Gastrin induces leukocyte-endothelial cell interactions in vivo and contributes to the inflammation...

The FASEB Journal • FJ Express Full-Length Article

Gastrin induces leukocyte-endothelial cell interactionsin vivo and contributes to the inflammation causedby Helicobacter pylori

Angeles Alvarez,* Sales Ibiza,† Carlos Hernandez,* Alberto Alvarez-Barrientos,‡

Juan V. Esplugues,*,1and Sara Calatayud**Departamento de Farmacologıa and †Unidad Mixta CNIC-UVEG, Facultad de Medicina,Universidad de Valencia, Valencia, Spain; and ‡Cytometry Unit, Fundacion Centro Nacional deInvestigaciones Cardiovasculares, Madrid, Spain

ABSTRACT Gastric mucosal inflammation causes hy-pergastrinemia, and gastrin receptors have been de-tected in several leukocyte types. We have analyzedwhether gastrin affects the leukocyte-endothelial cellinteractions in vivo by monitoring leukocyte rolling,adhesion, and emigration in rat mesenteric venulesusing intravital microscopy. Mesenteric superfusionwith exogenous gastrin increased these processes in aconcentration- and time-dependent manner, effectsprevented by the cholecystokinin (CCK)-2 receptorantagonists (proglumide, L-365,260) but not by theCCK-1 receptor antagonist devazepide. A similar re-sponse was induced by exogenous CCK or endog-enously released gastrin. CCK-2 receptors were localizedin mesenteric macrophages and polymorphonuclearleukocytes. This effect of gastrin is not modulated bysomatostatin and is independent of the endogenousrelease of histamine. To analyze whether hypergastrine-mia elicited by Helicobacter pylori (HP) modulates theinflammation induced by the germ, rats were chroni-cally administered with an extract of a CagA�/VacA�strain of HP. This protocol increased gastrinemia andinduced an inflammatory response in the rat mesentery.Blockade of CCK-2 receptors attenuated this responseand induced a qualitative change in the leukocyteinfiltrate suggestive of a receding inflammatory pro-cess. Our results reveal a new proinflammatory role ofgastrin that seems to contribute to the maintenance ofthe inflammation elicited by HP components.—Alvarez,A., Ibiza, S., Hernandez, C., Alvarez-Barrientos, A.,Esplugues, J. V., Sara Calatayud, S. Gastrin inducesleukocyte-endothelial cell interactions in vivo and con-tributes to the inflammation caused by Helicobacterpylori. FASEB J. 20, E1742–E1752 (2006)

Key Words: CCK-2 receptors � gastrin-releasing peptide

Gastrin is mainly known by its first described role asa gastric acid-secreting hormone and, more recently, byits ability to promote cellular growth. Peptides of thegastrin family activate two different receptors: theCCK-1 receptor, which has low affinity for gastrin but

high affinity for the related hormone cholecystokinin(CCK), and the CCK-2 receptor, which has high affinityfor both gastrin and CCK and mediates the acid-secretory as well as the proliferative effects of gastrin(1). Recent studies using gene arrays have started toidentify several previously unsuspected target genesdownstream of the CCK-2 receptor, and accumulatingevidence suggests that gastrin may have a wider rangeof effects (2).

Gastrin receptors have been observed in severalhuman leukocyte types. For instance, CCK-1 and CCK-2receptors or their mRNAs have been localized in lam-ina propria macrophages (3), peripheral blood mono-nuclear cells (4, 5), circulating polymorphonuclear(PMN) leukocytes (6), and PMNs within the tumorstroma of human colorectal cancers (7). In rats, pul-monary vascular endothelial cells and macrophageshave been shown to express both CCK-1 and CCK-2receptor genes (8). In parallel with these molecularstudies, gastrin and CCK have been shown to inducechanges in leukocyte functions such as chemotaxis,adherence, or phagocitosis in vitro. In particular, bothpeptides seem to be chemoattractant for human neu-trophils (9), monocytes (10), and lymphocytes (11).However, it has not been analyzed whether these cellu-lar effects of gastrin have any consequence on thefunctioning of the immune system in vivo.

It is important to note that infection by Helicobacterpylori (HP), the main cause of nonautoimmune chronicgastritis, increases gastrin secretion (12–15), and acorrelation between gastrinemia and the severity ofgastritis has been identified (16–18). This hypergas-trinemia has been regarded as a consequence of thelocal inflammatory process, since it was shown thatproinflammatory cytokines and activated monocytesstimulate gastrin release from antral G cells (19–21).However, if gastrin could modify leukocyte behavior in

1 Correspondence: Departamento de Farmacologıa, Fac-ultad de Medicina, Universidad de Valencia, Avd. BlascoIbanez 15, 46010-Valencia, Spain. E-mail: [email protected]

doi: 10.1096/fj.05-5696fje

E1742 0892-6638/06/0020-1742 © FASEB

vivo, its increased concentration in the HP-infectedgastric mucosa might affect the evolution of the inflam-matory process induced primarily by the microorgan-ism, i.e., gastritis. In the present study, we have evalu-ated this possibility in rats administered chronicallywith an extract of a CagA�/VacA� strain of HP, whichmimics the HP infection in humans.

MATERIALS AND METHODS

Intravital microscopy

Leukocyte-endothelial cell interactions were evaluated infasted male Sprague-Dawley rats (200–250 g), the details ofthe experimental preparation having been described previ-ously (22, 23). In brief, rats were anesthetized with sodiumpentobarbital (65 mg/kg, i.p.). A midline abdominal incisionwas made, and a segment of the midjejunum was exteriorizedand placed over a transparent pedestal for tissue transillumi-nation. A selected loop of the exposed mesentery was contin-uously superfused with bicarbonate buffer saline (pH 7.4,37°C, 2 ml/min) and observed through an orthostatic micro-scope equipped with a video camera. Images were capturedon videotape for playback analysis (final magnification of thevideo screen was �1300).

The mesentery was left to stabilize for a period of 30 min.When the agent to be tested was administered before theintravital microscopy experiment, images of two to threeunbranched mesenteric venules (with diameters between 25and 40 �m) were recorded for a period of 5 min per venule.When the agent was administered during the intravital mi-croscopy experiment, a single venule was selected and imageswere recorded for 5 min periods at baseline (time 0) and at 15min intervals over a 60 min period after administration.

The numbers of rolling, adherent, and emigrated leuko-cytes were determined off-line during playback analysis ofvideotaped images. Rolling leukocyte flux was assessed bycounting the number of leukocytes passing a reference pointin the vessel per minute. Leukocyte rolling velocity (Vwbc) wascalculated by measuring the time required for a leukocyte totraverse a distance of 100 �m along the length of the venuleand was expressed as micrometers per second. A leukocytewas considered to be adherent to the venular endothelium ifit remained stationary for a period equal to or exceeding 30 s.Adherent cells were expressed as the number of white bloodcells per 100 �m of venule. Leukocyte emigration induced bychronic treatments was evaluated as the total number ofinterstitial leukocytes per field. When the proinflammatoryagent was administered acutely, leukocyte emigration wasdetermined as the number of interstitial leukocytes in closeproximity to the selected vessel (50 �m above and below thevessel). Systemic arterial blood pressure, venular diameters,and centerline red blood cell velocity were evaluated on-line,and venular blood flow and venular wall shear rate (�) werecalculated as described previously (24).

Gastrin concentration was measured by radioimmunoassayin plasma from the portal blood collected in heparine at theend of the experiment. In some cases, the area of themesentery selected for the experiment was excised, fixed withparaformaldehyde (4% in PBS, pH 7.4) and stained withhematoxylin and eosin. The preparation was subsequentlyobserved under a clear field microscope (�40), and theinfiltrated leukocytes were counted (number per 0.5 cm2)and classified morphologically into PMN leukocytes, macro-phages and lymphocytes by an observer who was unaware ofthe treatment previously given.

