Pentoxifylline Prevents Loss of PP2A Phosphatase Activity andRecruitment of Histone Acetyltransferases to ProinflammatoryGenes in Acute Pancreatitis
Juan Sandoval, Javier Escobar, Javier Pereda, Natalia Sacilotto, Jose Luis Rodriguez,Luis Sabater, Luis Aparisi, Luis Franco, Gerardo Lopez-Rodas, and Juan SastreDepartment of Biochemistry and Molecular Biology (J.San., N.S., J.L.R., L.F., G.L-R.), Department of Physiology (J.E., J.P.,J.Sas.), School of Medicine, University of Valencia, Spain; Department of Surgery (L.S.) and Laboratory of Pancreatic Function(L.A.), University Clinic Hospital, Valencia, Spain
Received June 10, 2009; accepted August 7, 2009
ABSTRACTMitogen-activated protein kinases (MAPKs) are considered ma-jor signal transducers early during the development of acutepancreatitis. Pentoxifylline is a phosphodiesterase inhibitor withmarked anti-inflammatory properties through blockade of ex-tracellular signal regulated kinase (ERK) phosphorylation andtumor necrosis factor � production. Our aim was to elucidatethe mechanism of action of pentoxifylline as an anti-inflamma-tory agent in acute pancreatitis. Necrotizing pancreatitis in-duced by taurocholate in rats and taurocholate-treated AR42Jacinar cells were studied. Phosphorylation of ERK and ERKkinase (MEK1/2), as well as PP2A, PP2B, and PP2C serine/threonine phosphatase activities, up-regulation of proinflam-matory genes (by reverse transcription-polymerase chain reac-tion and chromatin immunoprecipitation), and recruitment oftranscription factors and histone acetyltransferases/deacety-lases to promoters of proinflammatory genes (egr-1, atf-3, inos,icam, il-6, and tnf-�) were determined in the pancreas during
pancreatitis. Pentoxifylline did not reduce MEK1/2 phosphory-lation but prevented the marked loss of serine/threonine phos-phatase PP2A activity induced by taurocholate in vivo withoutaffecting PP2B and PP2C activities. The rapid loss in PP2Aactivity induced by taurocholate in acinar cells was due to adecrease in cAMP levels that was prevented by pentoxifylline.Pentoxifylline also reduced the induction of early (egr-1, atf-3)responsive genes and abrogated the up-regulation of late (inos,icam, il-6, tnf-�) responsive genes and recruitment of transcrip-tion factors (nuclear factor �B and C/EBP�) and histone acetyl-transferases to their gene promoters during pancreatitis. Inconclusion, the beneficial effects of pentoxifylline—and pre-sumably of other phosphodiesterase inhibitors—in this diseaseseem to be mediated by abrogating the loss of cAMP levels andPP2A activity as well as chromatin-modifying complexes veryearly during acute pancreatitis.
Acute pancreatitis begins with a local inflammation of thepancreatic tissue that, in the severe forms, leads to a sys-temic inflammatory response, and eventually 20 to 30% ofcases result in death due to multiple organ failure (Bhatia etal., 2001). The early events and signaling mechanisms thatoccur in the pancreas, and in particular, in acinar cells, areconsidered of special pathophysiological relevance because
they may be translated into long-term inflammatory re-sponses that would determine the development of pancreati-tis (Ji et al., 2003). The early inflammatory cascade in acutepancreatitis should be ascribed not only to the inflammatoryinfiltrate but also to acinar cells, which may behave as in-flammatory cells (de Dios et al., 2005). Indeed, they respondto, produce, and release cytokines, chemokines, and adhesionmolecules (Grady et al., 1997; Gukovskaya et al., 1997; Za-ninovic et al., 2000; Ramudo et al., 2005; de Dios et al., 2006).
Numerous inflammatory mediators such as activated pan-creatic enzymes, cytokines, chemokines, free radicals, Ca2�,platelet-activating factor, adenosine, and neurogenic factorshave been involved in the pathogenesis of acute pancreatitis
This work was supported by the Spanish Ministry of Science and Innovation[Grants SAF2009-09500, SAF2006--06963, CSD-2007-00020] (to J.Sas.); andthe Spanish Ministry of Science and Innovation [Grants BFU2004-03616,BFU2007-63120, CSD2006-49] (to G.L.-R.).
Article, publication date, and citation information can be found athttp://jpet.aspetjournals.org.
(Denham et al., 1997; Grady et al., 1997; Bhatia et al., 1998;Satoh et al., 2000; Pereda et al., 2006). At present, elucida-tion of the signaling pathways involved in this inflammatorynetwork is underway. Protein kinases, and especially mito-gen-activated protein kinases (MAPKs), are considered ma-jor signal transducers early during the development of pan-creatitis (Hofken et al., 2000). Activation of MAPK requiresphosphorylation of threonine and tyrosine residues by up-stream MAPK kinases, and strength and duration of MAPKactivation may determine their biological effects (Marshall,1995). Inhibition of p38 decreased pancreatic and pulmonaryinjury in severe acute pancreatitis in rats (Yang et al., 1999).
Pentoxifylline [3,7-dimethyl-1-(5-oxohexyl)purine-2,6-di-one] exhibits marked anti-inflammatory properties mediatedmainly by inhibition of tumor necrosis factor � (TNF-�) pro-duction and by prevention of ERK phosphorylation andNF-�B activation (Schandene et al., 1992; Haddad et al.,2002a,b; Pereda et al., 2004). Accordingly, in knockout micedeficient in TNF-� receptors, the rate of mortality due tonecrotizing acute pancreatitis decreased because the sys-temic response was restrained (Denham et al., 1997). Simul-taneous blockade of the three major MAPKs—that is, extra-cellular signal regulated kinase (ERK), c-jun N-terminalkinase (JNK), and p38 kinase—by pentoxifylline and oxy-purinol greatly reduced the local and systemic inflammatoryresponse in necrotizing acute pancreatitis and decreased themortality rate (Pereda et al., 2004). Pentoxifylline also showsbeneficial effects in edematous pancreatitis (Gomez-Cam-bronero et al., 2000; Pereda et al., 2004; de Campos et al.,2008). However, the precise mechanism of action of pen-toxifylline responsible for these positive effects is not yetelucidated.
