related downregulation in 5 genes that was nearlyprevented by IB-MECA treatment. They represent 3 epithe-lial membrane transporters/ATPases for H+/K+ transport(nongastric, a-polypeptide), Na+/K+ transport (a1), or Ca2+

transport (plasma membrane 1). The other 2 genes are sterol-C4-methyl oxidaseY like and mitogen-activated protein kinase6 (MapK6; note that log2-fold changes are shown). Thesecond mode generally produced strong colitis-related upre-gulation in the remaining 34 genes ranging from 2- to 12.5-fold upregulation (a ratio of 1.0 means no difference fromcontrol; net fold changes are described with 1.0 subtracted forcontrol). IB-MECA nearly prevented the strong upregulationin 990% of the 34 genes. Upregulation of solute carrier family1, member 3 (Slc1a3) was prevented by IB-MECA. Otherupregulated genes included retinol binding proteins 1 and 2(rbp1, 2), but rbp1 was upregulated by 10-fold compared to 1.6-fold for rbp2. Four chemokine-related genes that weresignificantly upregulated include MIP-1a receptor gene,chemokine (C-X-C motif) ligand 2, chemokine (C-C motif)ligand 3, and C-X-C chemokine LIX. Upregulated cytokines

were IL-1b and small inducible cytokine A2. Two probe setsgave 3- and 5.5-fold changes in the small inducible cytokine;for IL-1b, different probes yielded 2.5- and 6-fold changes.Probe sets for complement component 3 (C3) revealed 5.2-and 4.0-fold upregulation; C4 was upregulated by 2.4-fold inTNBS colitis. Antioxidant enzymes/related genes that wereupregulated by TNBS colitis are SOD2, inducible nitric oxidesynthase (iNOS/NOS2), heme oxygenase-1 (Hmox1), andtransferrin. Two different probe sets for Hmox1 gave the same4-fold upregulation. Neural/related genes that were upregu-lated included neural cell adhesion molecule (Ncam1), activityand neurotransmitter-induced early gene protein 4 (Ania4),tachykinin gene encoding SP, neurokinin A, neurokinin K,neuropeptide g, neuropilin, and vimentin. The cluster includes4 growth factors and related genes that were upregulated,namely insulin-like growth factor binding protein 5 (Igfbp5),early growth response 1 (Egr1, 2.7- or 3.7-fold upregulation by2 different probes), insulin-like growth factor 1 (Igf1), andplatelet-derived growth factor receptor a (Pdgfra). Other genesare platelet selectin, C/EBPb (3.5- or 5-fold upregulation), theacute-phase reactant apolipoprotein E (ApoE), the benzodia-zepine receptor (Bzrp, peripheral benzodiazepine receptor),and several heat shock protein genes (HSP 70, 70-kDa protein1A, 27-kDa protein 1).

Real-time SYBR Green PCR Validation ofSelected Transcripts to Assess the Effect ofIB-MECA on Colitis-related Gene Dysregulation

In general, real-time SYBR green PCR analysis (Fig. 3)confirmed the relative changes in upregulation or down-regulation of 990% of the genes depicted in Figure 2 withgene chip microarray analysis. These genes represent aheterogeneous set of genes with distinct and unrelatedfunctions (see below). Several points deserve further consid-eration. In the cluster in Figure 3 in which we have analyzedall known genes by PCR, Igfbp5 is missing because the PCRconditions did not work. MapK6 did not show significantregulation, and the Na+/K+ transporting ATPase (a1)(Atp1a1) showed upregulation on TNBS treatment (in contrastto the chip results). Sc4mol,sterol-C4-methyl oxidaseY likegene, and 3 ATPase transporters were downregulated by TNBSand showed the same tendency as the chip results with IB-MECA treatment, although Sc4mol dysregulation was notsignificantly affected by IB-MECA (unlike chip data). All of the

FIGURE 2. Oral IB-MECA administration prevents colitis-induced alterations in gene expression patterns in themost exceptional clusters ofgenes. Two modes of colitis-related gene dysregulation profiles are depicted. A, Hierarchical gene cluster analysis of genes 1 through 7associated with the strongest colitis-related downregulation; B, histogram representation of the fold changes in genes 1 through 7;C, hierarchical gene cluster representation of genes 9 through 49 associated with the strongest colitis-related upregulation; and D,histogram representation of the effect of IB-MECA in preventing gene upregulation in genes 9 through 49. Thirty-nine different genes arerepresented. n = 3 animals/group; P G 0.00001 for differences. Pseudocolor scale: green denotes log2-fold downregulation comparedwith age-matched controls; red, Q2-fold upregulation compared with control; and black, no difference between groups. Lanes 1 and 2depict cluster data according to calculated mean ratios of 1, IBMECA plus TNBS to control, and 2, TNBS to control.







other genes were strongly upregulated, typically in the range of5- to 1000-fold, and oral IB-MECA treatment reduced or nearlyprevented dysregulation; receptor genes were upregulated by2.5- to 20-fold. The absolute magnitudes of the change in geneexpression varied tremendously as expected between genemicroarray and SYBR green PCR; PCR is much more sensitivethan microarray. Notably, upregulation of rbp1 exceeded rbp2as occurred with gene chip analysis. The highest upregulationoccurred in IL-1b, C3, chemokines, ApoE, benzodiazepinereceptor, platelet selectin, insulin-like growth factor 1, andsmall inducible cytokine (ranging from 20- to 500-foldupregulation); NOS2 (i.e., iNOS) was the most upregulatedgene at 91000-fold. The tendency for C3 upregulation to begreater than C4 upregulation is consistent with gene chip data.Note that IB-MECA treatment caused significant reduction orattenuation or nearly prevented gene dysregulation at valuesbetween P G 10j6 and 10j11 in most cases. TNBS causedsignificant dysregulation in all genes at values in the range ofP G 0.002 to P G 10j9 (with noted exceptions earlier). Overall,34 of 37 genes were confirmed by microarray and SYBRgreen PCR for both induction with TNBS and treatment withoral IB-MECA, yielding a 92% confidence in the microarrayanalysis for those genes.

In contrast to the previous 39 genes, SYBR green PCRanalysis of purine genes (Fig. 3) revealed some differences ingene dysregulation from those detected with gene chip micro-array analysis. It is notable that purine genes did not cluster inmicroarray analysis indicative for variable and heterogeneousdysregulation profiles. Therefore, PCR revealed upregulation ofA1R gene mRNA expression of $6-fold that was not affectedby IB-MECA treatment. Adenosine A3R expression wasupregulated by 2-fold but was not significantly altered by IB-MECA. A2aRs were undetectable by SYBR green PCR using2 different sets of primers. A2bR was not affected by TNBSinduction or IB-MECA treatment. In contrast, a 4-fold P2X4R

mRNA upregulation in TNBS colitis was nearly prevented byIB-MECA. Significant levels of upregulation of P2X1, P2X4,P2X7, P2Y2, and P2Y6 purinoceptor mRNA levels rangingfrom 5- to 20-fold upregulation was either prevented or mostlyblocked by IB-MECA treatment. It is noteworthy that P2YRupregulation was much greater than that of P2XRs.

Gene chip analysis with adenosine receptor probes forA1, A2a, A2b, and A3Rs on the chip revealed that only A2aand A2bR genes showed statistically significant differentialexpression in TNBS colitis. Therefore, downregulation ofA2bR gene mRNA expression was significantly blocked byIB-MECA treatment but not prevented. Upregulation of A2awas prevented by IB-MECA. The associated probabilityvalues are G0.000000001. Among all of the P2 purinergicreceptors on the chip (including P2X, P2Y, P2Y4, P2Y1, P2X1,P2X2, P2X3, P2X4, P2X7), only P2X4 receptors (P2rx4)showed statistically significant differential expression. The log2ratios showed approximately equal upregulation in inflamed(0.43) and treated (0.36) animals; therefore, IB-MECA did notreveal any blockade of P2X4R gene mRNA expression. Bothlog2 ratios are highly significant (P G 0.000000001). It isinteresting that the benzodiazepine receptor (Bzrp) gene, unlikethe purinoceptor genes, was shown to be regulated in a similarfashion by both techniques. However, Bzrp was stronglyupregulated by gene chip analysis, unlike purine receptor genes.