Experimental design

Effects of exogenous gastrointestinal hormones onleukocyte-endothelial cell interactions

The effects of gastrin (0.01–10 nM), pentagastrin (1 nM),CCK (1–100 nM), and somatostatin (1 �M) on leukocyte-endothelial cell interactions were evaluated using intravitalmicroscopy. Each of these agents was added to the superfu-sion buffer after baseline measurements were taken, and theireffects were evaluated over the following 60 min.

Some animals were pretreated before gastrin (1 nM) orCCK (10 nM) superfusion with an antagonist of the CCK-2receptor (proglumide, 30 mg/kg, i.p.; or L-365,260, 1 mg/kg,i.v.) or with the CCK-1 receptor antagonist devazepide (1mg/kg, i.v.).

To analyze whether gastrin was acting through the endog-enous release of histamine, the H1 receptor antagonist di-phenhydramine (2 �M) was cosuperfused with gastrin (1nM). The involvement of mast cell activation was evaluated byadministering the mast cell stabilizing agent cromolyn beforeinitiating surgery (20 mg/kg, i.v.) and thereafter cosuperfus-ing 0.33 mg/ml of the drug with gastrin (1 nM). The effect ofgastrin (1 nM) on mast cell activation was also analyzed bysupplementing the superfusion buffer with 0.001% ruthe-nium red, which is captured by activated mast cells (25).Superfusion with compound 48/80 at 1 �g/ml was used as apositive control of mast cell activation.

The selected doses of antagonists were taken from previousin vivo studies using either proglumide (26), L-365,260 (27),devazepide (27–29), diphenhydramine (30), or cromolyn(31). The dose of proglumide used all along this study did nothave any effect on the increase of leukocyte-endothelialinteractions induced by a common proinflammatory agentlike PAF (10�7M, data not shown).

Acute effects of endogenous gastrin on leukocyte-endothelialcell interactions

The endogenous release of gastrin was stimulated by twodifferent procedures. In a first set of experiments, an intra-venous infusion of gastrin releasing peptide (GRP, 300 pmol/kg/h) or saline was initiated after measurement of baselinesand leukocyte-endothelial cell interactions were evaluatedduring the following 60 min. A second group of animals wastreated with a maximal acid-inhibitory dose of the protonpump inhibitor omeprazole (40 mg/kg, p.o.), and leukocyteparameters were measured 5 h later, coinciding with thepeaking of the gastrin plasma level (32). When necessary, ratswere pretreated with proglumide (30 mg/kg, i.p., �30 min).

Effects of an HP extract on leukocyte-endothelialcell interactions

The extract of HP was prepared from a strain CagA�/VacA�

(NCTC 11638). The bacteria were grown in Brucella brothculture (107 CFU/plate) supplemented with 5% FBS in acarbon dioxide incubator and maintained at 37°C for 5 days.The HP organisms were harvested in 1 ml of 0.15 M NaCl andcentrifuged at 3,000 g for 25 min at 25°C as describedpreviously (33, 34). The pellet was then resuspended in anequal volume of sterile distilled water, which was vortex-mixed for 45 s and centrifuged at 3,000 g. The resultingsupernatant, containing water-extracted surface proteins, wasfinally filtered using a nylon mesh (0.2 �m) and stored at–20°C. The extracts were diluted with saline to normalize thetotal protein content to 100 �g/ml.

Rats received either the aqueous extract of HP or saline on

E1743GASTRIN AND INFLAMMATION BY HELICOBACTER PYLORI

days 1, 3, 5, and 7 (1 ml/rat, p.o.), and leukocyte-endothelialcell interactions were evaluated on day 14 (33). Some animalswere administered daily with the CCK-2 receptor antagonistproglumide (30 mg/kg/day, i.p., 14 days).

Detection of CCK-2 receptor mRNA by reversetranscriptase-polymerase chain reaction

CCK-2 receptor mRNA expression was analyzed in the mes-enteric tissues and in peritoneal cell suspensions from controlrats and in animals suffering an acute peritoneal inflamma-tion after a zymosan injection (10 mg/rat in PBS, i.p.). In thelatter case, peritoneal inflammatory cells were collected 4 hlater by means of an abdominal lavage with 20 ml of ice-coldPBS medium containing 2 mM EDTA (35). Rat gastric corpuswas used as a positive control.

Total RNA from mesenteric and gastric tissues or fromperitoneal cell suspensions was isolated with TriPure IsolationReagent (Roche Diagnostics, Indianapolis, IN) or withRNeasy Mini Kit (Quiagen, Valencia, CA), respectively, andwas treated with a DNA-free kit (Ambion, Austin, TX) toeliminate the possible traces of contaminating genomic DNA.Reverse transcription of this RNA was performed with Super-Script reverse transcriptase (RT; Invitrogen, Carlsbad, CA),using 0.5 �g oligo(dT16) and 40 U of RNase inhibitor (RocheDiagnostics, Indianapolis, IN). Minus-RT controls were in-cluded for each sample. For nested-polymerase chain reac-tion (PCR), reactions were produced in a LightCycler instru-ment (Roche Diagnostics) with the aid of a LightCycler-FastStart DNA Master SYBR Green I kit (Roche Diagnostics).The first round of PCR (annealing temperature: 57°C, 35cycles) from cDNA was performed with specific primers forCCK-2 receptor (36): 5�-CTTCATCCCGGGTGTGGTTATT-GCG-3� (sense) and 5�-CCCCAGTGTGCTGATGGTGGTAT-AGC-3� (antisense), generating a product of 669 bp. Afterpurification with Montage PCR (Millipore, Bedford, MA),PCR products were used as templates for the second round ofPCR (annealing temperature: 60°C, 25 cycles) with internalprimers designed in our laboratory: 5�-TGGCCTATGGA-CTCATCTCC-3� (sense) and 5�-AGCAGCCATCACTGTCT-TCC-3� (antisense), which give rise to a product of 189 bp.Specificity of PCR was confirmed by melting curve analysis,agarose gel electrophoresis and sequenciation of the ampli-fied product.

Immunohistochemical studies

Mesenteric windows were extracted, cleared of surroundingfat, whole-mounted on gelatin-coated slides, fixed for 10 minwith paraformaldehyde (4% in PBS pH 7.4), washed withPBS, and immersed in methanol (�20°C) for 5 min. Afterantigen retrieval with alpha-chymotrypsin (Sigma ChemicalCO) and blocking (10% goat serum, 1% BSA), specimenswere incubated with a rabbit polyclonal anti–cholecystoki-nin-B receptor antibody (Ab; Abcam, Cambridge, MA, 1:100,4°C, overnight). A goat anti-rabbit HRP conjugate (SouthernBiotech, Birmingham, AL, 1:100) was used as secondary Aband was incubated for 1 h at room temperature. Finally,tissues were incubated with 3,3�-diaminobenzidine (DAB)Enhanced Liquid substrate System for Immunohistochemis-try (Sigma Chemical, St. Louis, MO). In some cases, mesen-teric macrophages were identified by applying an analogousprotocol to a second sample from the same animal butchanging the primary Ab for a mouse monoclonal anti-macrophage � granulocyte Ab ([OX41], Abcam, 1:400) or amouse monoclonal antibody (mAb) against an ED2-like anti-gen ([HIS36], eBioscience, San Diego, CA, 1:100).

All experimental protocols were performed according to

the guidelines approved by the Ethical Committee for Exper-imental Research of the Faculty of Medicine of the Universityof Valencia.

Drugs

Gastrin, CCK, pentagastrin, somatostatin, GRP, omeprazole,proglumide, diphenhydramine, cromolyn, compound 48/80,and zymosan were all obtained from Sigma Chemical. Pento-barbital was from B. Braun Medical SA (Rubi, Barcelona,Spain). L-365,260 and devazepide were kindly donated by MLLaboratories PLC (Warrington, UK).

Statistical Analysis

All values are mean � sem. Data within groups were com-pared using an ANOVA (one-way ANOVA) followed by aNewman-Keuls post hoc test. The differences were consideredsignificant when the P value was �0.05.