Pentoxifylline is a well known phosphodiesterase inhibitor(Meskini et al., 1994) and this effect may be involved in itsanti-inflammatory action (Haddad et al., 2002b). The benefi-cial effects of phosphodiesterase inhibitors are associatedwith reduced secretion of proinflammatory cytokines, adher-ence of leukocytes to endothelium, and inhibition of leukocyteactivation (Klemm et al., 1995; Sekut et al., 1995). Rolipram,a phosphodiesterase IV-specific inhibitor, inhibits lypopo-lysaccharide-induced IL-6 production (Haddad et al., 2002a)and ameliorates cerulein-induced acute pancreatitis (Sato etal., 2006). cAMP might affect MAPK activation by acting onphosphatase activities, such as serine threonine phosphatasePP2A (Feschenko et al., 2002). Indeed, inactivation of MAPKby dephosphorylation is critical and may be triggered byserine threonine (Ser/Thr) phosphatases.
The aim of the present study was to investigate themechanism of action of pentoxifylline as anti-inflamma-tory agent and potential therapy in acute pancreatitis, inparticular, to explain its effects on MAPK activation andTNF-� production.
Materials and MethodsAnimals. Young male Wistar rats were used in the experiments.
Animals were cared for and handled in accordance with the Decla-ration of Helsinki and the European regulations (Council Directive86/609/EEC), and the studies were approved by the Research Com-mittee of the University of Valencia. Animals were anesthetized withintraperitoneal administration of ketamine (80 mg/kg b.wt.) andacepromacine (2.5 mg/kg b.wt.) before sacrifice.
Experimental Model of Acute Pancreatitis and Treatmentwith Pentoxifylline. Acute necrotizing pancreatitis was induced inrats by retrograde infusion into the biliopancreatic duct of sodiumtaurocholate (3.5%) (Sigma-Aldrich, St. Louis, MO) (Pereda et al.,2004). Lipase activity was measured and histological studies wereperformed to confirm the appropriate induction of necrotizing pan-creatitis. Lipase activity in plasma was 47 � 19 IU/liter (n � 6) incontrols and it increased markedly at 6 h after pancreatitis induction(1280 � 230 IU/liter; n � 6) as well as at 6 h in rats with pancreatitistreated with pentoxifylline (945 � 310 IU/liter; n � 6). Rats wereanesthetized as previously mentioned before induction of pancreati-tis and before sacrifice at 0, 30 min, 1, 3, and 6 h after the infusionof taurocholate. Where indicated, rats were treated with pentoxifyl-line (12 mg/kg b.wt.) just after taurocholate treatment. Pentoxifyl-line (12 mg/kg b.wt.) was infused into the femoral vein (0.066 ml/min) for 30 min, as in Pereda et al. (2004).
Culture of AR42J Acinar Cells. The AR42J cell line, derivedfrom an exocrine rat pancreatic tumor (CRL 1492; American TypeCulture Collection, Manassas, VA), was grown in Dulbecco’s modi-fied Eagle’s medium (Invitrogen, Carlsbad, CA) containing 25 mMglucose, 100 �g/ml penicillin, 100 �g/ml streptomycin, and 25 �g/mlfungizone, supplemented with 10% fetal bovine serum.
Measurement of Serine/Threonine Phosphatase Activities.Pancreata were removed from control rats and rats with acute pan-creatitis at 1 h after induction. Pancreatic tissue was homogenized in50 mM Tris-HCl, pH 7.4, containing 150 mM NaCl, 5 mM dithio-threitol, 2 mM EDTA, 0.1% Triton X-100, and the following proteaseinhibitor cocktail (Sigma-Aldrich): 1.5 mM 4-(2-aminoethyl)benzene-sulfonyl fluoride, 1,2 �M aprotinin, 60 �M bestatin, 21 �M E-64, 30�M leupeptin, 15 �M pepstatin A. Samples were centrifuged for 1 hat 100,000g at 4°C. Supernatants were collected and subjected toexclusion chromatography on Sephadex G-25 columns to eliminatefree phosphates. Eluates were used to measure the activities ofPP2A, PP2B (calcineurin), and PP2C serine/threonine phosphatasesusing a serine/threonine phosphopeptide and the Serine/ThreoninePhosphatase Assay System provided by Promega and following themanufacturer instructions.
Measurement of cAMP Levels. cAMP levels were measured inAR42J cells using the cyclic AMP EIK kit (Cayman Chemical Co.,Ann Arbor, MI) following the instructions of the manufacturer.
Western Blotting. Pancreatic specimens were frozen at �80°Cuntil they were homogenized in extraction buffer (100 mg/ml) on ice.The extraction buffer contained 10 mM Tris-HCl, pH 7.5, 0.25 Msucrose, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate,50 mM sodium fluoride, 100 �M sodium orthovanadate, and thepreviously mentioned protease inhibitor cocktail (Sigma-Aldrich).Phospho ERK 1/2 (p42/p44) and phospho-MEK1/2 were determinedby Western blotting and chemiluminescence detection by use of thePhototope-HRP Detection kit (Cell Signaling Technology, Danvers,MA). The following antibodies were used: phospho-p44/42 MAPK(ERK) (Thr 202/Tyr 204) and phospho-MEK1/2 (Ser217/221) antibod-ies (Cell Signaling Technology).
Chromatin Immunoprecipitation (ChIP) Assay and RNApolChIP. Rats were sacrificed at the indicated times after taurocholatetreatment and their pancreas were excised and immersed in 1%formaldehyde during 10 min at room temperature to cross-link thechromatin. ChIP and RNApol ChIP procedures were performed asdescribed previously (Sandoval et al., 2004).