Microarray Analysis of Genes ModestlyDysregulated by TNBS

Figures 4 and 5 depict additional clusters of genes withmoderate colitis-related downregulation or downregulation andupregulation, respectively. In general, all genes were signifi-cantly dysregulated by colitis (P G 0.000000001), but oral IB-MECA treatment blocked dysregulation in some of the genes.In the remainder, IB-MECA had a modest effect ondysregulated genes. In some genes, IB-MECA converted a

FIGURE 3. SYBR green PCR validation analysis of genes represented in the most exceptional hierarchical clusters of colitis-relatedgene dysregulation profiles. AYC, SYBR green PCR analysis illustrating the alteration of gene expression between normal colon, TNBStreatment, and IB-MECA plus TNBS treatment in individual samples. A, ACTB, housekeeping gene b-actin. B, Effect of IB-MECA onAtp1b1, Na+/K+ transporting b1 polypeptide dysregulation. C, Partial inhibition of the upregulation of the chemokine ligand 2(C-X-C motif, Ccl3) by IB-MECA. Graphs show relative fluorescence units (y axis) vs cycle numbers (x axis); orange lines showcycle threshold used for calculation of gene expression; measurements are averages from data in 3 samples/animal for eachtreatment group and triplicate PCR determinations for each animal. Graphs are shown for b-actin (A1), a housekeeping geneused as loading control; Atp1b1 (A2), a gene that is repressed by TNBS treatment (green, blue, and red lines) and derepressedby IB-MECA treatment (blue, purple, and pink lines); and Ccl3 (C), a gene induced by TNBS treatment (green, purple, andpink lines) and partially repressed by IB-MECA treatment (blue, purple, and pink lines). D, Real-time SYBR green PCR histogramanalysis of the effect of IB-MECA on a selected set of 39 genes that showed the greatest colitis-related dysregulation. SYBRgreen PCR analysis also revealed purine receptor gene dysregulation in TNBS-induced colitis. A1R or A3R gene dysregulation isnot significantly altered by oral IB-MECA treatment. A2bR was not affected by TNBS induction or IB-MECA treatment.Significant levels of upregulation of P2X1, P2X4, P2X7, P2Y2, and P2Y6 purinoceptor mRNA levels ranging from 5- to 20-foldupregulation were either prevented or mostly blocked by IB-MECA treatment. Relative expression in normal colon is defined as1. Fold changes are calculated as ratios of values of TNBS treated to control and TNBS plus IB-MECA to control. Probabilityvalues of significant differences between TNBS treatment and IB-MECA are shown on the histogram for each gene.




downregulation to an upregulation or the reverse. For instance,in the gene clusters depicted in Fig. 4, TNBS caused moderatedownregulation of IL-18 and cathepsin E, but IB-MECAprevented and strongly upregulated these genes. CD48 antigenwas upregulated by IB-MECA. Additional examples are

shown in Fig. 5. In addition, 19 genes that displayed colitis-related downregulation were moderately or not affected byoral IB-MECA treatment. Examples of these genes areneurodap 1, MAP kinase 3, glutamine synthase 1, glutamatedehydrogenase 1, chloride channel 2, and PKC z (Fig. 4).

FIGURE 4. Gene cluster of colitis-dysregulated genes with moderate sensitivity to IB-MECA treatment. For instance, 19 of the genesthat displayed colitis-related downregulation were moderately or not affected by oral IB-MECA treatment. Pseudocolor scale: greendenotes log2-fold downregulation compared with age-matched controls; red, Q2-fold upregulation compared with control; andblack, no difference between groups. Lanes 1 and 2 depict cluster data according to calculated mean ratios of 1, IBMECA plus TNBSto control, and 2, TNBS to control.




TABLE 2. EASE Analysis of Exceptional Microclusters

Gene Category List Hits EASE Score SD SidakGene Identifiers(LocusLinks)

AffymetrixProbesets Descriptive Gene Name

Downregulation

ATPase activity/coupled totransmembranemovement of ions

3 9.26E-04 4.17E-02 24211 M74494_g_at ATPase, Na+/K+transporting,a1; ATPase H+/K+transporting, nongastric, apolypeptide; ATPase Na+/K+

transporting b1 polypeptide

25650 M90398_at

171028 rc_AI230614_s_at

Upregulation

Response tobiotic stimulus

11 4.11E-05 8.19E-03 24232 L23088_at Platelet selectin; early growthresponse 1; IL-1b,chemokine (C-C motif)ligand 3; complementcomponent 4a; chemokine(C-X-C motif) ligand 2; CXCchemokine LIX;apolipoprotein E, smallinducible cytokine A2 (2�);complement component 3;SOD2**

24233 M18416_at

24330 M98820_g_at

24494 U22414_at

24770 U42719_at

24787 U45965_at

25542 U90448_at

25651 X04979_at

25728 X17053cds_s_at

60665 X17053mRNA_s_at

114105 X52477_at

Y00497_s_at

Response to stress 9 2.28E-04 4.44E-02 24232 J02722cds_at Heme oxygenase 1; IL-1-b;glutamate dehydrogenase 1;chemokine (C-C motif)ligand 3, complementcomponent 4a; chemokine(C-C motif) ligand 2;apolipoprotein E; smallinducible cytokine A2 (2�);complement component 3;SOD2

24233 M98820_g_at

24451 rc_AI179610_at

24494 U22414_at

24770 U42719_at

24787 U45965_at

25542 X04979_at

25728 X17053cds_s_at

114105 X17053mRNA_s_at

X52477_at

Y00497_s_at




Effect of Oral IB-MECA Treatment on ColonicHistopathology, Clinical Score, and Progressionof Inflammation

TNBS-induced colitis leads to macroscopic and micro-scopic signs of inflammation (see Fig. 6AY I) compared withage-matched vehicle/control animals (Fig. 6AYC). TNBS/ethanol induction caused variable regions of necrosis inmucosa (arrows) with regions of hemorrhage in both mucosaand submucosa (arrowhead) and extensive infiltration ofinflammatory cells (Fig. 6D). Injury to the mucosa leads toulceration, villus atrophy, and dense transmural infiltration ofinflammatory cells (Fig. 6E). Major adhesions were visible,and ulcerations occurred at Q1 sites in association withhyperemia and bowel wall thickening. Oral administration ofIB-MECA (1.5 mg/kg weight b.i.d. for 6 days) nearlyprevented all visible signs of inflammation and colitis, andthe gut wall appeared normal except for a few inflammatorycells still present in the submucosa and circular muscle layers(Fig. 6GY I) (n = 4 animals/group).

IB-MECA treatment improved the clinical score,appetite, and weight gain in TNBS-treated rats (Fig. 6JYM).IB-MECA nearly prevented the colitis because the clinicalscore remained modestly but significantly higher than age-matched controls (Fig. 6J), and food intake fully recoveredafter 72 to 48 h from TNBS induction (Fig. 6K). About 12 hafter the first oral dose of IB-MECA (e.g., 1 day after TNBSinduction), TNBS-induced rats receiving IB-MECA wereeating very little food compared with controls, and theirfood intake was the same as that of TNBS-induced animals.Within 48 h, IB-MECA/TNBSY treated rats began to eat asmuch as age-matched controls and significantly more thanTNBS rats (Fig. 6K). As a result, total food intake in IB-MECAY treated rats remained significantly lower than

controls (Fig. 6L), but the weight loss profile during the7-day period from induction indicates that IB-MECA treat-ment was effective in preventing the 15% body weight losscaused by TNBS induction (Fig. 6M).