RESULTS

Acute effects of agonists and antagonists of CCK-gastrin receptors on leukocyte-endothelialcell interactions

Mesenteric venules exposed to gastrin superfusion ex-perienced a concentration- and time-dependent in-crease in leukocyte-endothelial cell interactions. Gas-trin increased the number of rolling leukocytes, whiledecreasing their rolling velocity. This effect was rapid inonset (15 min) and was observed with all the gastrinconcentrations used (0.01-1 nM). This response wasfollowed by significant increases in the number ofadherent leukocytes and leukocytes emigrated into theinterstitial tissue. These two effects were observed prin-cipally after superfusion with gastrin 0.1 and 1 nM (Fig.1). Similar effects were induced by superfusion withpentagastrin 1 nM (rolling flux(60 min): 70�18, rollingvelocity(60 min): 50�4; adhesion(60 min): 10�2; emigra-tion(60 min): 2.3�0.3, P�0.05 vs. buffer and P�0.05 vs.basal values in all cases). Mesenteric superfusion withCCK also increased the interaction between leukocytesand venular endothelial cells (Fig. 2). The resultsincluded in Figs. 1 and 2 correspond to the maximaleffects of these hormones (induced by gastrin 1 nM andCCK 10 nM) since similar responses were obtained by10-fold increased concentrations (data not shown).

Pretreatment with proglumide prevented both theincrease in leukocyte rolling, adhesion, and emigrationand the reduction in rolling velocity induced by gastrin(1 nM). The proinflammatory action of gastrin was alsoinhibited by pretreatment with another CCK-2 receptorantagonist, L-365,260. However, pretreatment with theCCK-1 antagonist devazepide failed to modify the ef-fects of gastrin on leukocyte-endothelial cell interac-tions (Fig. 3). In a similar way, the proinflammatoryeffects of CCK (10 nM) were prevented by proglumiderather than devazepide (Fig. 4).

The proinflammatory effects induced by gastrin(1 nM) were unaltered by the addition of somatostatin1 �M to the superfusion buffer (Table 1). This concen-

E1744 Vol. 20 November 2006 ALVAREZ ET AL.The FASEB Journal

tration of somatostatin did not affect per se any of theleukocyte parameters (rolling flux(60 min): 23�4, rollingvelocity(60 min): 108�11; adhesion(60 min): 2.6�0.5; em-igration(60 min): 0.8�0.2). Gastrin effects were not mod-ified by either administration of an antagonist of the H1receptor for histamine (diphenhydramine) or stabiliza-

tion of mast cells with cromolyn (Table 1). Further-more, superfusion with gastrin did not induce ruthe-nium red uptake in mast cells.

Basal hemodynamics were similar in all groups andnone of the treatments produced significant changes inthese parameters.

Figure 1. Effects of exogenous gastrin on leu-kocyte-endothelial cell interactions in rat mes-entery. Gastrin superfusion induced a concen-tration- and time-dependent increase inleukocyte rolling flux (A), adhesion (C), andemigration (D), with a parallel reduction inrolling velocity (B), in mesenteric postcapillaryvenules. Animals were divided into 4 groups:buffer, gastrin 0.01, 0.1, and 1 nM, and param-eters were measured at 0, 15, 30, and 60 min.Results are mean � sem. *P � 0.05, **P � 0.01vs. respective basal value (0 min).

Figure 2. Effects of exogenous CCK on leuko-cyte-endothelial cell interactions in rat mesen-tery. CCK superfusion induced a concentration-and time-dependent increase in leukocyte roll-ing flux (A), adhesion (C), and emigration (D),with a parallel reduction in rolling velocity (B),in mesenteric postcapillary venules. Animalswere divided into 3 groups: buffer, CCK 1 and10 nM, and parameters were measured at 0, 15,30, and 60 min. Results are mean � sem. **P �0.01 vs. respective basal value (0 min).


Acute effects of endogenous gastrin on leukocyte-endothelial cell interactions

GRP infusion significantly increased portal gastrine-mia (36�5 vs. 23�2 pM in control animals,P�0.05) and leukocyte-endothelial cell interactions,

which were reduced by pretreatment with proglu-mide (Fig. 5). Similarly, omeprazole pretreatmentresulted in hypergastrinemia (68�8 vs. 20�1 pM incontrol animals, P�0.05) and augmented interac-tions between leukocytes and the venular endothe-lium, which were also prevented by proglumide (Fig.

Figure 3. Effects of CCK/gastrin receptor antag-onists on gastrin-induced leukocyte/endothe-lium interactions. Rats were pretreated with twodifferent CCK-2 receptor antagonists (proglu-mide 30 mg/kg, i.p.; or L-365,260 1 mg/kg, i.v.)or with a CCK-1 receptor antagonist (devaz-epide 1 mg/kg, i.v.). Rolling flux (A), rollingvelocity (B), adhesion (C), and emigration (D)were evaluated in mesenteric postcapillaryvenules superfused with gastrin 1 nM for 60min. Results are mean � sem. *P � 0.05, **P �0.01, ***P � 0.001 vs. buffer; �P � 0.05, ��P �0.01, ��P � 0.001 vs. gastrin 1 nM.

Figure 4. Effects of CCK-gastrin receptor antag-onists on CCK–induced leukocyte/endothe-lium interactions. Rats were pretreated with aCCK-2 receptor antagonist (proglumide 30mg/kg, i.p.) or a CCK-1 receptor antagonist(devazepide 1 mg/kg, i.v.). Rolling flux (A),rolling velocity (B), adhesion (C), and emigra-tion (D) were measured before (time 0) and15, 30, and 60 min after the initiation of CCKsuperfusion. Results are mean � sem. �P �0.05, ��P � 0.01, ��P � 0.001 vs. untreatedgroup.


5). None of these treatments affected the hemody-namic parameters.

HP extract causes mesenteric inflammation. Effect ofa CCK-2 receptor antagonist

Rats chronically administered with an aqueous extract of aCagA�/VacA� strain of HP showed an active inflammatoryresponse in the mesenteric tissue as denoted by a highernumber of rolling leukocytes moving at a slower rate, anincreased leukocyte adhesion, and a significant increase inthe number of leukocytes emigrated into the interstitiumsurrounding the selected venules (Fig. 6). Analysis of the

hematoxylin and eosin-stained mesenteric tissue also re-vealed the presence of an inflammatory infiltrate in HP-treated animals (322�20 leukocytes per field in HP-treatedrats vs. 44�2 leukocytes per field in control animals,P�0.001), consisting mainly of PMN leukocytes (79�1%),while macrophages and lymphocytes amounted to 15 � 1and 6 � 1% of the total, respectively.

Chronic administration of the aqueous extract of HPalso caused hypergastrinemia (39�7 vs. 19�1 pM gas-trin in control rats, P�0.05), and daily treatment ofthese animals with the CCK-2 receptor antagonist pro-glumide significantly reduced the number of rolling,adherent and emigrated leukocytes (Fig. 6). Hemody

TABLE 1. Effects of somatostatin, an H1 receptor antagonist (diphenhydramine), or a mast cell stabilizing agent (cromolyn) ongastrin-induced leukocyte-endothelial cell interactions in rat mesenteric postcapillary venulesa

Gastrin � Somatostatin � Diphenhydramine � Cromolyn

Rolling Flux0 min 29 � 4 25 � 3 24 � 2 26 � 260 min 75 � 6** 74 � 12** 82 � 6** 78 � 9**Rolling Velocity0 min 101 � 2 100 � 6 94 � 5 108 � 560 min 38 � 6** 49 � 5** 52 � 6** 55 � 3**Adhesion0 min 1.3 � 0.5 1.7 � 0.6 1.4 � 0.2 1.0 � 0.460 min 8.8 � 1.0** 6.0 � 1.2** 8.6 � 1.0** 7.6 � 1.3**Emigration0 min 0.3 � 0.3 0.5 � 0.3 0.2 � 0.2 0.2 � 0.260 min 2.5 � 0.3** 2.3 � 0.3** 2.4 � 0.4** 2.2 � 0.4**

a Values correspond to measurements made in basal conditions (0 min) and after 60 min of superfusion with gastrin 1 nM in animalsuntreated (n4) or cotreated with somatostatin (n4), diphenhydramine (n5), or cromolyn (n5). Results are mean � sem. **P � 0.01 vs.corresponding basal value.