The following antibodies were used against AP-1 (sc-1694), ATF2(sc-187), ATF3 (sc-188), CBP (sc-369), C/EBP� (sc-150), SP1 (sc-59),EGR1 (sc-110), �-HD1 (sc-6298), NF-�B (sc-109), PCAF (sc-8999),and Sin3A (sc-994) from Santa Cruz Biotechnology (Santa Cruz, CA)for ChIP assay, and against RNA polymerase II (sc-899 from SantaCruz Biotechnology) for RNApol ChIP.
PCR Analysis of Immunoprecipitated Chromatin. Input, IP,and NoAb fractions were analyzed by PCR with the appropriateprimer pairs to amplify products corresponding to either the pro-moter or the coding regions of the target genes. Primer sequences
were as follows: egr-1 (promoter): forward 5-GTAGAACCCCGGC-CTGACTC-3 and reverse 5-AGGCTCCTGGAGTTCCCAGC-3;icam-1 (promoter): forward 5-GGGATGGCCGTCCTGACTA-3 andreverse 5-GCCACTTTCCCGGAAACCT-3; il-6 (promoter): forward5-ATCAGCCCCACCCACTCTGG-3 and reverse 5-CGCCTGATGCT-GGCTGCTGG-3; inos (promoter): forward 5-GGTGCAGCTAA-GAAAAGCCTCC-3 and reverse 5-TTTATACCCATCCACGCTCTGC-3; tnf-� (promoter): forward 5-GGTGAGGACGGAGAGGAGATT-3and reverse 5-TGGGAGTTAGTACCAGGGTGTTC-3; atf3 (coding re-gion): forward 5-GGGTCACTGGTGTTTGAGGATT-3 and reverse5-GCTTGTTCTGGATGGCGAAT-3; egr-1 (coding region): forward 5-CAGTGGCCTTGTGAGCATGA-3 and reverse 5-GACGATGAAG-CAGCTGGAGAA-3; icam-1 (coding region): forward 5-TGTCGGT-GCTCAGGTATCCA-3 and reverse 5-TTCACCTGCACGGATCCA-3;il-6 (coding region): forward 5-AGAGGCACCTCAGTGGCTGC-3 andreverse 5-TGGGCTGACCTGAGACCTGC-3; inos (coding region): for-ward 5-ACTGGACCACCGCTGTCAGG-3 and reverse 5-CCTGCTT-TGCCACTTGCCAG-3.
RNA Extraction and Determination of Steady-State Levelsof mRNA. A small piece of the pancreas was excised and immedi-ately immersed in 1 ml of RNA-later solution (Ambion, Austin, TX)to stabilize the RNA. Total RNA was isolated from pancreas andfrom AR42J cell cultures by the guanidinium thiocyanate method(Chomczynski and Sacchi, 1987). The isolated RNA was size-frac-tioned by electrophoresis (2 �g/lane) in a 1% agarose/formalin geland stained with ethidium bromide to assess the quality of the RNA.The cDNA used as template for amplification in the PCR assay wasconstructed by reverse transcription reaction by use of SuperScript II(Invitrogen) with random hexamers as primers starting with 1 �g ofRNA. As a PCR internal control, 18S rRNA was simultaneouslyamplified, as indicated by Sandoval et al. (2004). Real-time PCR wasperformed by using the double-stranded DNA binding dye SyberGreen PCR Master mix (Applied Biosystems, Foster City, CA) in anABI GeneAmp 7000 Sequence Detection System. Each reaction wasperformed in triplicate and the melting curves were constructed withuse of Dissociation Curves Software (Applied Biosystems) to ensurethat only a single product was amplified. 18S rRNA was also ana-lyzed as real-time RT-PCR control. The following specific primerswere used: atf-3: forward 5-GGGTCACTGGTGTTTGAGGATT-3and reverse 5-GCTTGTTCTGGATGGCGAAT-3; egr-1: forward5-CAGTGGCCTTGTGAGCATGA-3 and reverse 5-GACGAT-GAAGCAGCTGGAGAA-3; icam-1: forward 5-TGTCGGTGCTCAG-GTATCCA-3 and reverse 5-TTCACCTGCACGGATCCA-3; inos:forward 5-AGCGGCTCCATGACTCTCA-3 and reverse 5-TGCAC-
CCAAACACCAAGGT-3; il-6: forward 5-TGTCGGTGCTCAGG-TATCCA-3 and reverse 5-TTCACCTGCACGGATCCA-3; tnf-�:forward 5- CAGCCGATTTGCCATTTCAT-3 and reverse 5-TCCT-TAGGGCAAGGGCTCTT-3; 18S rRNA: forward 5-AGTCCCTGC-CCTTTGTACACA-3 and reverse 5-GATCCGAGGGCCTCACTA-AAC-3. The threshold cycle (CT) was determined and the relativegene expression was expressed as follows: fold change � 2�(CT),where CT � CTtarget � CThousekeeping, and (CT) � CTtreated �CTcontrol.
Statistical Analysis. Results are expressed as mean � S.D. withthe number of experiments given in parentheses. Statistical analysiswas performed in two steps. First, a one-way analysis of variancewas carried out to find significance in the overall comparison ofgroups. Then, differences between individual groups were investi-gated by the Scheffe test. Differences were considered to be signifi-cant at p � 0.05.
ResultsPhosphorylation of ERK and MEK1/2 in Acute Pan-
creatitis. Effect of Pentoxifylline. Phosphorylation ofERK and MEK1/2 was measured by Western blotting as anindex of their activation in pancreas in the course of acutepancreatitis. Figure 1, A and B, shows a rapid and intensephosphorylation of both ERK and MEK1/2 after induction oftaurocholate-induced pancreatitis. Indeed, at 30 min afterinduction the phosphorylation of ERK and MEK1/2 was max-imal and declined progressively thereafter, especially in thecase of ERK.