Effect of Acute IB-MECA Treatment on FreeRadical Production in TNBS-Inflamed Colon

The antioxidant role of A3Rs is a target for its beneficialeffect that was explored directly by EPRI of superoxideanions. EPRI analysis using DMPO to spin trap hydroxyl freeradicals (R-OH) showed that TNBS-induced colitis leads to aseveral-fold increase in R-OH (Fig. 7A, B); R-OH is low innormal colon. Acute exposure to IB-MECA (0.1Y10 mmol/L) for 20 min significantly reduces the R-OH by 30% to 70%in whole-thickness gut (shown) or mucosa (not shown) ininflamed gut (Figs. 7A, B). The source of the R-OH issuperoxide anions because SOD abolishes it. Exogenousapplication of adenosine (100 mmol/L) or a cocktail ofEHNA plus NBTI (1 mmol/L) that preserves endogenousadenosine released caused a similar inhibition of R-OH.

SYBR Green PCR Analysis of Free RadicalEnzyme/Pathways

Two important antioxidant enzymes in the gut mucosawith overlapping activities are glutathione peroxidase 1 and 2(GPx1 and GPx2). SYBR green PCR revealed a 7-foldupregulation (Fig. 3). Gene chip microarray analysis and/SYBR green PCR revealed strong colitis-related upregulationof the various antioxidant/oxidant enzymes SOD1, SOD2,iNOS, and heme oxygenase (Hmox1) and iron-metabolicpathways (transferrin, transferrin receptor, iron-responsiveelement binding protein). Oral IB-MECA treatment blockedor prevented such upregulation in these genes.

TABLE 3. Gene Merge Analysis

Gene Category Study Faction e-Score Affymetrix Probe Sets Descriptive Gene Name

Downregulation

Amino acid metabolism 2/32 0.032 rc_AI044610_s_at Dopa decarboxylase

rc_AI179613_at Glutamate dehydrogenase 1

Potassium ion transport 2/4 0.014 M74494_g-at ATPase, H+/K+ transporting, a1

rc_AI230614_s_at ATPase, Na+/K+transporting, b1 polypeptide

Sodium ion transport 2/4 0.007 M74494_g_at ATPase, H+/K+ transporting, a1

rc_AI230614_s_at ATPase, Na+/K+ transporting, b1 polypeptide

Upregulation

Inflammatory response 3/28 0.001 U22414_at Chemokine (C-C motif) ligand 3

M98820_s_at IL-1b

M29866_s_at Complement component 3

Retinoid metabolism 2/28 0.025 M19257_at Retinol-binding protein 1

M13949_at Retinol-binding protein 2




Differential Dysregulation in Neural andNonneural Layers

SOD1 and SOD2 are important for EPRI analysis ofR-OH because we are trapping extracellular superoxide withDMPO that is dismutated by SOD2 to H2O2. We show that

SOD was differentially upregulated in neural and nonneuralmuscle layers: it was upregulated 0.52-fold in longitudinalmuscleYmyenteric plexus, 0.21-fold in mucosa, and 2.16-fold in submucosa (all changes, P G 0.000001; n = 3 rats).Gene chip analysis indicated that inducible nitric oxide

FIGURE 5. Another gene cluster moderately dysregulated by TNBS colitis with strong sensitivity to oral IB-MECA treatment. Someupregulated genes converted to downregulation after oral IB-MECA treatment. Pseudocolor scale: green denotes log2-folddownregulation compared with age-matched controls; red, Q2-fold upregulation compared with control; and black, no differencebetween groups. Lanes 1 and 2 depict cluster data according to calculated mean ratios of 1, IBMECA plus TNBS to control, and 2,TNBS to control.




FIGURE 6. IB-MECA prevents cell and tissueinjury in TNBS-induced colitis in rats. Colonsections from age-matched control (AYC),TNBS-treated (DYF), and IB-MECAY treatedTNBS rats (GY I). AYC, Normal histologicalprofile. D, TNBS administration caused areas ofnecrosis in mucosa (M; arrows) with evidentregions of hemorrhage in both mucosa andsubmucosa (SM; arrowhead) and infiltration ofinflammatory cells (curved, hollow arrow).E, Injured mucosa resulting in ulceration (ar-rows), atrophy of villi (arrowheads), edema,fibrosis, and inflammatory cell infiltration (hol-low arrows). F, Atrophy of villi (arrowhead) anddense infiltration of all of the layers (arrows).GY I, Oral administration of IB-MECA (1.5 mg/kg b.i.d. for 6 days) nearly prevented the colitis.A few immune cells more than in age-matchedcontrols are still visible in submucosa (arrows,GY I) and circular muscle (CM) layer (arrow-head, G). Hematoxylin and eosin (H & E)staining. Scale bar = 50 mm. J, Clinical scoringconfirms effects of IB-MECA in protectingagainst histopathological damage (K). Foodintake is nearly normal in IB-MECAY treated rats(P G 0.05). L, Total food intake (7 days) is muchgreater in IB-MECAY treated TNBS animals thanin TNBS colitis animals. M, IB-MECA treatmentprevented weight loss that occurs as a result ofTNBS colitis.




synthase mRNA is elevated in colitis by 2-fold in mucosa,0.30-fold in longitudinal muscleYmysenteric plexus, and 4-fold in SMP (P G 0.000001), suggesting a putative role fornitric oxide in ONOO- and nitrotyrosine-induced transmuraldamage. The increase in iNOS mRNAwas similar to previousestimates from PCR.37 NO derived from iNOS mediates, inpart, inflammation-induced epithelial transport dysfunction.37

Genes Induced by NF-kBGenes that are reported to be regulated by the transcrip-

tion factor are summarized in Figure 8. TNBS colitis causedupregulation in these genes, including chemokines, comple-ment, cytokines, adhesion molecules, heme oxygenase, andSOD2. IB-MECA blocked such dysregulation.

Neural Gene DysregulationNeural genes that are dysregulated by TNBS colitis are

illustrated in Figure 9. IB-MECA prevented such dysregulation.

DISCUSSIONTo the best of our knowledge, this is the first study to

examine the prophylactic effect of ADOA3Rs in experimentalcolitis using high-density oligonucleotide microarray analysisof gene expression profiles. Overall, 5.4% of the gene pool inthe RNU34 gene chip represent the most exceptional genecluster, revealing 2 modes of colitis-related gene dysregulation(upregulation or downregulation). These findings, validated byreal-time quantitative SYBR green PCR for both induction ofcolitis and treatment with oral IB-MECA, are consistent withhistopathology and clinical observations of the severity ofinflammation and tissue injury.

Certain types of genes were strongly upregulated andothers were downregulated by colitis. A distinguishingfeature of a subset of genes (rbp1, IL-1b,C3, chemokines,iNOS, Igf1, platelet selectin, Hmox1, Bzrp, and the smallinducible cytokine) is the strong 50- to 1000-fold geneupregulation by SYBR green PCR. Despite obvious differ-ences in etiology between chemical (TNBS) induction of ex-perimental colitis and immune-based IBD in humans, aremarkable number of different functional gene families thatare dysregulated in rat colitis resemble those in humanIBD.28 This suggests that the TNBS colitis model is asuitable IBD model for molecular gene dysregulation studieson chemokine/related genes (i.e., MIP-1a receptor gene,chemokine C-C motif), cytokine (IL-1b), complement (C3,C4a), growth factor (Igfbp5, Egf1, Pdgfra), receptor(i.e., Bzrp, purine receptors), HSP (HSP70, HSP27), retinoidmetabolism (rbp1), oxidant/antioxidant pathway (SOD2,Hmox1, GPX1, iNOS, ferritin), neural and remodeling genes,and mucosal transporters (i.e., Atp2b1, Ca2+transporter).Some upregulated genes are involved in the inflammatoryresponse or are a response to a biotic stimulus or stress(Tables 2 and 3).