Figure 5. Leukocyte-endothelial cell interac-tions induced by two drugs inducing acutehypergastrinemia and effects of a CCK-2 recep-tor antagonist. Rats were treated with gastrinreleasing peptide (GRP, 300 pmol/kg/h, i.v.infusion) or omeprazole (40 mg/kg, p.o.) andtheir effects on leukocyte rolling flux (A), roll-ing velocity (B), adhesion (C), and emigration(D) were analyzed in mesenteric postcapillaryvenules 60 min after the initiation of GRP/saline i.v. infusion or 5 h after omeprazole/vehicle administration. Some animals were pre-treated with proglumide (30 mg/kg, i.p.). *P �0.05 and ***P � 0.001 vs. respective control;�P � 0.05, ��P � 0.01 and ��P � 0.001 vs.GRP � vehicle of proglumide group; #P � 0.05,##P � 0.01 and ###P � 0.001 vs. omeprazole �vehicle of proglumide group. (saline i.v.: n7,GRP: n10, proglumide�GRP: n5, vehiclep.o.: n14, omeprazole: n13, proglumide�omeprazole: n6).


namic parameters were similar in all groups. Hematox-ylin and eosin staining revealed that rats receivingproglumide also presented an inflammatory infiltratebut with fewer leukocytes (244�12 vs. 322�20 leuko-cytes per field, P�0.01) and differing proportions ofeach leukocyte type. In contrast to what was observed inHP-treated rats (see previous paragraph), the infiltratein rats receiving proglumide contained similar num-bers of PMN leukocytes and macrophages (44�3 and50�4%, respectively; Fig. 7).

Analysis of the expression of the CCK-2 receptor

CCK-2 receptor mRNA, analyzed by nested RT-PCR,was detected in both mesenteric tissues of control ratsand peritoneal cell suspensions of zymosan-treated rats(Fig. 8).

Immunohistochemical studies revealed the presenceof CCK-2 receptor in two cellular types of mesenterictissue, which were later identified in the hematoxylinand eosin staining as PMN leukocytes and macro-phages. The identity of macrophages was further con-firmed by comparing the morphology of non-PMNCCK-2 receptor positive cells with cells stained with twodifferent macrophage membrane markers in parallelimmunohistochemical studies (Fig. 9). Endothelial cellstested negative for the CCK-2 receptor immunostaining.

DISCUSSION

The present study demonstrates for the first time thatgastrin has a proinflammatory action in vivo and sug-

gests that this peptide may directly contribute to theinflammatory response elicited by HP.

We have observed by intravital microscopy that gas-trin enhances the interactions between the flowingleukocytes and the venular endothelium in rat mesen-tery. Gastrin increases the number of rolling leuko-cytes, while it reduces their rolling velocity and en-hances leukocyte adhesion and emigration into theinterstitium. Since these are the initial steps in theformation of inflammatory foci, our results indicatethat gastrin induces a proinflammatory activity. As saidbefore, peptides of the gastrin family mediate theireffects on target tissues by activating two differentreceptors named CCK-1 and CCK-2. Our results suggestthat the proinflammatory activity of gastrin is a CCK-2receptor-mediated process since it can be reversed byits specific antagonists proglumide and L-365,260 butnot by the CCK-1 antagonist devazepide. Significantincreases in leukocyte rolling, adhesion, and emigra-tion were also observed with pentagastrin and CCKsuperfusion and when the endogenous release of gas-trin was stimulated by the neuropeptide GRP or thepotent antisecretory drug omeprazole. In all cases,these events were partially reversed by pretreatmentwith proglumide.

CCK-2 receptor mRNA was detected in rat mesen-teric tissue extracts and, through immunohistochemis-try studies, those receptors were localized in mesentericmacrophages and PMN leukocytes. Likewise, CCK-2receptor mRNA was also present in extracts fromleukocyte suspensions obtained by peritoneal lavage ofzymosan-treated rats, thus confirming the location ofthe receptor in rat inflammatory cells. These results aresimilar to those previously reported for humans (3–8).

Figure 6. Leukocyte-endothelial cell interac-tions induced by the HP extract and effect of aCCK-2 receptor antagonist. Rats were treatedchronically (days 1, 3, 5, and 7) with the aque-ous extract of HP or saline and their effects onleukocyte rolling flux (A), rolling velocity (B),adhesion (C), and emigration (D) were ana-lyzed in mesenteric postcapillary venules on day14. Animals were divided into 4 groups: saline� vehicle of proglumide (n7), saline � pro-glumide (n5), HP extract � vehicle of proglu-mide (n5), and HP extract � proglumide(n7). Results are mean � se. *P � 0.05 and***P � 0.001 vs. saline � vehicle of proglumidegroup; �P � 0.05 and ��P � 0.01 vs. HP extract� vehicle of proglumide group.


CCK-2 receptors are G protein-coupled receptors that,on binding to their peptide agonists, activate proteinsof the Gq or G12,13 subfamilies (37, 38). The signal-ing pathways initiated by these proteins share severalelements with those that lead to leukocyte activationafter chemokine binding to their G protein-coupledreceptors or after �2-integrin engagement (39–42).Although it is beyond the scope of the present study, wesuspect that some of these elements account for the

proinflammatory activity of gastrin that we observed invivo and/or the modulation of leukocyte adhesion andemigration previously demonstrated in vitro (11, 43).

In the HP-infected mucosa both elements, significantgastrin levels (16–18) and infiltrated leukocytes can befound, and, considering the effects observed with gas-trin, a modulatory effect of the peptide on the localleukocytes function should be expected. Natural colo-nization by HP is observed exclusively in humans andprimates, and different strategies have been used to getvalid murine and rat models of infection. We haveadministered an extract obtained from bacterial sus-pensions containing many of the pathogenic compo-nents of the germ (44). Similar HP extracts havepreviously been shown to induce an acute inflamma-tory response in the digestive tract (45–47), which isgenerally attributed to a direct effect of bacterial com-ponents on leukocytes (34, 45, 48–50). In our experi-mental design, the extract was administered for 1 wk,and the experiments were performed 7 days later, whenincreased gastrin plasma levels were detectable (33, thisstudy). At this stage, despite the suspension of admin-istration of the extract during a relatively long period,there was still an active inflammatory process in the

Figure 7. Leukocyte infiltration in the mesenteryof rats treated with the HP extract and effect ofa CCK-2 receptor antagonist. Rats received saline� vehicle of proglumide (A), HP extract �vehicle of proglumide (B), saline � proglumide(C), or HP extract � proglumide (D). Chronicadministration (days 1, 3, 5, and 7) of theaqueous extract of Helicobacter pylori induces theformation of an inflammatory infiltrate in the ratmesentery mainly composed of PMN leukocytes(B). Daily treatment with proglumide reducesthe number of extravasated leukocytes and in-creases relative amount of macrophages in theinflammatory infiltrate (D). Tissues are stainedwith hematoxylin and eosin. Arrows denote mes-enteric macrophages. Bar 50 �m.