In a previous work, we demonstrated that pentoxifyllinediminished ERK phosphorylation in pancreas in tauro-cholate-induced pancreatitis (Pereda et al., 2004). To testwhether this effect of pentoxifylline was mediated by inhibi-tion of MEK1/2, the MAPK kinase mainly responsible forERK phosphorylation, the effect of pentoxifylline on MEK1/2phosphorylation was measured. Our results show that pen-toxifylline does not prevent MEK1/2 phosphorylation at 1 hafter induction of pancreatitis (Fig. 1C). The absence of pen-toxifylline effect on MEK1/2 phosphorylation was also con-firmed at 6 h after induction (Fig. 1C). Alternatively, and toexplain the prevention of ERK phosphorylation by pentoxi-
Fig. 1. Phosphorylation of ERK (A) and MEK1/2 (B) in the course of acutepancreatitis. Effect of pentoxifylline (PTX) (C). Phospho-ERK1/2 andphospho-MEK1/2 were determined by Western blotting by use of tubulinas reference. Shown are representative Western blottings of four differentexperiments.
Fig. 2. Activities of PP2A, PP2B (calcineurin), and PP2C serine/threoninephosphatases in taurocholate-induced acute pancreatitis. Effect of pen-toxifylline. The activities of these three phosphatases were measured inpancreas at 1 h after induction of necrotizing pancreatitis. The number ofrats per group was 5 for PP2A, and 4 for PP2B and PP2C. The statisticaldifference is indicated as follows: �, P � 0.05 versus control; #, P � 0.05versus taurocholate. PTX, pentoxifylline; TAU, taurocholate.
Pentoxifylline Prevents Loss of PP2A in Pancreatitis 611
fylline, we decided to focus on serine/threonine phosphataseactivities as a potential target for this anti-inflammatoryagent.
Serine/Threonine Phosphatase Activities in AcutePancreatitis and in AR42J Acinar Cells. Effects of Pen-toxifylline. The activities of PP2A, calcineurin, and PP2Cserine/threonine (Ser/Thr) phosphatases were measured inpancreas at 1 h after pancreatitis induction. Figure 2 showsthat the activities of these three phosphatases were reducedby 50, 57, and 29%, respectively, at 1 h after induction.Pentoxifylline abrogated the loss in PP2A activity, but it didnot prevent the decreases in PP2B (calcineurin) and PP2C.The activities of dual MAPK phosphatases (MKP1, MKP2,and MKP3) were also measured, and pentoxifylline showedno effect on these phosphatases (results not shown).
To elucidate the mechanism responsible for the loss ofPP2A activity in taurocholate-induced pancreatitis and itsprevention by pentoxifylline, this phosphatase activity wasmeasured in AR42J acinar cells incubated with taurocholate.Figure 3A shows that in vitro taurocholate induced a rapidand transient decrease in PP2A activity in AR42J cells. Thisfall in PP2A activity was prevented by pentoxifylline or dibu-tyryl cAMP (see Fig. 3B). Therefore, the transient loss ofPP2A activity seems to be mediated by cAMP. Accordingly,taurocholate caused a rapid decrease in cAMP levels thatwas abrogated by pentoxifylline (Fig. 3C). Furthermore, pen-toxifylline also prevented the rapid ERK phosphorylationthat occurred in AR42J cells incubated with taurocholateunder the same conditions (Fig. 3D).
Expression of Early Responsive Genes (egr-1 andatf-3) in Acute Pancreatitis. Effect of Pentoxifylline.EGR-1 and ATF-3 are transcription factors induced rapidlyin signaling pathways triggered by stress and cytokines (Ji etal., 2003). Our results show that egr-1 was rapidly and mark-edly up-regulated in pancreas—more than 20-fold—afterpancreatitis induction (Figs. 4 and 5). Nevertheless, afterpentoxifylline administration the induction of this gene at1 h, although still significant, was only half the level of thosein the untreated group (Fig. 5). Likewise, the rapid andmarked up-regulation of atf-3 was also reduced by pentoxi-fylline (Fig. 5).
Expression of Late Responsive Genes (inos, icam-1,il-6, and tnf-�) in Pancreas in Pancreatitis. Effect ofPentoxifylline. Inducible nitric oxide synthase (inos), inter-cellular adhesion molecule 1 (icam-1), and interleukin 6 (il-6)were up-regulated in the course of taurocholate-induced pan-creatitis, as demonstrated by binding of RNA polymerase IIto their coding regions (Fig. 4), an index of actual transcrip-tional activity (Sandoval et al., 2004). The expression of inos,icam-1, and il-6 was measured by RT-PCR in pancreas in thecourse of acute pancreatitis at 0, 30 min, 1, 3, and 6 h. inoswas up-regulated more than 3-fold at 1 h and later more than6-fold at 3 and 6 h, whereas icam-1 was up-regulated morethan 3-fold at 1 h after induction and maintained around thislevel thereafter (results not shown). Il-6 was already up-regulated 15-fold at 1 h after induction and later it wasfurther induced (results not shown). Pentoxifylline reducedto a great extent the increase in inos expression and abro-gated the up-regulation of both icam-1 and il-6 at 6 h afterinduction (Fig. 5). In accordance with previous studies (Go-mez-Cambronero et al., 2000; Pereda et al., 2004), pentoxi-fylline also completely prevented the increase in tnf-� expres-
sion in the course of acute pancreatitis in the present study(results not shown).