Our study on gene dysregulation provides convincingdata that an ADOA3R agonist can prevent development ofchronic inflammation and most gene dysregulation, histopa-thology, clinical signs of damage, and weight loss in TNBS-induced colitis. Three days after TNBS induction, animalstreated with IB-MECA were eating as much as TNBS-induced animals, and they did not lose any weight, in contrastto TNBS animals. This indirectly suggests that IB-MECAalso may block the development of the acute phase and

FIGURE 7. Acute ex vivo IB-MECA/adeno-sine treatment reduces superoxide anion/hydroxyl radical production. A, Riboflavingenerates hydroxyl radicals. A small butdetectable hydroxyl radical response occursin normal gut. In inflamed gut, there is a 3-fold increase in free radical production. IB-MECA (10 mmol/L) treatment (20 min ofacute exposure) reduces the response.Adenosine suppresses the response. A cock-tail of EHNA plus NBTI that blocks break-down and reuptake of eADO, respectively,mimics the effects of adenosine or IB-MECA.The hydroxyl radicals are derived fromsuperoxide anions because SOD (1 mmol/L)blocks them in inflamed gut. EPRI is used todetermine hydroxyl free radical productionusing DMPO spin trapping. Tissues were

preincubated for 20 min in Krebs_ oxygenated buffer at 37-C and then preincubated with vehicle or drugs for 20 min. EPRI of DMPOspin trapping was done for 10 min. B, Hydroxyl radical generation. Pooled data show that an increase in radical production occurs incolitis and is reduced $50% by IB-MECA (P G 0.01). n = 3 to 4 tissues/condition. A.U. indicates arbitrary units.




its progression to chronic inflammation. At the dose given,IB-MECA could not prevent all gene dysregulation inducedby TNBS colitis by day 7. In other studies, limited markeranalysis suggested that IB-MECA may be beneficial in DDSor IL-10 spontaneous models of colitis or other models. AnADO kinase inhibitor also was shown to be of benefit inDDS colitis.16,17 The potential of ADO receptor agonistssuch as IB-MECA in the treatment of active chronicinflammation once established remains unknown in exper-imental colitis or IBD. IB-MECA was more effective inprotecting against rat TNBS-induced colitis than in prevent-ing murine colitis in previous studies on DDS or IL-10knockout mice. Differences in susceptibility to IB-MECAmay reflect differences in mechanism of induction, speciesdifferences, and the primary role of oxidative stress inthe injury and colitis induced by TNBS in contrast to DDSor IL-10 knockout mice.

An important finding is that a cluster of genes that werestrongly downregulated by colitis were Na+, amino acid, andK+ and Ca2+ transporters (i.e., H+/K+ transporter, nongastric;Na+/K+ transporter, a1, a3 probes; Ca2+ transporter, plasma

membrane 1; Na+/K+ transporter, b1 polypeptide) that arealso indicative of the extent of mucosal barrier injury and lossof epithelial functional integrity. These genes can be targetedfor therapy in studies focused on mucosal barrier function,integrity, and transport in the rat TNBS colitis model. Thesetransporters are important in the normal function andmaintenance of electrolyte and water balance across theepithelial membranes. IB-MECA almost prevented down-regulation in these transport mechanisms and protectedagainst mucosal and transmural cell injury, reflected in thelack of histopathology and normal clinical scores/profiles. Itshould be noted, however, that downregulation of the Na+/K+

transporter 1a was only modestly affected by IB-MECAtreatment. Upregulation of Na+/K+ transporter also is pre-vented by IB-MECA.

Our study did not delineate which receptor is involvedin the protective action of IB-MECA. However, the thera-peutic oral dose used in rats (1.5 mg/kg) is likely sufficient toactivate other ADORs that have a higher affinity thanADOA3R.5,6 All 4 ADOR subtypes are expressed (differ-entially) in immune cells and neutrophils38,39 and in the

FIGURE 8. Summary of TNBS colitisY inducedgenes that were reported to be regulated bythe redox-sensitive transcription factor NF-kB.Upregulation of redox-sensitive genes isblocked by oral IB-MECA treatment.




intestinal tract in neural and nonneural elements,8 and theirexpression is similar in rats and humans (F.L.C., unpublishedobservations). However, ADOA3Rs are a more likely targetfor the effects of oral IB-MECA administration thanADOA1Rs, although A2aRs also have potential. Activationof A1Rs by IB-MECA would not be beneficial becausechronic A1R activation could lead to rapid desensitization/internalization of A1Rs and disinhibition in the entericnervous system, adding to neural dysfunction.8,40 In addition,neutrophils contain A1Rs, A2Rs, and A3Rs that produceopposite effects. A1Rs cause chemotaxis and phagocy-tosis,38,39 and these proinflammatory actions limit the benefitsof targeting A1Rs in IBD and certainly argue against theinvolvement of A1Rs in the benefits of oral IB-MECA.Interestingly, ADOA1R/mRNAs are upregulated by a varietyof conditions, including glucocorticoid treatment, ethanolwithdrawal, and oxidative stress, rabbit ileitis,30 and TNBScolitis. In contrast, activation of the upregulated ADOA3R orA2aR activation by eADO exerts anti-inflammatory effects39,41

in human neutrophils, inhibits free radicals, stimulates anti-oxidant enzymes, and/or has putative neuroprotective effects42

that may contribute to the effects of IB-MECA in vivo.Neuronal plasticity in the enteric nervous system is a

prominent feature of the rat colitis model.32 Hyperexcitabilityin gut sensory afferent AH neurons and facilitation ofsynaptic transmission occurs during chronic TNBS colitis.32

Activity-dependent signaling pathways lead to neuronalinduction of early genes encoding various proteins, including

transcription factors involved in neuronal functional plasti-city.43,44 Dysregulation (upregulation or downregulation)occurred in a variety of neural genes encoding neuropeptides(Tac1, for substance P, neurokinin A, neuropeptide K, andneuropeptide g; Penk-rs-pre-pro-enkephalin), neural recep-tors (Bzrp, galanin, GABA-A, bradykinin-B, neuropilin,purine receptors), neural kinases (Ania4, Ania2), neural celladhesion (Ncam1), and neuronal development (Nnat, neuro-natin; Nrp, neuropilin). Inflammatory mediators are known toaffect all components of gut neural reflexes and to influencereceptor expression and neuronal phenotypic changes thatmay affect release of cytokines such as IL-1. IL-1b inhibitsacetylcholine release and increases iNOS expression/activity.In TNBS colitis, the activity of myenteric neurons in theinflamed gut may be under the ongoing influence of IL-1bthat is strongly upregulated and is implicated in colonicdysmotility in experimental colitis.45 Oral IB-MECA waseffective against neural gene dysregulation in TNBS colitis.The neuroprotective effects of ADOA3R agonists have beendemonstrated after chronic administration46 that may be, inpart, indirect by inhibiting the production of the proinflam-matory cytokines MIP-1a, TNF-a, or IFN-g in macrophagesor monocytes, known to express ADOA3Rs.5,6,47 Theseinflammatory cytokines play a key role in IBD,1 and IB-MECA was effective against the production of these cyto-kines in other colitis models.16 This study showed that geneinduction of various cytokine/chemokine/proinflammatorygenes was blocked by IB-MECA.

FIGURE 9. Neural genes that are dysregulated by TNBS colitis are sensitive to oral IB-MECA treatment. Neuromedin andbenzodiezepine receptors are the most dysregulated neural genes. Their dysregulation is highly sensitive to oral IB-MECA treatment.