Figure 8. Expression of CCK-2 receptor mRNA in rat mesen-tery. CCK-2 receptor mRNA is present in mesenteric tissues ofcontrol rats and peritoneal cell suspensions obtained fromzymosan-treated rats as analyzed by nested RT-PCR. Lane 1,100 bp MW marker; lane 2, cDNA of gastric corpus; lane 3,-RT control; lane 4, cDNA of mesenteric tissues; lane 5, -RTcontrol; lane 6, cDNA of peritoneal cells; lane 7, -RT control;lane 8, water. Specificity of reaction was assessed throughagarose gel electrophoresis (this figure), melting curve anal-ysis and sequenciation of the amplified product.


gastrointestinal tract. We observed increased leukocyterolling, adhesion, and emigration in the mesentericvenules of rats receiving the HP extract, which coin-cides with the high levels of proinflammatory cytokinesseen in plasma from animals following the same proto-col (33). It is important to note that daily treatmentwith the antagonist proglumide significantly reducedthe leukocyte-endothelial cell interactions induced bythe HP extract and the number of interstitial leuko-cytes. On this basis, it can be speculated that gastrin,released in response to HP-induced inflammation, aidsthe recruitment of leukocytes by activating local mac-rophages or granulocytes, which in turn would contrib-ute to the persistence of the inflammatory process. Infact, when this activity is impeded by proglumide, therat mesentery presents an altered pattern of infiltra-tion, with more macrophages and less PMN leukocytes,suggestive of a receding inflammatory process. It waspreviously observed in HP-infected patients a directrelationship between gastrinemia and the number ofmucosal PMN (16) and mononuclear (16, 17) leuko-cytes. At the time, and based on the observed stimula-tory effect of proinflammatory cytokines and activatedmonocytes on gastrin release from cultured antral Gcells (19–21), this correlation was explained as hyper-gastrinemia being the consequence of the inflamma-tory process. Our results complement this idea bysuggesting that gastrin released in response to localinflammation exerts a positive feedback on the inflam-matory process.

Gastrin stimulates gastric acid secretion not only bydirect activation of CCK-2 receptors on parietal cellsbut also through the release of histamine from entero-chromaffin-like cells within the gastric mucosa (1).Histamine is a mediator of acute inflammatory reac-tions, helping to recruit rolling leukocytes and increas-ing albumin leakage in postcapillary venules via H1receptor activation (51). The main sources of histaminein our experimental preparation are mastocytes, andthere is some evidence suggesting that activation of theCCK-2 receptor induces mast cell degranulation in theintestine (52). Our experiments, however, indicate that

histamine release is not involved in the proinflamma-tory action of gastrin since it was unaffected by stabili-zation of mastocytes with cromolyn or blockade of H1receptors with diphenhydramine.

The acid secretory effect of gastrin is physiologicallycounteracted by somatostatin, the dual action of whichdirectly inhibits parietal cells and reduces the release ofgastrin from G cells. HP decreases somatostatin synthe-sis (15, 53), which could partly explain the gastrinhypersecretion, and somatostatin displays anti-inflam-matory properties (54–57). Thus, gastrin proinflamma-tory activity may be limited by the physiological releaseof this peptide or enhanced when its secretion isreduced by HP infection. However, in the presentstudy, somatostatin did not alter in any way the basalleukocyte-endothelial cell interactions nor did it reducethe proinflammatory action of gastrin, which rules outa regulatory role for somatostatin in this process.

In summary, our results reveal that gastrin has aproinflammatory effect and suggest that local endo-crine G cells exposed to HP-induced inflammationsecrete more gastrin and contribute to the progressionof the focus. Further research is required to define therelevance of this effect of gastrin in other pathophysi-ological contexts.

The present study was awarded the 2002 Prize of SalvatFoundation from Salvat Laboratories and has been supportedby grants PI020461 and C03/02 from Ministerio de Sanidad yConsumo, SAF2004–06211 and SAF2005–01366 from Minis-terio de Educacion y Cultura and 05/043 from GeneralitatValenciana. The authors thank Dra. M. A. Ferrus (Grupo deInvestigacion sobre Helicobacter, Dpto. de Biotecnologıa, Uni-versidad Politecnica de Valencia) for preparing the HPextract.

REFERENCES

1. Walsh, J. H. (1994) Gastrin. In Gut Peptides: Biochemistry andPhysiology (Walsh, J. H., and Dockray, G. J., eds) pp. 75–121,Raven Press, New York

2. Dockray, G., Dimaline, R., and Varro, A. (2005) Gastrin: oldhormone, new functions. Pflugers Arch. 449, 344–355

Figure 9. Immunostaining for CCK-2 receptor inrat mesentery. Positive staining is evident inmacrophages (A) and PMN leukocytes (B). Nostaining was observed in the absence of primaryAb (not depicted). Insets in A correspond toimmunostaining for 2 different macrophagemarkers: A) monoclonal anti-macrophage �granulocyte Ab [OX41], Abcam; B) mAb againstan ED2-like antigen [HIS36], eBioscience).


3. Mezey, E., Palkovits, M. (1992) Localization of targets forantiulcer drugs in cells of the immune-system. Science 258,1662–1665

4. Sacerdote, P., Wiedermann, C. J., Wahl, L. M., Pert, C. B., andRuff, M. R. (1991) Visualization of cholecystokinin receptors ona subset of human monocytes and in rat spleen. Peptides 12,167–176

5. Schmitz, F., Schrader, H., Otte, J. M., Schmitz, H., Stuber, E.,Herzig, K. H., and Schmidt, W. E. (2001) Identification ofCCK-B/gastrin receptor splice variants in human peripheralblood mononuclear cells. Regul. Pept. 101, 25–33

6. Iwata, N., Murayama, T., Matsumori, Y., Ito, M., Nagata, A.,Taniguchi, T., Chihara, K., Matsuo, Y., Minowada, J., andMatsui, T. (1996) Autocrine loop through cholecystokinin-B/gastrin receptors involved in growth of human leukemia cells.Blood 88, 2683–2689

7. Schmitz, F., Otte, J. M., Stechele, H. U., Reimann, B., Ban-asiewicz, T., Folsch, U. R., Schmidt, W. E., and Herzig, K. H.(2001) CCK-B/gastrin receptors in human colorectal cancer.Eur. J. Clin. Invest. 31, 812–820

8. Cong, B., Li, S. J. H., Ling, Y. L., Yao, Y. X., Gu, Z. Y., Wang, J. X.,and You, H. Y. (2003) Expression and cell-specific localization ofcholecystokinin receptors in rat lung. World J. Gastroenterol. 9,1273–1277

9. DelaFuente, M., Carrasco, M., and Hernanz, A. (1997) Modula-tion of human neutrophil function in vitro by gastrin. J.Endocrinol. 153, 475–483

10. Sacerdote, P., Ruff, M. R., and Pert, C. B. (1988) Cholecystoki-nin and the immune system: receptor-mediated chemotaxis ofhuman and rat monocytes. Peptides 9, 29–34

11. Carrasco, M., Hernanz, A., and DelaFuente, M. (1997) Effect ofcholecystokinin and gastrin on human peripheral blood lym-phocyte functions, implication of cyclic AMP and interleukin 2.Regul. Pept. 70, 135–142

12. Kelly, S. M., Crampton, J. R., and Hunter, J. O. (1993) Helico-bacter-pylori increases gastric antral juxtamucosal pH. Dig. Dis.Sci. 38, 129–131

13. Lichtenberger, L. M., Dial, E. J., Romero, J. J., Lechago, J.,Jarboe, L. A., and Wolfe, M. M. (1995) Role of luminal ammoniain the development of gastropathy and hypergastrinemia in therat. Gastroenterology 108, 320–329

14. Dial, E. J., Hall, L. R., Romero, J. J., and Lichtenberger, L. M.(1996) Rats with gastritis have increased sensitivity to the gastrinstimulatory effects of luminal ammonia. Gastroenterology 110,801–808

15. Odum, L., Petersen, H. D., Andersen, I. B., Hansen, B. F., andRehfeld, J. F. (1994) Gastrin and somatostatin in Helicobacter-pylori infected antral mucosa. Gut 35, 615–618

16. Wagner, S., Haruma, K., Gladziwa, U., Soudah, B., Gebel, M.,Bleck, J., Schmidt, H., and Manns, M. (1994) Helicobacter-pylori infection and serum pepsinogen-A, pepsinogen-C, andgastrin in gastritis and peptic-ulcer: significance of inflamma-tion and effect of bacterial eradication. Am. J. Gastroenterol. 89,1211–1218