Recruitment of Transcription Factors, HistoneAcetyltransferases, and Histone Deacetylases to thePromoters of Proinflammatory Genes. Effects of Pen-toxifylline. Epigenetic mechanisms that modulate chroma-tin structure, in particular, recruitment of transcription fac-tors and histone acetylation, determine the up-regulation ofproinflammatory genes. Figure 6A shows that egr-1 induc-tion is mediated, at least in part, by binding of SP-1 factor toits promoter. At 6 h, binding of EGR-1 itself to its own
Fig. 3. Taurocholate causes a rapid decrease in PP2A activity and cAMPlevels and induces ERK phosphorylation in AR42J acinar cells. Preven-tion by pentoxifylline. A, time course of PP2A activity in AR42J acinarcells incubated with 0.3% taurocholate. PP2A activity was measured at 0,5, 15, and 30 min after addition of taurocholate. B, prevention of tauro-cholate-induced decrease in PP2A activity in AR42J cells by pentoxifyl-line (12 mg/liter) or 0.5 mM dibutyryl cAMP (dbAMPc). Cells were incu-bated with 0.3% taurocholate for 5 min. C, cAMP levels in AR42J cellsincubated with 0.3% taurocholate for 5 min. Effect of pentoxifylline (12mg/liter). D, ERK phosphorylation in AR42J cells incubated with 0.3%taurocholate for 5 min. Effect of pentoxifylline (12 mg/liter). Phospho-ERK was detected by Western blotting. There were four to six experi-ments for each condition. The statistical difference is indicated as follows:�, P � 0.05 versus control or time 0; #, P � 0.05 versus taurocholate.
promoter might lead to negative feedback because egr-1 ex-pression started to decrease at 6 h after induction, concomi-tantly with EGR-1 recruitment. Pentoxifylline clearly dimin-ished the binding of SP-1 to the promoter of egr-1 at 1 h afterinduction.
Up-regulation of inos, icam and il-6 is mediated, at least inpart, by NF-�B and C/EBP� recruitment (Fig. 6, B–D). Pen-toxifylline markedly reduced the recruitment of both NF-�Band C/EBP� to the promoter of the three genes.
The level of histone acetylation is regulated by the oppositeaction of histone acetyltransferases and histone deacety-
lases, and their recruitment to gene promoters. In our exper-imental model of acute pancreatitis, the induction of proin-flammatory genes was associated with recruitment of histoneacetyltransferases to the gene promoters. Thus, up-regula-tion of icam-1 and tnf-� was associated with binding of CBPand PCAF, two major histone acetyltransferases, whereasegr-1 and inos induction was associated only with CBP re-cruitment (Fig. 7). Pentoxifylline increased binding of his-tone deacetylases to the promoters of icam-1 and inos, and inthe case of egr-1, icam-1, and tnf-� it prevented the binding ofhistone acetyltransferases (Fig. 7). Consequently pentoxifyl-line reduced the level of acetylation in the histones located atthe promoters of the proinflammatory genes leading to theirrepression.
DiscussionInactivation of MAPK by dephosphorylation is critical and
may occur through serine/threonine (Ser/Thr) phosphatases,dual-specificity (Thr/Tyr) phosphatases (MKPs), and proteintyrosine phosphatases. The role of protein phosphatases inacute pancreatitis seems to be complex. Calcineurin—or Ser/Thr phosphatase PP2B—mediates pancreatic zymogen acti-vation in acinar cells (Husain et al., 2007). Inhibition ofprotein tyrosine phosphatases causes dissociation of cell con-tacts in pancreatic acini as a prerequisite for the develop-ment of pancreatic edema (Schnekenburger et al., 2005).
Fig. 4. Binding of RNA polymerase II to exons of proinflammatory genes(egr-1, atf3, inos, icam-1, and il-6) in the pancreas in the course of acutepancreatitis. The binding of RNA polymerase II to exons (RNAPol-ChIP)is an index of transcriptional activity. No Ab, no antibody. These imagesare representative of four different experiments.
Fig. 5. Up-regulation of egr-1, atf3,icam-1, inos, and il-6 mRNAs in pan-creas in acute pancreatitis. Effect ofpentoxifylline. mRNA expression wasmeasured by real-time RT-PCR (seeMaterials and Methods). The numberof rats per group was 4. C, control; T,taurocholate; P, pentoxifylline. Thestatistical difference is indicated asfollows: *, P � 0.05, **, P � 0.01 ver-sus control; #, P � 0.05, ##, P � 0.01versus taurocholate.
Pentoxifylline Prevents Loss of PP2A in Pancreatitis 613
Furthermore, expression of the transmembrane protein ty-rosine phosphatase CD45 decreases in the course of acutepancreatitis in parallel with the up-regulation of TNF-� (deDios et al., 2006). Hence, CD45 seems to negatively controlthe production of cytokines. On the other hand, up-regulationof Thr/Tyr phosphatases (MKPs) and protein tyrosine phos-phatases SHP-1 and SHP-2 is an early event during acutepancreatitis (Hofken et al., 2000).
Our results point to Ser/Thr phosphatase PP2A as a keymodulator of the inflammatory cascade through the ERKpathway and histone acetylation. It is known that PP2A mayact as a negative regulator of MAPK signaling through the
ERK pathway (Letourneux et al., 2006). On the other hand,histone acetylation is associated with chromatin-modifyingcomplexes and gene activation (Shahbazian and Grunstein,2007).
Under basal conditions there is a well regulated balancebetween protein kinases and Ser/Thr phosphatases (Hunter,1995). Hence, any loss of phosphatase activity either triggersactivation of MAPK pathways or keeps them activated, lead-ing to the uncontrolled inflammatory network. In acute pan-creatitis we have found a rapid and marked decrease simul-taneously in three of the major Ser/Thr phosphataseactivities, that is, PP2A, calcineurin, and PP2C. Among
Fig. 6. Recruitment of transcriptionfactors to the promoters of egr-1 (A),icam-1 (B), nos-2 (C), and il-6 (D) inpancreas in acute pancreatitis. Effectof pentoxifylline. P, pentoxifylline; T,taurocholate; No Ab, no antibody.These images are representative offour different experiments.
them, PP2A is well known as a key regulator of MAPK, inparticular, of ERK activation. Accordingly, in acute pancre-atitis there is a rapid and marked phosphorylation of ERKthat is caused not only by MEK1/2 phosphorylation, but alsoby the decrease in PP2A activity.