Our study revealed novel information on dysregulationof 19 receptor genes. We can predict that because all of thesegenes are important in the neurophysiology of the gut, theirdysregulation would impair normal gut function. Significantpurine receptor gene upregulation occurs in purinergicADOA1, A3, A2a, P2X1, P2X3, P2X7, P2Y2, and P2Y6receptor genes. Such widespread dysregulation would resultin significant disruption in normal gut functions, motility,secretion, neural reflexes, and coordination reflexes knownto be modulated by such receptors.7 Y9 These receptors arepresent in the human digestive tract and/or in various rodentspecies studied to date7 (F.L.C., unpublished observations).Except for putative A3Rs targeted by oral IB-MECA, it re-mains unknown whether other purine receptors serve anybeneficial role in colitis. Expression of ADOA1R or ADOA3Ror receptor mRNA has been shown to be sensitive to intestinalinflammation in a rabbit ileitis/CD model,30 and ADOA1Rdownregulation leads to disruption of neurotransmission.31 IB-MECA partially blocked or prevented dysregulation ofselected P2X and P2Y receptor subtypes without affectingADORA1R or ADOA3R or downregulated P2Y1R, P2Y4R,or P2X2R gene products. P2Y nucleotide receptor genedysregulation was much more sensitive to IB-MECA treat-ment than P2X receptor genes. Therefore, there is selectiveinfluence of the orally active ADOA3R agonist IB-MECA onpreventing dysregulation of subsets of purine genes in TNBS-induced colitis. This study revealed novel complex puri-neYpurine receptor interactions in chronic gut inflammationthat were not previously reported in other tissues.

Strong upregulation of the peripheral benzodiazepinereceptor (Bzrp) gene occurs in the gut of colitis animals. Bzrpis involved in oxidative metabolism and steroidogenesis48

and is localized to surface epithelium, absorptive goblet cells,and enteric neurons of rat colon.49 Upregulation of Bzrpoccurs in human neurodegenerative/inflammatory diseasesand monocytes.50,51 Expression of Bzrp also is subject toregulation by chronic stress such as TNBS colitis (i.e., EASEanalysis). A protective role is suggested for Bzrp in TNBScolitis that may be related, in part, to its anti-inflammatoryactions.52 Its expression and functional importance in IBDremain unknown.

TNBS may be converted to ROS,53 which can con-tribute to injury and inflammation of the colon. In the inflamedmucosa, the main sources of free radicals are activatedmucosal leukocytes (neutrophils and macrophages), xanthineoxidase during ischemia/reperfusion, and arachidonic acidmetabolites. Oxidative stress may lead to tissue injury by ROSinteracting with proteins, lipids, carbohydrates, and DNA andstimulation of neutrophil chemotaxis, 5-lipoxygenase, andphospholipase A2.37 Finally, ROS can activate transcriptionfactors such as NF-kB, c-fos, c-myc, and c-jun37 involved incytokine release, hypertrophic signaling, and cell growth.Superoxide generation may play a role in IBD.1 In rat TNBS

colitis, SOD2 was strongly elevated in the colonic tissues in afailed attempt to protect the gut against ROS injury, whereasa decrease in SOD was reported in patients with IBD,54

which would increase tissue injury. Species differences maybe the result of the longer duration of the oxidative stressperiod in IBD, leading to further compromises in antioxidantdefense mechanisms. SOD2 is induced by IL1-b (and TNF-a)and serves a protective function against oxidative damage.IBD in humans are immune-mediated diseases, with dysre-gulated T cell activity1 leading to inflammation and ROSproduction. An SOD mimetic reduced ICAM-1 upregulation,expression of P-selectin, and lipid peroxidation products inTNBS colitis;55 these genes were upregulated in our study.ICAM-1 and P-selectin were shown to be expressed inendothelial and epithelial cells and neutrophils in the ratTNBS colon.55 LeukocyteYendothelial interactions and neu-trophil migration from the vessel require P-selectin, b2

integrins, and ICAM-1.56 IB-MECA was effective in sup-pressing these oxidative stress and chemotactic mechanismsin TNBS colitis.

Thirty-three upregulated genes are targets for redox-sensitive transcriptional regulation through NF-kB (i.e., chemo-kines, cytokines, complement, SOD2, heme oxygenase, iNOS,NCAM 1, ICAM 1, selectin, Bzrp, IL-1b, Jun B proto-oncogene, ADOA1R), and others for heat shock factor 1(HSF1). Virtually all of these genes were upregulated bycolitis, and their sensitivity to oral IB-MECA is consistent withthe hypothesis that activation of ADOA3R prevents colitis bysuppressing free radical production and redox-sensitive tran-scriptional regulation through NF-kB, HSF1, and other factorsinvolved in the induction of a wide variety of genes includingproinflammatory genes. Antisense oligonucleotide to NF-kBp65 treatment was shown to ameliorate inflammation evenafter induction of colitis.1

eADO release into the extracellular environment underconditions of oxidative stress is believed to play a cytoprotec-tive role.21 Neurons/nerve terminals, neutrophils, and endothe-lial cells have been reported to release high levels of ADO atsites of inflammation, infection, and metabolic stress. Duringinflammation, eADO may become high enough to exertimmunomodulatory, neuroprotective, immunosuppressive,and antioxidant effects through putative low-affinityADOA3Rs. With respect to the last effect, free radicals areimplicated in IBD and TNBS colitis, and activation ofADOA3Rs is known to stimulate the antioxidant defensemechanisms.29 SOD or GPX knockout mice are used to studyfree radicals in colitis.57,58 SODs convert superoxide radicals(O2

j) to H2O2 and then to water by GPX or catalase. SOD2is inducible by oxidative insults and cytokines. It is notsurprising that in TNBS colitis SOD2 is tremendouslyupregulated, given the 3-fold elevation in free radical gen-eration in the inflamed gut. An increase in peroxidation ofmembrane lipids also suggests the involvement of oxidative




stressYmediated mechanisms of cell/tissue injury in both ratTNBS colitis and IBD.30 The transmural damage and genedysregulation seen in rat TNBS colitis also provide anotherlink with human CD/IBD. The ability of IB-MECA or ADO tosuppress free radicals in ex vivo inflamed colon indicates thatADO has direct antioxidant effects in inflamed gut. This maybe one mechanism by which it exerts protection. Theinhibitory effect of IB-MECA on GPX1, GPX2, Hmox1,SOD2, iNOS, and iron-handling genes supports its antioxidantmode of action in TNBS colitis and mimics the effects ofeADO at putative ADOA3Rs.

ROS regulates the antioxidant genes heme oxygenase(Hmox1) and GSH.59 The expression of the small heat shockprotein (HSP) Hmox1 was strongly upregulated in theinflamed intestine, and it may afford neuroprotection.60

HSP27, which was elevated 5-fold by TNBS, also has beensuggested to have neuroprotective effects60 and anti-inflammatory effects in pancreatitis.61 Autoantibodies againstvarious HSPs exist in autoimmune diseases and IBD.13

Ferritin, the major intracellular iron storage protein, isresponsive to intracellular oxidative stress and ROS gener-ated during gut inflammation. Related proteins, transferrin,the transferrin receptor, and iron-responsive element-bindingprotein, are downregulated in TNBS colitis.62 Thesegenes are restored by oral IB-MECA treatment, indicatingthat IB-MECA was able to restore normal gut function.Overall, IB-MECA seems to be highly effective in protectingagainst the deleterious effects of proinflammatory mediators,including free radicals, restoring oxidant/antioxidant balancein experimental colitis, and potentially having some efficacyin IBD.