17. Graham, D. Y., Go, M. F., Lew, G. M., Genta, R. M., and Rehfeld,J. F. (1993) Helicobacter-pylori infection and exaggerated gas-trin-release: effects of inflammation and progastrin processing.Scand. J. Gastroenterol. 28, 690–694

18. Liu, Y., Vosmaer, G. D., Tytgat, G. N., Xiao, S. D., and Ten Kate,F. J. (2005) Gastrin (G) cells and somatostatin (D) cells inpatients with dyspeptic symptoms: helicobacter pylori associatedand non-associated gastritis. J. Clin. Pathol. 58, 927–931

19. Weigert, N., Schaffer, K., Schusdziarra, V., Classen, M., andSchepp, W. (1996) Gastrin secretion from primary cultures ofrabbit antral G cells: Stimulation by inflammatory cytokines.Gastroenterology 110, 147–154

20. Lehmann, F. S., Golodner, E. H., Wang, J. Y., Chen, M. C. Y.,Avedian, D., Calam, J., Walsh, J. H., Dubinett, S., and Soll, A. H.(1996) Mononuclear cells and cytokines stimulate gastrin re-lease from canine antral cells in primary culture. Am. J. Physiol.270, G783–G788

21. Beales, I., Blaser, M. J., Srinivasan, S., Calam, J., PerezPerez,G. I., Yamada, T., Scheiman, J., Post, L., and DelValle, J. (1997)Effect of Helicobacter pylori products and recombinant cyto-kines on gastrin release from cultured canine G cells. Gastroen-terology 113, 465–471

22. Alvarez, A., Cerda-Nicolas, M., Nabah, Y. N. A., Mata, M.,Issekutz, A. C., Panes, J., Lobb, R. R., and Sanz, M. J. (2004)Direct evidence of leukocyte adhesion in arterioles by angioten-sin II. Blood 104, 402–408

23. Alvarez, A., Hermenegildo, C., Issekutz, A. C., Esplugues, J. V.,and Sanz, M. J. (2002) Estrogens inhibit angiotensin II-inducedleukocyte-endothelial cell interactions in vivo via rapid endothe-lial nitric oxide synthase and cyclooxygenase activation. Circ.Res. 91, 1142–1150

24. Calatayud, S., Sanz, M. J., Canet, A., Bello, R., de Rojas, F. D.,and Esplugues, J. V. (1999) Mechanisms of gastroprotection bytransdermal nitroglycerin in the rat. Br. J. Pharmacol. 127,1111–1118

25. Gaboury, J. P., Johnston, B., Niu, X. F., and Kubes, P. (1995)Mechanisms underlying acute mast cell-induced leukocyte roll-ing and adhesion in vivo. J. Immunol. 154, 804–813

26. Doong, M. L., Lu, C. C., Kau, M. M., Tsai, S. C., Chiao, Y. C.,Chen, J. J., Yeh, J. Y., Lin, H., Huang, S. W., Chen, T. S., et al.(1998) Inhibition of gastric emptying and intestinal transit byamphetamine through a mechanism involving an increasedsecretion of CCK in male rats. Br. J. Pharmacol. 124, 1123–1130

27. Blandizzi, C., Lazzeri, G., Colucci, R., Carignani, D., Tognetti,M., Baschiera, F., and Del Tacca, M. (1999) CCK1 and CCK2receptors regulate gastric pepsinogen secretion. Eur. J. Pharma-col. 373, 71–84

28. Wang, L., Martinez, V., Barrachina, M. D., and Tache, Y. (1998)Fos expression in the brain induced by peripheral injection ofCCK or leptin plus CCK in fasted lean mice. Brain Res. 791,157–166

29. Glatzle, J., Wang, Y. H., Adelson, D. W., Kalogeris, T. J., Zittel,T. T., Tso, P., Wei, J. Y., and Raybould, H. E. (2003) Chylomi-cron components activate duodenal vagal afferents via a chole-cystokinin A receptor-mediated pathway to inhibit gastric motorfunction in the rat. J. Physiol. London 550, 657–664

30. Piqueras, L., Kubes, P., Alvarez, A., O’Connor, E., Issekutz, A. C.,Esplugues, J. V., and Sanz, M. J. (2000) Angiotensin II inducesleukocyte-endothelial cell interactions in vivo via AT(1) andAT(2) receptor-mediated P-selectin upregulation. Circulation102, 2118–2123

31. Mathison, R., and Davison, J. S. (1995) Regulation of jejunalarterioles by capsaicin-sensitive nerves in Nippostrongylus bra-siliensis-sensitized rats. J. Pharmacol. Exp. Ther. 273, 337–343

32. Decktor, D. L., Pendleton, R. G., Kellner, A. T., and Davis, M. A.(1989) Acute effects of ranitidine, famotidine and omeprazoleon plasma gastrin in the rat. J. Pharmacol. Exp. Ther. 249, 1–5

33. Brzozowski, T., Konturek, P. C., Konturek, S. J., Kwiecien, S.,Pajdo, R., Karczewska, E., Stachura, J., and Hahn, E. G. (1999)Water extracts of Helicobacter pylori delay healing of chronicgastric ulcers in rats: Role of cytokines and gastrin-somatostatinlink. Digestion 60, 22–33

34. Mai, U. E. H., PerezPerez, G. I., Wahl, L. M., Wahl, S. M., Blaser,M. J., and Smith, P. D. (1991) Soluble surface-proteins fromHelicobacter-pylori activate monocytes macrophages by lipopo-lysaccharide-independent mechanism. J. Clin. Invest. 87, 894–900

35. Haribabu, B., Verghese, M. W., Steeber, D. A., Sellars, D. D.,Bock, C. B., and Snyderman, R. (2000) Targeted disruption ofthe leukotriene B(4) receptor in mice reveals its role in inflam-mation and platelet-activating factor-induced anaphylaxis. J.Exp. Med. 192, 433–438

36. Morisset, J., Wong, H., Walsh, J. H., Laine, J., and Bourassa, J.(2000) Pancreatic CCK(B) receptors: their potential roles insomatostatin release and delta-cell proliferation. Am. J. Physiol.Gastrointest. Liver Physiol. 279, G148–G156

37. Wank, S. A. (1998) G protein-coupled receptors in gastrointes-tinal physiology. I. CCK receptors: an exemplary family. Am. J.Physiol. 274, G607–G613

38. Rozengurt, E., and Walsh, J. H. (2001) Gastrin, CCK, signaling,and cancer. Annu. Rev. Physiol. 63, 49–76

39. Niggli, V. (2003) Signaling to migration in neutrophils: impor-tance of localized pathways. Int. J. Biochem. Cell Biol. 35, 1619–1638

40. Ley, K. (2002) Integration of inflammatory signals by rollingneutrophils. Immunol. Rev. 186, 8–18

41. Dib, K., andAndersson, T. (2000) Beta 2 Integrin signaling inleukocytes. Front Biosci. 5, D438–D451

42. Jones, G. E. (2000) Cellular signaling in macrophage migrationand chemotaxis. J. Leuk. Biol. 68, 593–602


43. DelaFuente, M., Drummond, J., DelRio, M., Carrasco, M., andHernanz, A. (1996) Modulation of murine peritoneal macro-phage functions by gastrin. Peptides 17, 219–224

44. Engstrand, L. (1995) Potential animal-models of Helicobacter-pylori infection in immunological and vaccine research. FEMSImmunol. Med. Microbiol. 10, 265–270

45. Yoshida, N., Granger, D. N., Evans, D. J., Evans, D. G., Graham,D. Y., Anderson, D. C., Wolf, R. E., and Kvietys, P. R. (1993)Mechanisms involved in Helicobacter-pylori-induced inflamma-tion. Gastroenterology 105, 1431–1440

46. Elizalde, I., Gomez, J., Panes, J., Lozano, M., Casadevall, M.,Ramirez, J., Pizcueta, P., Marco, F., deRojas, F. D., Granger,D. N. et al. (1997) Platelet activation in mice and humanHelicobacter pylori infection. J. Clin. Invest. 100, 996–1005