Our in vitro results in acinar cells show that the rapiddecrease in cAMP levels is involved in the transient loss ofphosphatase PP2A activity leading to ERK activation. Thus,the metabolically stable analog dibutyryl cAMP completelyprevented the fall in PP2A activity in accordance with thecAMP-regulated PP2A activation reported previously (Fe-schenko et al., 2002). The decrease in intracellular cAMPlevels could be mediated by the adenylate cyclase inhibitioninduced by adenosine through A1 receptors. Indeed, accumu-lation of intracellular adenosine and its release into the ex-tracellular milieu are linked to excessive ATP catabolismthat occurs rapidly in acute pancreatitis (Luthen et al., 1995).
Furthermore, this process is relevant in the pathophysiologyof acute pancreatitis because administration of a selective A1agonist induces several features of the disease, such as leu-kocyte infiltration and interstitial edema in the pancreas andhyperamylasemia (Satoh et al., 2000). Based on our findings,activation of A1 receptors would lead to the decrease in PP2Acontributing decisively to the inflammatory cascade and leu-kocyte infiltration in acute pancreatitis.
Nevertheless, it should be taken into account that theeffects of extracellular adenosine on immune cells depend onthe interplay of A1 and A2 receptors and on the concentrationof adenosine (Sitkovsky et al., 2004). High-affinity A1 recep-tors inhibit adenylyl cyclase, but A2 receptors activate it.Furthermore, extracellular adenosine exhibits anti-inflam-matory effects acting through high-affinity A2A receptors(Lukashev et al., 2004; Sitkovsky and Ohta, 2005). Conse-quently, A1 receptors might be activated first at very lowadenosine levels followed later on by the stimulation of im-munosuppressive A2A receptors to trigger “OFF” signals inthe immune response.
cAMP seems to play a key role in the inflammatory re-sponse in acute pancreatitis. Accordingly, phosphodiesteraseinhibitors—such as rolipram or pentoxifylline—amelioratetaurocholate and cerulein-induced acute pancreatitis (Go-mez-Cambronero et al., 2000; Pereda et al., 2004; Sato et al.,2006). The beneficial actions of phosphodiesterase inhibitorsseem to be mediated by inhibition of leukocyte activation andinfiltration decreasing secretion of TNF-� and IL-6 and ad-herence of leukocytes to the endothelium (Klemm et al., 1995;Sekut et al., 1995; Haddad et al., 2002a; Pereda et al., 2004).On one hand, cAMP presumably enhances the activity ofcAMP-dependent protein kinase A and reduces the produc-tion of initiator proinflammatory cytokines TNF-� and IL-1�.On the other hand, and in accordance with our findings,maintenance of cAMP levels would avoid the loss of PP2Aactivity reducing the activation of ERK as well as the induc-tion of proinflammatory cytokines. Regarding the early re-sponsive gene egr-1, the reduction of its up-regulation medi-ated by pentoxifylline is also relevant in pancreatitis sincethe induction of MCP-1, IL-6, and ICAM-1 were diminishedin EGR-1-deficient mice (Ji et al., 2003).
Chromatin remodelling during induction of proinflamma-tory genes depends on phosphorylation of transcription fac-tors and subsequently their recruitment to gene promotersand histone acetylation. The latter mechanism opens thechromatin favoring anchorage of the transcriptional complexcontaining RNA polymerase II and gene up-regulation. Con-sequently, down-regulation of ERK phosphorylation eventu-ally may abrogate recruitment of transcription factors andhistone acetyltransferases, in favor of histone deacetylasesthat lead to deacetylation-mediated chromatin compaction.In acute pancreatitis, induction of proinflammatory genessuch as inos, icam-1, and il-6 involves mainly recruitment ofNF-�B and C/EBP�, both subjected to activation by phos-phorylation. Therefore, prevention of ERK phosphorylationwith pentoxifylline abrogates recruitment of NF-�B, C/EBP�,and histone acetyltransferases to the promoters togetherwith release of histone deacetylases from them. All thesemechanisms lead to a significant blockade of the inflamma-tory cascade.
These findings are in accordance with the well known roleof NF-�B as an early key event sufficient to initiate the
Fig. 7. Binding of histone deacetylases (mSin3A and HD1) and histoneacetyltransferases (CBP and PCAF) to the promoters of egr-1 (A), inos(B), icam-1 (C), and tnf-� (D) in pancreas at 3 h after induction ofnecrotizing pancreatitis. Effect of pentoxifylline (PTX). No Ab, no anti-body. These images are representative of four different experiments.
Pentoxifylline Prevents Loss of PP2A in Pancreatitis 615
inflammatory response in acute pancreatitis (Gukovsky etal., 1998; Chen et al., 2002). In addition, our results supportthe hypothesis that the earliest events in the evolution ofacute pancreatitis occur within acinar cells (Steer and Sa-luja, 1993) and that acinar cells behave as inflammatory cells(de Dios et al., 2005).
It is also noteworthy that the beneficial effects of pentoxi-fylline are mediated not only by inhibition of tnf-� up-regu-lation but also by blocking expression of two other relevantmediators of the inflammatory response such as icam-1 andil-6. Indeed, up-regulation of icam-1 mediates neutrophil ad-hesion to pancreatic acinar cells (Zaninovic et al., 2000) andIL-6 is considered a reliable marker of severity in acutepancreatitis. Consequently, pentoxifylline exhibits broadanti-inflammatory effects by keeping normal PP2A activityand it is not acting specifically on TNF-� production. In thisregard it should be highlighted that pentoxifylline reducesthe up-regulation of early responsive genes—such as egr-1—before blocking induction of late-responsive genes such astnf-�.