Lamina propria mononuclear cells express higher levelsof IL-1b in IBD (both UC and CD). IL-1b release issustained throughout the clinical course, with a prominentneutrophil infiltrate in IBD patients (CD or UC).1 In TNBScolitis, IL-1b was one of the genes with the strongestupregulation. IL-1 was a sensitive indicator of mucosal in-flammation, and its values correlated well with myeloperox-idase activity in TNBS/ethanol-induced colitis in the rat.1

Caspase-1 is an IL-1bYconverting enzyme (ICE) that cleavesand activates IL-18 and IL-1b. ICE expression and release ofthese cytokines contribute to gut inflammation. Manipulationsthat interfere with ICE expression or effect protect againstinflammation.63 The strong upregulation of the IL-1b gene isprobably in part a consequence of ICE downregulation in TNBScolitis. In IBD, there appears to be a disturbed balance betweenproinflammatory and anti-inflammatory cytokines. Higher levelsof proinflammatory cytokines occur in IBD (IL-1, IL-8, TNF-a,IL-6) from macrophages, lymphocytes, and polymorpho-nuclear leukocytes. Their synthesis is induced by the activationof NF-kB.64 IL-1 is produced primarily by macrophages, and itsexpression is increased in inflammatory lesions of patients withIBD. An imbalance between IL-1 and IL-1R antagonist occurs

in inflamed mucosa.64 Data indicate that complement activationalso contributes to the pathogenesis of IBD. Serum concen-trations of C3 were higher in CD than in UC, which can be usedin the differential diagnosis to distinguish patients with CD andUC.65 Our findings with upregulation in C3 and C4 in TNBScolitis are consistent with IBD. Overall, the ability of IB-MECAto block IL-1b, ICE, IL-18, C3, and C4a gene dysregulationcontributes to its anti-inflammatory effects and protectionagainst tissue injury.

The expression of colonic chemokines is nonselectivelyupregulated in IBD. Chemokines promote leukocyte migrationto areas of inflammation and initiate cell activation events.Macrophage inflammatory proteins 1a and 1b (MIP-1a, MIP-1b) are b chemokines that attract monocytes and T lympho-cytes and are involved in their activation.66 Epithelialcells, macrophages, T lymphocytes, neutrophils, and plasmacells in IBD mucosa produce chemokines. It has beensuggested that the Bdegree of local inflammation and damagein IBD depends on local chemokine expression in IBDtissues.[66 Recent findings indicated that activated plateletsmay contribute to mucosal inflammation and pathogenesisof UC and CD by mediating the increase in the CD40-dependent expression of adhesion molecules (ICAM-1,VCAM- 1) and chemokine production by intestinal micro-vascular endothelial cells, which is followed by T celladhesion to these cells.67 Abnormalities in platelet numbersand function occur in IBD patients,68 and activated plateletsinduce neutrophil activation and degranulation and dischargechemoattractant molecules such as MIP-1a (this study) thatparticipate in leukocyte recruitment.68 In activated IBDplatelets, there is upregulation of VCAM-1 and ICAM-1, 2key leukocyte adhesion molecules (reviewed above). IB-MECA inhibits upregulation of these genes, an effect thatmay contribute to its protective effect. Various chemokine orplatelet genes that are strongly upregulated in TNBS colitis arehighly sensitive to IB-MECA treatment and are suitabletherapeutic/screening targets.

MAP kinase 6 was greatly downregulated compared toother kinases or other downregulated genes in TNBS colitis.The mitogen-activated protein kinase pathways regulate mostcellular processes, including defense mechanisms such asstress reactions and inflammation in IBD. Several MAPkinase inhibitors reached clinical trials with promise in CD.69

They are key elements in the regulation of all stages ofinflammation and are activated by G-protein coupled recep-tors or cytokine receptors (i.e., IL-1R, TNF-aR). Activationof MAP kinases can lead to subsequent activation of nuclearkinases and transcriptional regulation (AP-1, cAMP responseelement binding protein). All MAP kinase signaling cascadesare key players in the initiation and propagation ofinflammation, and considerable cross-talk occurs with otherinflammatory pathways such as NF-kB.69 Downregulation ofMapK6 in TNBS colitis is likely to represent a host-adoptive




response during colitis in a failed attempt to limit or preventthe inflammatory response. IB-MECA protected against gutinjury and prevented downregulation of MapK6.

The iNOS/NO signaling pathway also could play a keyrole in the pathogenesis of IBD. In rat colitis, iNOS was themost upregulated gene product (91000-fold by PCR).Increased iNOS mRNA or iNOS protein occurs in IBD andin human epithelial cells exposed to cytokines.70,71 In con-trast, the inducible Hmox1 is believed to act as a protectiveantioxidant mechanism. Hmox1 converts heme into biliverdin,CO, free ferrous iron, and eventually to bilirubin and bilepigments. These actions provide a potent antioxidant systemthat limits both ROS- and lipid peroxidationYassociateddamage of tissues and anticomplement effects.72,73 Hmox1 isessential for iron homeostasis. Heme and NO donors arepotent inducers of Hmox1 in the human intestinal epithelialcells,70 suggesting potential interactions in IBD. In rat colitis,the remarkable upregulation of Hmox1 may result from theupregulated iNOS expression and elevation of NO to activatethe pathway. Ferritin sequestration of ferrous iron reduces itsavailability to catalyze lipid peroxidation and oxygen radicalformation.74 Iron also regulates NO synthase gene expres-sion.60 The overexpression of Hmox1 in TNBS colitis maybe involved in immunomodulation, host defense, antioxidant,and neuroprotective mechanisms.60

CONCLUSIONSValidated gene microarray analysis established patterns

of gene dysregulation in rat TNBS colitis that are largelyconsistent with IBD. We conclude that our validated high-density oligonucleotide microarray analysis is a powerfultechnique for molecular gene dysregulation studies to assessthe beneficial effects of purine-based or other drugs in exper-imental colitis. The ADOA3R is a new potential therapeutictarget for IBD that selectively modulates certain purinereceptor genes altered by colitis. These included upregulations(e20-fold in mRNA levels) in ADOA1R, ADOA3R, P2X1R,P2X4R, P2X7R, P2Y2R, and P2Y6R receptor mRNA withobvious implications for altering gut reflexes in colitis. Otherpurinergic receptors such as P2X2R, P2Y1R, and P2Y4R thatwere downregulated were insensitive to the ADOA3R agonist.Oral administration of an ADOA3R agonist is effective inprotecting the intestine from experimental colitis. The benefi-cial effects of ADOA3R activation could potentially involveanti-inflammatory, antioxidant, or neuroprotective effects (thisstudy), as well as modulation of serotonin release fromenterochromaffin cells.40 Inhibitory ADOA3R exists on enter-ochromaffin cells, and eADO activates them to inhibit 5-HTrelease. The ADOA3R agonist IB-MECA is being tested inclinical trials for other inflammatory disorders and is apparentlywithout toxicity. A clinical implication of our study is thepotential for testing ADOA3R agonists or other purine-based

ADO analogue drugs (i.e., ADOA2aR agonists) in theprophylactic treatment of postoperative recurrence of CD afterileocolonic resection shown to be responsive to another purine,6-mercaptopurine.75

REFERENCES1. Fedorak RN, Madesen KL. Naturally occurring and experimental models

of inflammatory bowel disease. In: Kirsner JB, ed. Inflammatory BowelDisease. 5th ed. Philadelphia: WB Saunders; 2000:113Y143.

2. Cronstein BN, Naime D, Ostad E. The anti-inflammatory mechanism ofmethotrexate: increased adenosine release at inflamed sites diminishedleukocyte accumulation in an in vivo model of inflammation. J ClinInvest. 1993;92:2675Y2682.

3. Marak GE Jr, de Kozak Y, Faure JP, et al. Pharmacologic modulation ofacute ocular inflammation. I. Adenosine. Ophthalmic Res. 1988;20:220Y226.