47. Kurose, I., Granger, D. N., Evans, D. J., Evans, D. G., Graham,D. Y., Miyasaka, M., Anderson, D. C., Wolf, R. E., Cepinskas, G.,and Kvietys, P. R. (1994) Helicobacter pylori-induced microvas-cular protein leakage in rats: role of neutrophils, mast-cells, andplatelets. Gastroenterology 107, 70–79

48. Mai, U. E. H., PerezPerez, G. I., Allen, J. B., Wahl, S. M.,Blaser, M. J., and Smith, P. D. (1992) Surface-proteins fromHelicobacter-pylori exhibit chemotactic activity for human-leukocytes and are present in gastric-mucosa. J. Exp. Med. 175,517–525

49. Nielsen, H., Andersen, L. P. (1992) Chemotactic activity ofHelicobacter-pylori sonicate for human polymorphonuclearleukocytes and monocytes. Gut 33, 738–742

50. Craig, P. M., Territo, M. C., Karnes, W. E., and Walsh, J. H.(1992) Helicobacter-pylori secretes a chemotactic factor formonocytes and neutrophils. Gut 33, 1020–1023

51. Asako, H., Kurose, I., Wolf, R., Defrees, S., Zheng, Z. L., Phillips,M. L., Paulson, J. C., and Granger, D. N. (1994) Role of H1receptors and P-selectin in histamine-induced leukocyte rollingand adhesion in postcapillary venules. J. Clin. Invest. 93, 1508–1515

52. Juanola, C., Giralt, M., Jimenez, M., Mourelle, M., and Vergara,P. (1998) Mucosal mast cells are involved in CCK disruption ofMMC in the rat intestine. Am. J. Physiol. Gastrointest Liver Physiol.38, G63–G67

53. Moss, S. F., Legon, S., Bishop, A. E., Polak, J. M., and Calam, J.(1992) Effect of Helicobacter-pylori on gastric somatostatin induodenal-ulcer disease. Lancet 340, 930–932

54. Karalis, K., Mastorakos, G., Chrousos, G. P., and Tolis, G. (1994)Somatostatin analogs suppress the inflammatory reaction in-vivo. J. Clin. Invest. 93, 2000–2006

55. Scheiman, J. M., Tillner, A., Pohl, T., Oldenburg, A., Angermul-ler, S., Gorlach, E., Engel, G., Usadel, K. H., and Kusterer, K.(1997) Reduction of non-steroidal anti-inflammatory drug in-duced gastric injury and leucocyte endothelial adhesion byoctreotide. Gut 40, 720–725

56. Karalis, K., Mastorakos, G., Sano, H., Wilder, R. L., andChrousos, G. P. (1995) Somatostatin may participate in theantiinflammatory actions of glucocorticoids. Endocrinology136, 4133– 4138

57. Helyes, Z., Pinter, E., Nemeth, J., Keri, G., Than, M., Oroszi, G.,Horvath, A., and Szolcsanyi, J. (2001) Anti-inflammatory effectof synthetic somatostatin analogues in the rat. Br. J. Pharmacol.134, 1571–1579

Received for publication January 31, 2006.Accepted for publication June 2, 2006.


The FASEB Journal • FJ Express Summary

Gastrin induces leukocyte-endothelial cell interactionsin vivo and contributes to the inflammation causedby Helicobacter pylori

Angeles Alvarez,* Sales Ibiza,† Carlos Hernandez,* Alberto Alvarez-Barrientos,‡

Juan V. Esplugues,*,1and Sara Calatayud**Departamento de Farmacologıa and †Unidad Mixta CNIC-UVEG, Facultad de Medicina,Universidad de Valencia, Valencia, Spain; and ‡Cytometry Unit, Fundacion Centro Nacional deInvestigaciones Cardiovasculares, Madrid, Spain

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-5696fje

SPECIFIC AIMS

Infection by Helicobacter pylori (HP), the main cause ofnonautoimmune chronic gastritis, increases gastrin se-cretion, and a correlation between gastrinemia and theseverity of gastritis has been identified. Gastrin recep-tors have been observed in several leukocytes, andgastrin and cholecystokinin (CCK) have been shown toinduce changes in leukocyte functions such as chemo-taxis, adherence, or phagocitosis in vitro. However, ithas not been analyzed whether these cellular effects ofgastrin have any consequence on the functioning of theimmune system in vivo. If this was the case, its increasedconcentration in the HP-infected gastric mucosa mightaffect the evolution of the inflammatory process in-duced primarily by the microorganism, i.e., gastritis. Inthe present study, we have evaluated this possibility inrats administered chronically with an extract of aCagA�/VacA� strain of HP, which mimics the HPinfection in humans.

PRINCIPAL FINDINGS

1. Gastrin and related peptides induce leukocyte-endothelial cell interactions

Gastrin superfusion induced a concentration- and time-dependent increase in leukocyte-endothelial cell inter-actions in the mesenteric venules as evaluated by intra-vital microscopy in anesthetized rats. Gastrin increasedthe number of rolling leukocytes, while decreased theirrolling velocity. This effect was rapid in onset (15 min)and was observed with all the gastrin concentrationsused (0.01-1nM). This response was followed by signif-icant increases in the number of adherent leukocytesand leukocytes emigrated into the interstitial tissue.Similar effects were induced by superfusion with pen-tagastrin (1nM) or CCK (1-10nM).

2. Effects of gastrin and CCK on leukocyte-endothelial cell interactions are mediated throughCCK-2 receptors

Pretreatment with proglumide (30 mg/kg, i.p.) pre-vented both the increase in leukocyte rolling, adhesion,and emigration and the reduction in rolling velocityinduced by gastrin (1nM). These effects of gastrin werealso inhibited by pretreatment with another CCK-2receptor antagonist L-365,260 (1 mg/kg, i.v.) but unaf-fected by pretreatment with the CCK-1 antagonistdevazepide (1 mg/kg, i.v.). In a similar way, the in-crease in leukocyte-endothelial cell interactions in-duced by CCK (10nM) was prevented by proglumideand not by devazepide.

Endogenous release of gastrin was stimulated by anintravenous infusion of gastrin-releasing peptide (300pmol/kg/h) or by a single maximal dose of the antise-cretory drug omeprazole (40 mg/kg, p.o.), and bothtreatments induced significant increases in leukocyte-endothelial cell interactions, which were also preventedby pretreatment with proglumide.

3. Analysis of the expression of the CCK-2 receptor

CCK-2 receptor mRNA, analyzed by nested reversetranscriptase-polymerase chain reaction, was detectedin both mesenteric tissues of control rats and peritonealcell suspensions of zymosan-treated rats. Immunohisto-chemical studies revealed the presence of CCK-2 recep-tor in two cellular types of mesenteric tissue, whichwere identified as polymorphonuclear leukocytes andmacrophages. Endothelial cells tested negative for theCCK-2 receptor immunostaining.

1 Correspondence: Departamento de Farmacologıa, Facultadde Medicina, Universidad de Valencia, Avd. Blasco Ibanez 15,46010-Valencia, Spain. E-mail: [email protected]

doi: 10.1096/fj.05-5696fje

2396 0892-6638/06/0020-2396 © FASEB

4. Effects of gastrin on leukocyte-endothelial cellinteractions are not mediated by histamine normodulated by somatostatin

The increase in leukocyte-endothelial cell interactionsinduced by gastrin (1nM) was not modified by either anantagonist of the H1 receptor for histamine diphenhy-dramine (2 �M cosuperfused) or stabilization of mastcells with cromolyn (20 mg/kg, i.v. plus cosuperfusionwith 0.33 mg/ml). The effects of gastrin (1nM) werealso unaltered by the addition of somatostatin (1nM) tothe superfusion buffer.