Nevertheless, it should be emphasized that the anti-in-flammatory action of pentoxifylline as a phosphodiesteraseinhibitor in acute pancreatitis seems to be mediated, at leastin part, by prevention of the rapid and presumably transientloss of PP2A phosphatase activity. Therefore, to achieve max-imal therapeutic efficiency phosphodiesterase inhibitorsshould be administered very early at the onset of the diseaseor just before a situation of risk such as endoscopic retro-grade cholangiopancreatography.
In conclusion, we have provided new insights into the keyrole of PP2A phosphatase as an early event in the inflamma-tory network in acute pancreatitis and we have also eluci-dated, at least in part, the mechanism of action of pentoxi-fylline as a potential therapy in this disease. The beneficialeffects of pentoxifylline–and presumably of other phosphodi-esterase inhibitors—seem to be mediated by abrogating theloss of cAMP levels and PP2A activity and by preventingchanges in chromatin remodeling very early in the course ofacute pancreatitis.
Acknowledgments
We thank Landy Menzies for revising the manuscript.
ReferencesBhatia M, Neoptolemos JP, and Slavin J (2001) Inflammatory mediators as thera-
peutic targets in acute pancreatitis. Curr Opin Investig Drugs 2:496–501.Bhatia M, Saluja AK, Hofbauer B, Frossard JL, Lee HS, Castagliuolo I, Wang CC,
Gerard N, Pothoulakis C, and Steer ML (1998) Role of substance P and theneurokinin 1 receptor in acute pancreatitis and pancreatitis-associated lung in-jury. Proc Natl Acad Sci U S A 95:4760–4765.
Chen X, Ji B, Han B, Ernst SA, Simeone D, and Logsdon CD (2002) NF-kappaBactivation in pancreas induces pancreatic and systemic inflammatory response.Gastroenterology 122:448–457.
Chomczynski P and Sacchi N (1987) Single-step method of RNA isolation by acidguanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156–159.
de Campos T, Deree J, Martins JO, Loomis WH, Shenvi E, Putnam JG, and CoimbraR (2008) Pentoxifylline attenuates pulmonary inflammation and neutrophil acti-vation in experimental acute pancreatitis. Pancreas 37:42–49.
de Dios I, Ramudo L, Alonso JR, Recio JS, Garcia-Montero AC, and Manso MA (2005)CD45 expression on rat acinar cells: involvement in pro-inflammatory cytokineproduction. FEBS Lett 579:6355–6360.
de Dios I, Ramudo L, García-Montero AC, and Manso MA (2006) Redox-sensitivemodulation of CD45 expression in pancreatic acinar cells during acute pancreati-tis. J Pathol 210:234–239.
Denham W, Yang J, Fink G, Denham D, Carter G, Ward K, and Norman J (1997)Gene targeting demonstrates additive detrimental effects of interleukin 1 andtumor necrosis factor during pancreatitis. Gastroenterology 113:1741–1746.
Feschenko MS, Stevenson E, Nairn AC, and Sweadner KJ (2002) A novel cAMP-
stimulated pathway in protein phosphatase 2A activation. J Pharmacol Exp Ther302:111–118.
Gomez-Cambronero L, Camps B, de La Asuncion JG, Cerda M, Pellín A, Pallardo FV,Calvete J, Sweiry JH, Mann GE, Vina J, et al. (2000) Pentoxifylline amelioratescerulein-induced pancreatitis in rats: role of glutathione and nitric oxide. J Phar-macol Exp Ther 293:670–676.
Grady T, Liang P, Ernst SA, and Logsdon CD (1997) Chemokine gene expression inrat pancreatic acinar cells is an early event associated with acute pancreatitis.Gastroenterology 113:1966–1975.
Gukovskaya AS, Gukovsky I, Zaninovic V, Song M, Sandoval D, Gukovsky S, andPandol SJ (1997) Pancreatic acinar cells produce, release, and respond to tumornecrosis factor-alpha. Role in regulating cell death and pancreatitis. J Clin Invest100:1853–1862.
Gukovsky I, Gukovskaya AS, Blinman TA, Zaninovic V, and Pandol SJ (1998) EarlyNF-kappaB activation is associated with hormone-induced pancreatitis. Am JPhysiol 275:G1402–G1414.
Haddad JJ, Land SC, Tarnow-Mordi WO, Zembala M, Kowalczyk D, and LauterbachR (2002a) Immunopharmacological potential of selective phosphodiesterase inhi-bition. I. Differential regulation of lipopolysaccharide-mediated proinflammatorycytokine (interleukin-6 and tumor necrosis factor-alpha) biosynthesis in alveolarepithelial cells. J Pharmacol Exp Ther 300:559–566.
Haddad JJ, Land SC, Tarnow-Mordi WO, Zembala M, Kowalczyk D, and LauterbachR (2002b) Immunopharmacological potential of selective phosphodiesterase inhi-bition. II. Evidence for the involvement of an inhibitory-kappaB/nuclear factor-kappaB-sensitive pathway in alveolar epithelial cells. J Pharmacol Exp Ther300:567–576.
Hofken T, Keller N, Fleischer F, Goke B, and Wagner AC (2000) Map kinasephosphatases (MKP’s) are early responsive genes during induction of ceruleinhyperstimulation pancreatitis. Biochem Biophys Res Commun 276:680–685.
Hunter T (1995) Protein kinases and phosphatases: the yin and yang of proteinphosphorylation and signaling. Cell 80:225–236.
Husain SZ, Grant WM, Gorelick FS, Nathanson MH, and Shah AU (2007) Caerulein-induced intracellular pancreatic zymogen activation is dependent on calcineurin.Am J Physiol Gastrointest Liver Physiol 292:G1594–G1599.
Ji B, Chen XQ, Misek DE, Kuick R, Hanash S, Ernst S, Najarian R, and Logsdon CD(2003) Pancreatic gene expression during the initiation of acute pancreatitis:identification of EGR-1 as a key regulator. Physiol Genomics 14:59–72.