4. Schrier DJ, Lesch ME, Wright CD, et al. The anti-inflammatory effectsof adenosine receptor agonists on the carrageenan-induced pleuralinflammatory response in rats. J Immunol. 1990;145:1874Y1879.

5. Hasko G, Szabo C. Regulation of cytokine and chemokine production bytransmitters and co-transmitters of the autonomic nervous system.Biochem Pharmacol. 1998;56:1079Y1087.

6. Hasko G, Kuhel DG, Chen JF, et al. Adenosine inhibits IL-12 and TNF-alpha production via adenosine A2a receptor-dependent and independentmechanisms. FASEB J. 2000;14:2065Y2074.

7. Cooke HJ, Wunderlich J, Christofi FL. BThe force be with you:[ ATP ingut mechanosensory transduction. News Physiol Sci. 2003;18:43Y49.

8. Christofi FL. Unlocking mysteries of gut sensory transmission: isadenosine the key? News Physiol Sci. 2001;16:201Y207.

9. Cooke HJ, Javed N, Christofi FL. Enteric neural reflexes and secretion.In: Goetzl E, Blennerhassett M, Bienestock J, eds. Autonomic NervousSystem, Autonomic Neuroimmunology: Vol 16. Reading, UK: HarwoodAcademic Publishers; 2003:35Y59.

10. Cronstein BN, Van de Strouwe M, Druska L, et al. Nonsteroidal anti-inflammatory agents inhibit stimulated neutrophil adhesion to endothe-lium: adenosine dependent and independent mechanisms. Inflammation.1994;18:323Y335.

11. Bouma M, Gvan den Wildenberg FA, Buurman WA. The anti-inflammatory potential of adenosine in ischemia-reperfusion injury:established and putative beneficial actions of a retaliatory metabolite.Shock. 1997;8:313Y320.

12. Eigler A, Sinha B, Hartmann G, et al. Taming TNF: strategies to restrainthis proinflammatory cytokine. Immunol Today. 1997;18:487Y492.

13. Tishler M, Shoenfeld Y. Anti-heat shock protein antibodies in rheumaticand autoimmune diseases. Semin Arthritis Rheumatism. 1996;26:558Y563.

14. Heidland UE, Heintzen MP, Michel CJ, et al. Effect of adjunctiveintracoronary adenosine on myocardial ischemia, hemodynamic functionand left ventricular performance during percutaneous transluminalcoronary angioplasty: clinical access to ischemic preconditioning? CoronArtery Dis. 2000;11:421Y428.

15. Mahaffey KW, Puma JA, Barbagelata NA, et al. Adenosine as an adjunctto thrombolytic therapy for acute myocardial infarction: results of amulticenter, randomized, placebo-controlled trial: the Acute MyocardialInfarction Study of Adenosine (AMISTAD) trial. J Am Coll Cardiol.1999;34:1711Y1720.

16. Mabley J, Soriano F, Pacher P, et al. The adenosine A3 receptor agonist,N6-(3-iodobenyl)-adenosine-5-N-methyluronamide, is protective in twomurine models of colitis. Eur J Pharmacol. 2003;466:323Y329.

17. Siegmund B, Rieder F, Albrich S, et al. Adenosine kinase inhibitorGP515 improves experimental colitis in mice. J Pharmacol Exp Ther.2001;296:99Y105.

18. Kowaluk EA, Mikusa J, Wismer CT, et al. ABT-702 (4-amino-5-(3-bromophenyl)-7-(6-morpholino-pyridin-3-yl)pyrido[2,3-d]pyrimidine), anovel orally effective adenosine kinase inhibitor with analgesic and anti-inflammatory properties, II: in vivo characterization in the rat. J PharmacolExp Ther. 2000;295:1165Y1174.

19. Borman S, Washington C, Washington EN. A3 receptors: compoundsthat activate or inhibit adenosine A3 receptors are being studied for




potential therapeutic use in heart disease and cancer. Sci Technol. 2001;79:37Y40.

20. Dragunow M, Faull RL. Neuroprotective effects of adenosine. TrendsPharmacol Sci. 1988;9:193Y194.

21. Cronstein B, Naime D, Firestein G. The anti-inflammatory effects of anadenosine kinase inhibitor are mediated by adenosine. Arthritis Rheum.1995;38:1040Y1045.

22. Ahonen A, Kyosola K, Penttila O. Enterochromaffin cells in macrophagesin ulcerative colitis and irritable colon. Ann Clin Res. 1976;8:1Y7.

23. Apasov S, Koshiba M, Redegeld F, et al. Role of extracellular ATPand P1 and P2 classes of purinergic receptors in T-cell developmentand cytotoxic T lymphocyte effector functions. Immunol Rev. 1995;146:5Y19.

24. Broussas M, Cornillet-Lefebvre P, Potron G, et al. Inhibition offMLP-triggered respiratory burst of human monocytes by adenosine:involvement of A3 adenosine receptor. J Leukoc Biol. 1999;66:495Y501.

25. Costenla AR, de Mendonca A, Ribeiro JA. Adenosine modulates synapticplasticity in hippocampal slices from aged rats. Brain Res. 1999;851:228Y234.

26. Linden J. Cloned adenosine A3 receptors: pharmacological properties,species differences and receptor functions. Trends Pharmacol Sci. 1994;15:298Y306.

27. Christofi FL, Zhang H, Yu JG, et al. Differential gene expression ofadenosine A1, A2a, A2b, and A3 receptors in the human entericnervous system. J Comp Neurol. 2001;439:46Y64.

28. Lawrance IC, Fiocchi C, Chakravarti S. Ulcerative colitis and Crohn_sdisease: distinctive gene expression profiles and novel susceptibilitycandidate genes. Hum Mol Genet. 2001;10:445Y456.

29. Maffirwar S, Dhanraj D, Somani S, et al. Adenosine acts as anendogenous activator of the cellular antioxidant defense system.Biochem Biophys Res Commun. 1994;201:508Y515.

30. Sundaram U, Hassanain H, Suntres Z, et al. Rabbit chronic ileitis leadsto up-regulation of adenosine A1/A3 gene products, oxidative stress, andimmune modulation. Biochem Pharmacol. 2003;65:1529Y1538.

31. De Man JG, Seerden TC, De Winter BY, et al. Alteration of thepurinergic modulation of enteric neurotransmission in the mouse ileumduring chronic intestinal inflammation. Br J Pharmacol. 2003;139:172Y184.

32. Linden D, Sharkey K, Mawe G. Enhanced excitability of myenteric AHneurones in the inflamed guinea-pig distal colon. J Physiol. 2003;47:589Y601.

33. Peterson TC, Cleary CE, Shaw AM, et al. Therapeutic role for bismuthcompounds in TNBS-induced colitis in the rat. Dig Dis Sci. 2000;45:466Y473.

34. Auer H, Lyianarachchi S, Newsom D, et al. Chipping away at the chipbias: RNA degradation in microarray analysis. Nat Genet. 2003;35:292Y293.

35. Huang Y, Prasad M, Lemon W, et al. Gene expression in papillarythyroid carcinoma reveals highly consistent profiles. Proc Natl Acad SciU S A. 2001;98:15044Y15049.

36. Souza HP, Liu X, Samouilov A, et al. Quantitation of superoxidegeneration and substrate utilization by vascular NAD(P)H oxidase. Am JPhysiol Heart Circ Physiol. 2002;282:H466YH474.

37. Pavlick KP, Laroux FS, Fuseler J, et al. Role of reactive metabolites ofoxygen and nitrogen in inflammatory bowel disease. Free Radic BiolMed. 2002;33:311Y322.