5. HP extract causes mesenteric inflammation. Effectsof a CCK-2 receptor antagonist

The extract of HP was prepared from a strain CagA�/VacA� (NCTC 11638) and was administered on days 1,3, 5, and 7 (1 ml/rat, p.o.). Leukocyte-endothelial cellinteractions were evaluated on day 14. Rats receivingthe HP extract showed an active inflammatory responsein the mesenteric tissue as denoted by a higher numberof rolling leukocytes moving at a slower rate, an in-creased leukocyte adhesion, and a significant increasein the number of leukocytes emigrated into the inter-stitium surrounding the selected venules (Fig. 1). Anal-ysis of the eosin and hematoxylin-stained mesenteric

tissue also revealed the presence of an inflammatoryinfiltrate in HP-treated animals (322�20 leukocytes perfield in HP-treated rats vs. 44�2 leukocytes per field incontrol animals, P�0.001), consisting mainly of poly-morphonuclear leukocytes (79�1%), whereas macro-phages and lymphocytes amounted to 15 � 1 and 6 �1% of the total, respectively.

Chronic administration of the aqueous extract of HPalso caused hypergastrinemia (39�7 vs. 19�1 pM gas-trin in control rats, P�0.05), and daily treatment ofthese animals with the CCK-2 receptor antagonist pro-glumide significantly reduced the number of rolling,adherent, and emigrated leukocytes (Fig. 1). Eosin andhematoxylin staining revealed that rats receiving pro-glumide also presented an inflammatory infiltrate butwith fewer leukocytes (244�12 vs. 322�20 leukocytesper field, P�0.01) and differing proportions of eachleukocyte type. In contrast to what was observed inHP-treated rats (see previous paragraph), the infiltratein rats receiving proglumide contained similar num-bers of polymorphonuclears and macrophages (44�3and 50�4%, respectively).

CONCLUSIONS AND SIGNIFICANCE

The present study demonstrates for the first time thatgastrin has a proinflammatory action in vivo and pointsto a role for gastrin hypersecretion in the inflammatoryreaction caused by HP.

Gastrin enhances the interactions between theflowing leukocytes and the venular endothelium inrat mesentery, increasing the number of rolling,adherent, and emigrated leukocytes. These are theinitial steps conducting to the formation of inflam-matory foci and, thus, our results indicate that gas-trin possess a proinflammatory activity. Similar re-sponses were obtained with pentagastrin and CCKand when the endogenous release of gastrin wasstimulated by the neuropeptide GRP or the potentantisecretory drug omeprazole. Gastrin and CCKmediate their effects on target tissues by activatingthe G protein-coupled CCK-1 and CCK-2 receptors.Their proinflammatory activity seems to be a CCK-2receptor-mediated process since it can be reversed byits specific antagonists proglumide and L-365,260 butnot by the CCK-1 antagonist devazepide. The CCK-2receptor mRNA was detectable in rat mesenterictissue extracts, and immunohistochemistry studieslocalized those receptors in mesenteric macrophagesand polymorphonuclear leukocytes.

Gastrin partly stimulates gastric acid secretionthrough the release of histamine from enterochromaf-fin-like cells. Histamine is a mediator of acute inflam-matory reactions, and there is some evidence suggest-ing that activation of the CCK-2 receptor maydegranulate mast cells, an important source of hista-mine in the mesentery. However, our experimentsindicate that histamine release is not involved in theproinflammatory action of gastrin, since it was unaf-

Figure 1. Mesenteric inflammation induced by HP extractand effect of a CCK-2 receptor antagonist. Rats were treatedchronically (days 1, 3, 5, and 7) with an aqueous extract of HPor saline, and the effects on leukocyte rolling flux (A), rollingvelocity (B), adhesion (C), and emigration (D) were analyzedin mesenteric postcapillary venules on day 14. Animals weredivided into 4 groups: saline � vehicle of proglumide (n�7),saline � proglumide (n�5), HP extract � vehicle of proglu-mide (n�5), and HP extract � proglumide (n�7). Resultsare mean � sem. *P � 0.05 and ***P � 0.001 vs. saline �vehicle of proglumide group; �P � 0.05 and ��P � 0.01 vs.HP extract � vehicle of proglumide group.

2397GASTRIN AND INFLAMMATION BY HELICOBACTER PYLORI

fected by stabilization of mastocytes with cromolyn orblockade of H1 receptors with diphenhydramine. Theacid secretory effect of gastrin is physiologically coun-teracted by somatostatin, and some anti-inflammatoryproperties have been attributed to this peptide. Never-theless, somatostatin did not modify the effects ofgastrin, which rules out a regulatory role for somatosta-tin in its proinflammatory action.

In the HP-infected mucosa, significant gastrin levelsand infiltrated leukocytes can be found and, consider-ing the effects observed with gastrin, a modulatoryeffect of the peptide on local leukocytes functionshould be expected. We have administered an extractobtained from bacterial suspensions containing manyof the pathogenic components of the germ for 1 wk andthe experiments were performed 7 days later, whenincreased gastrin plasma levels were detectable. At thisstage, there was still an active inflammatory process inthe gastrointestinal tract. We observed increased leuko-cyte rolling, adhesion, and emigration in the mesen-teric venules of rats receiving the HP extract (Fig. 1),while high levels of proinflammatory cytokines inplasma were seen in previous studies following the sameprotocol. It is important to note that daily treatmentwith the antagonist proglumide significantly reducedthe leukocyte-endothelial cell interactions induced bythe HP extract and the number of interstitial leuko-cytes. On this basis, it can be speculated that gastrin,

released in response to HP-induced inflammation, aidsthe recruitment of leukocytes by activating local mac-rophages or granulocytes, which in turn would contrib-ute to the persistence of the inflammatory process. Infact, when this activity is impeded by proglumide, therat mesentery presents an altered pattern of infiltra-tion, with more macrophages and less polymorphonu-clear leukocytes, suggestive of a receding inflammatoryprocess. It was previously observed in HP-infected pa-tients a direct relationship between gastrinemia and thenumber of mucosal polymorphonuclear and mononu-clear leukocytes. At the time, and based on the ob-served stimulatory effect of proinflammatory cytokinesand activated monocytes on gastrin release from cul-tured antral G cells, this correlation was explained ashypergastrinemia being the consequence of the inflam-matory process. Our results complement this idea bysuggesting that gastrin released in response to localinflammation exerts a positive feedback on the inflam-matory process (Fig. 2).

In summary, our results reveal that gastrin has aproinflammatory effect and suggest that local endo-crine G cells exposed to HP-induced inflammationsecrete more gastrin and contribute to the progressionof the focus. Further research is required to define therelevance of this effect of gastrin in other pathophysi-ological contexts.

Figure 2. Infection by Helicobacter pyloricauses gastrin hypersecretion while re-duces the secretion of somatostatin. Sev-eral mechanisms have been proposed forthese hormonal changes: 1) rebound re-sponse to the elevated pH that accompa-nies the infection; 2) effect of the ammo-nia generated by bacterial urease ongastrin-secreting G cells; 3) direct effectsof bacterial components on endocrinecells; 4) stimulatory effect of pro-inflam-matory cytokines released from infil-trated leukocytes on G cells. Consideringthe proinflammatory effect of gastrin, thepresence of CCK-2 receptors in polymor-phonuclear leukocytes and macrophages,and the inhibitory effect of a CCK-2 re-ceptor antagonist on the inflammationinduced by H. pylori observed in the present study, we postulate that gastrin, released in response to HP-induced inflammation,aids the recruitment of leukocytes by activating local macrophages or granulocytes and contributes to the persistence of theinflammatory process. Thus, our theory is that gastrin released in response to local inflammation exerts a positive feedback onthe H. pylori induced gastritis. HP, Helicobacter pylori; EpC, epithelial cells; PMN, polymorphonuclear leukocytes; MØ,macrophages; EC, endothelial cells. Continuous lines denote positive influences; dotted lines denote negative influences.

2398 Vol. 20 November 2006 ALVAREZ ET AL.The FASEB Journal

Gastrin induces leukocyte-endothelial cell interactions in vivo and contributes to the inflammation...

Documents

Transcript of Gastrin induces leukocyte-endothelial cell interactions in vivo and contributes to the inflammation...