Klemm P, Harris HJ, and Perretti M (1995) Effect of rolipram in a murine model ofacute inflammation: comparison with the corticoid dexamethasone. Eur J Phar-macol 281:69–74.
Letourneux C, Rocher G, and Porteu F (2006) B56-containing PP2A dephosphorylateERK and their activity is controlled by the early gene IEX-1 and ERK. EMBO J25:727–738.
Lukashev D, Ohta A, Apasov S, Chen JF, and Sitkovsky M (2004) Cutting edge:physiologic attenuation of proinflammatory transcription by the Gs protein-coupled A2A adenosine receptor in vivo. J Immunol 173:21–24.
Luthen R, Niederau C, Niederau M, Ferrell LD, and Grendell JH (1995) Influence ofductal pressure and infusates on activity and subcellular distribution of lysosomalenzymes in the rat pancreas. Gastroenterology 109:573–81.
Marshall CJ (1995) Specificity of receptor tyrosine kinase signaling: transient versussustained extracellular signal-regulated kinase activation. Cell 80:179–185.
Meskini N, Nemoz G, Okyayuz-Baklouti I, Lagarde M, and Prigent AF (1994)Phosphodiesterase inhibitory profile of some related xanthine derivatives phar-macologically active on the peripheral microcirculation. Biochem Pharmacol 47:781–788.
Pereda J, Sabater L, Aparisi L, Escobar J, Sandoval J, Vina J, Lopez-Rodas G, andSastre J (2006) Interaction between cytokines and oxidative stress in acute pan-creatitis. Curr Med Chem 13:2775–2787.
Pereda J, Sabater L, Cassinello N, Gomez-Cambronero L, Closa D, Folch-Puy E,Aparisi L, Calvete J, Cerda M, Lledo S, et al. (2004) Treatment of experimentalacute pancreatitis by simultaneous inhibition of TNF-� production and xanthineoxidase activity. Role of MAP kinases. Ann Surg 240:108–116.
Ramudo L, Manso MA, Sevillano S, and de Dios I (2005) Kinetic study of TNF-alphaproduction and its regulatory mechanisms in acinar cells during acute pancreatitisinduced by bile-pancreatic duct obstruction. J Pathol 206:9–16.
Sandoval J, Rodríguez JL, Tur G, Serviddio G, Pereda J, Boukaba A, Sastre J, TorresL, Franco L, and Lopez-Rodas G (2004) RNAPol-ChIP: a novel application ofchromatin immunoprecipitation to the analysis of real-time gene transcription.Nucleic Acids Res 32:e88.
Sato T, Otaka M, Odashima M, Kato S, Jin M, Konishi N, Matsuhashi T, andWatanabe S (2006) Specific type IV phosphodiesterase inhibitor ameliorates cer-ulein-induced pancreatitis in rats. Biochem Biophys Res Commun 346:339–344.
Satoh A, Shimosegawa T, Satoh K, Ito H, Kohno Y, Masamune A, Fujita M, andToyota T (2000) Activation of adenosine A1-receptor pathway induces edemaformation in the pancreas of rats. Gastroenterology 119:829–836.
Schandene L, Vandenbussche P, Crusiaux A, Alegre ML, Abramowicz D, Dupont E,Content J, and Goldman M (1992) Differential effects of pentoxifylline on theproduction of tumour necrosis factor-� (TNF-�) and interleukin-6 (IL-6) by mono-cyte and T cells. Immunology 76:30–34.
Schnekenburger J, Mayerle J, Kruger B, Buchwalow I, Weiss FU, Albrecht E,Samoilova VE, Domschke W, and Lerch MM (2005) Protein tyrosine phosphatasekappa and SHP-1 are involved in the regulation of cell-cell contacts at adherensjunctions in the exocrine pancreas. Gut 54:1445–1455.
Sekut L, Yarnall D, Stimpson SA, Noel LS, Bateman-Fite R, Clark RL, Brackeen MF,Menius JA Jr, and Connolly KM (1995) Anti-inflammatory activity of phosphodi-esterase (PDE)-IV inhibitors in acute and chronic models of inflammation. ClinExp Immunol 100:126–132.
Shahbazian MD and Grunstein M (2007) Functions of site-specific histone acetyla-tion and deacetylation. Annu Rev Biochem 76:75–100.
Sitkovsky MV, Lukashev D, Apasov S, Kojima H, Koshiba M, Caldwell C, Ohta A,
and Thiel M (2004) Physiological control of immune response and inflammatorytissue damage by hypoxia-inducible factors and adenosine A2A receptors. AnnuRev Immunol 22:657–682.
Sitkovsky MV and Ohta A (2005) The ‘danger’ sensors that STOP the immuneresponse: the A2 adenosine receptors? Trends Immunol 26:299–304.
Steer ML and Saluja AK (1993) Experimental acute pancreatitis studies of the earlyevents that lead to cell injury, in The Pancreas: Biology, Pathophysiology andDiseases (Go VLW, Di Magno EP, Gardner JD, Lebenthal E, Rebar HA, andScheele GA, eds) pp 489–499, Raven, New York.
Yang J, Murphy C, Denham W, Botchkina G, Tracey KJ, and Norman J (1999)
Evidence of a central role for p38 map kinase induction of tumor necrosis factoralpha in pancreatitis-associated pulmonary injury. Surgery 126:216–222.
Zaninovic V, Gukovskaya AS, Gukovsky I, Mouria M, and Pandol SJ (2000) Ceruleinupregulates ICAM-1 in pancreatic acinar cells, which mediates neutrophil adhe-sion to these cells. Am J Physiol Gastrointest Liver Physiol 279:G666–G676.
Address correspondence to: Dr. Juan Sastre, Department of Physiology,School of Pharmacy, University of Valencia, Avda. Vicente Andres Estelles s/n,46100 Burjasot (Valencia), Spain. E-mail: [email protected]
Pentoxifylline Prevents Loss of PP2A in Pancreatitis 617