38. Gessi S, Varani K, Merighi S, et al. A3 adenosine receptors in humanneutrophils and promyelocytic HL60 cells: a pharmacological andbiochemical study. Mol Pharmacol. 2002;61:415Y424.

39. Sullivan GW, Linden J. Role of A2A adenosine receptors in inflamma-tion. Drug Dev Res. 1998;45:103Y112.

40. Christofi FL, Kim M, Wunderlich JE, et al. Endogenous adenosinedifferentially modulates 5-hydroxytryptamine release from a humanenterochromaffin cell model. Gastroenterology. 2004;127:188Y202.

41. Sajjadi FG, Takabayashi K, Foster AC, et al. Inhibition of TNF-alphaexpression by adenosine: role of A3 adenosine receptors. J Immunol.1996;156:3435Y3442.

42. Muller CE. Medicinal chemistry of adenosine A3 receptor ligands. CurrTopics Med Chem. 2003;3:445Y462.

43. Walton M, Henderson C, Mason-Parker S, et al. Immediate early genetranscription and synaptic modulation. J Neurosci Res. 1999;58:96Y106.

44. Kandel ER. The molecular biology of memory storage: a dialoguebetween genes and synapses. Science. 2001;294:1030Y1038.

45. Morteau O, More J, Pons L, et al. Platelet-activating factor andinterleukin 1 are involved in colonic dysmotility in experimental colitisin rats. Gastroenterology. 1993;104:47Y56.

46. Von Lubitz DKJE. Adenosine A3 receptor and brain: a culprit, a hero, ormerely yet another receptor? Ann N Y Acad Sci. 1997;825:49Y67.

47. McWhinney CD, Dudley MW, Bowlin TL, et al.Activation of adenosineA3 receptors on macrophages inhibits tumor necrosis factor-alpha. Eur JPharmacol. 1996;310:209Y216.

48. Gavish M, Bachman I, Shoukrun R, et al. Enigma of the peripheralbenzodiazepine receptor. Pharmacol Rev. 1999;51:630Y646.

49. Bribes E, Carriere D, Goubet C, et al. Immunohistochemical assessmentof the peripheral benzodiazepine receptor in human tissues. J HistochemCytochem. 2004;52:19Y28.

50. Faggioli P, Giani L, Chianese R, et al. Increase in peripheralbenzodiazepine receptors on monocytes in fibromyalgia. Rheumatology(Oxford). 2004;43:1224Y1225.

51. Wilms H, Claasen J, Rohl C, et al. Involvement of benzodiazepinereceptors in neuroinflammatory and neurodegenerative diseases: evi-dence from activated microglial cells in vitro. Neurobiol Dis. 2003;14:417Y424.

52. Lazzarini R, Malucelli BE, Palermo-Neto J. Reduction of acuteinflammation in rats by diazepam: role of peripheral benzodiazepinereceptors and corticosterone. Immunopharmacol Immunotoxicol. 2001;23:253Y265.

53. Grisham MB, Volkmer C, Tso P, et al. Metabolism of trinitrobenzenesulfonic acid by the rat colon produces reactive oxygen species.Gastroenterology. 1991;101:540Y547.

54. Verspaget HW, Mulder TP, Van der Sluys Veer A, et al. Reactive oxygenmetabolites and colitis: a disturbed balance between damage and protection: aselective review. Scand J Gastroenterol Suppl. 1991;188:44Y51.

55. Cuzzocrea S, Mazzon E, Dugo L, et al. Protective effects of M40403, asuperoxide dismutase mimetic, in a rodent model of colitis. Eur JPharmacol. 2001;432:79Y89.

56. Butcher EC. Specificity of leukocyte-endothelial interactions anddiapedesis: physiologic and therapeutic implications of an activedecision process. Res Immunol. 1993;144:695Y698.

57. Krieglstein CF, Cerwinka WH, Laroux FS, et al. Regulation of murineintestinal inflammation by reactive metabolites of oxygen and nitrogen:divergent roles of superoxide and nitric oxide. J Exp Med. 2001;194:1207Y1218.

58. Esworthy RS Aranda R Martin MG, et al. Mice with combineddisruption of GPX1and GPX2 genes have colitis. Am J PhysiolGastrointest Liver Physiol. 2001;281:G848YG855.

59. Alayash AI. Oxidative mechanisms of hemoglobin-based blood sub-stitutes. Artif Cells Blood Substit Immobil Biotechnol. 2001;29:415Y425.

60. Panahian N, Yoshiura M, Maines MD. Overexpression of hemeoxygenase-1 is neuroprotective in a model of permanent middle cerebralartery occlusion in transgenic mice. J Neurochem. 1999;72:1187Y1203.

61. Kubisch C, Dimagno MJ, Tietz AB, et al. Overexpression of heat shockprotein Hsp27 protects against cerulein-induced pancreatitis. Gastro-enterology. 2004;127:275Y286.

62. Roeser HP. In: Jacobs A,WorwoodM, eds. Iron Metabolism in Biochemistryand Medicine, Vol 2, New York: Academic Press; 1980:605Y640.

63. Siegmund B. Interleukin-1beta converting enzyme (caspase-1) inintestinal inflammation. Biochem Pharmacol. 2002;64:1Y8.

64. Rogler G, Andus T. Cytokines in inflammatory bowel disease. World JSurg. 1998;22:382Y389.

65. Bene L, Fust G, Fekete B, et al. High normal serum levels of C3 and C1inhibitor, two acute-phase proteins belonging to the complement system,occur more frequently in patients with Crohn_s disease than ulcerativecolitis. Dig Dis Sci. 2003;48:1186Y1192.

66. Banks C, Bateman A, Payne R, et al. Chemokine expression in IBD:mucosal chemokine expression is unselectively increased in bothulcerative colitis and Crohn_s disease. J Pathol. 2002;199:28Y35.




67. Danese S, Monte LA, Sturm C, et al. Platelets trigger a CD40-dependentinflammatory response in the microvasculature of inflammatory boweldisease patients. Gastroenterology. 2003;124:1249Y1264.

68. Danese S, de la Motte C, Fiocchi C. Platelets in inflammatory boweldisease: clinical, pathogenic, and therapeutic implications. Am J Gastro-enterol. 2004;99:938Y945.

69. Waetzig GH, Schreiber S. Review article: mitogen-activated proteinkinases in chronic intestinal inflammation: targeting ancient pathways totreat modern diseases. Aliment Pharmacol Ther. 2003;18:17Y32.

70. Cavicchi M, Gibbs L, Whittle BJR. Inhibition of inducible nitric oxidesynthase in the human intestinal epithelial cell line, DLD-1, by theinducers of heme oxygenase 1, bismuth salts, heme, and nitric oxidedonors. Gut. 2000;47:771Y778.

71. Salzman A, Denenberg AG, Ueta I, et al. Induction and activity of nitricoxide synthase in cultured human intestinal epithelial monolayers. Am JPhysiol. 1996;270:G565YG573.

72. Maines MD. The heme oxygenase system: a regulator of secondmessenger gases. Annu Rev Pharmacol Toxicol. 1997;37:517Y554.

73. Applegate LA, Luscher P, Tyrrell RM. Induction of heme oxygenase: ageneral response to oxidant stress in cultured mammalian cells. CancerRes. 1991;51:974Y978.

74. Aust SD. Ferritin as a source of iron and protection from iron-inducedtoxicities. Toxicol Lett. 1995;82-83:941Y944.

75. Hanauer SB, Korelitz BI, Rutgeerts P, et al. Postoperative maintenanceof Crohn_s disease remission with 6-mercaptopurine, mesalamine, orplacebo: a 2 year trial. Gastroenterology. 2004;127:723Y729.




ADOA3R as a Therapeutic Target in Experimental Colitis: Proof by Validated High-density Oligonucleotide Microarray Analysis

Documents