ORIGINAL ARTICLE

Neutrophil gelatinase-associated lipocalin (NGAL/Lcn2)is upregulated in gastric mucosa infectedwith Helicobacter pylori

Warner Alpízar-Alpízar & Ole Didrik Laerum & Martin Illemann & José A. Ramírez &

Adriana Arias & Wendy Malespín-Bendaña & Vanessa Ramírez & Leif R. Lund &

Niels Borregaard & Boye Schnack Nielsen

Received: 22 June 2009 /Revised: 13 August 2009 /Accepted: 14 August 2009 /Published online: 1 September 2009# Springer-Verlag 2009

Abstract Helicobacter pylori infection is one of the mostsignificant risk factors for gastric cancer. The infection isestablished early in life and persists lifelong leading to asustained chronic inflammation. Iron is essential for mostliving organisms. Bacteria use several mechanisms toacquire iron from their hosts, including the synthesis ofthe potent iron chelators known as siderophores. Hosts cellsmay express the siderophore-binding protein neutrophilgelatinase-associated lipocalin (NGAL/lipocalin-2 (Lcn2))in response to infection, thus preventing bacterial ironuptake. We have characterized here the pattern of expres-sion of NGAL/Lcn2 in gastric mucosa (45 non-neoplasticand 38 neoplastic tissue samples) and explored the connec-tion between NGAL/Lcn2 expression and H. pylori

infection. Immunohistochemical analysis showed highNGAL/Lcn2 expression in normal and gastritis-affectedmucosa compared to low expression in intestinal meta-plasia, dysplasia, and gastric cancer. In normal and gastritis-affected mucosa (n=36 tissue samples), NGAL/Lcn2 wasmore frequently seen in epithelial cells located at the neckand base of the glands in H. pylori-positive cases than insimilar epithelial cells of noninfected cases (Fisher’s exacttest, p=0.04). In conclusion, the high expression of NGAL/Lcn2 in normal and gastritis-affected mucosa infected withH. pylori suggests that NGAL/Lcn2 is upregulated locallyin response to this bacterial infection. It is discussed whetherthis may have a causal relation to the development of gastriccancer.

W. Alpízar-Alpízar :O. D. LaerumThe Gade Institute,University of Bergen and Department of Pathology,Haukeland University Hospital,Bergen, Norway

W. Alpízar-Alpízar :W. Malespín-Bendaña :V. RamírezCancer Research Program, Health Research Institute (INISA),University of Costa Rica,San José, Costa Rica

W. Alpízar-Alpízar :M. Illemann :B. S. NielsenThe Finsen Laboratory,Rigshospitalet,Copenhagen, Denmark

J. A. Ramírez :A. AriasDepartment of Pathology,Dr. Rafael A. Calderón Guardia Hospital,San José, Costa Rica

L. R. LundDepartment of Biology, Section for Celland Developmental Biology, University of Copenhagen,Copenhagen, Denmark

N. BorregaardThe Granulocyte Research Laboratory,Department of Haematology, University of Copenhagen,Rigshospitalet,Copenhagen, Denmark

W. Alpízar-Alpízar (*)The Gade Institute, Department of Pathology,Haukeland University Hospital,5021 Bergen, Norwaye-mail: awarnercr@yahoo.come-mail: Warner.Alpizar@student.uib.no

Present Address:B. S. NielsenExiqon A/S, Diagnostic Product Development,Vedbæk, Denmark

Virchows Arch (2009) 455:225–233DOI 10.1007/s00428-009-0825-8

Keywords NGAL/Lcn2 .Helicobacter pylori .

Inflammation . Gastritis . Intestinal metaplasia .

Gastric cancer

AbbreviationsIM Intestinal metaplasiamAb Monoclonal antibodypAbs Polyclonal antibodiesTNP Trinitrophenyl haptenNGAL Neutrophil gelatinase-associated lipocalinLcn2 Lipocalin-2

Introduction

Gastric cancer is the second most common cause of cancerdeaths worldwide [1]. It is the final result of a multistepprocess initiated by environmental factors, including dietand Helicobacter pylori infection [2, 3]. H. pylori infectionis one of the most important risk factors for this malignancy[4, 5]. The infection is usually established early in life andpersists lifelong in the absence of treatment [6]. Theinfection leads to a sustained chronic inflammation charac-terized by infiltration of a number of inflammatory cells inthe gastric mucosa and expression of inflammatory mediatorsby immune and epithelial cells [7]. It is the combination ofbacterial factors, host immune response, and environmentalinsults that drives the stepwise transformation starting withmucosal atrophy through metaplasia and dysplasia to overtgastric cancer [8, 9].

Iron is an essential element for a number of metabolicprocesses in almost all living organisms. During infection,bacteria utilize several mechanisms to obtain iron from theirhosts, including synthesis of so-called siderophores, whichchelate Fe3+ with high affinity and facilitate its transportinto the pathogen [10–12]. An important host defensemechanism, which interferes with bacterial uptake of iron hasbeen unraveled by the discovery of neutrophil gelatinase-associated lipocalin (NGAL or lipocalin-2 (Lcn2)) [13, 14] orNGAL-homologous proteins (24p3 in mouse and NRL inrats) [15] as siderophore-binding proteins [16]. NGAL/Lcn2binds some bacterial siderophores and prevents their uptakeinto bacteria [17]. NGAL/Lcn2 expression has been shownto be upregulated in response to inflammation in epithelialcells of several mucosal surfaces including the gastrointes-tinal tract and the lower respiratory tract [18–20]. Elevatedexpression of the NGAL-homologous proteins has also beenfound in animal models (mice and rats) in response tobacterial infections [21, 22]. The significance of this wasdemonstrated in a mouse model, where NGAL-deficientmice challenged with a clinical strain of Escherichia coli(H9049 strain) had a substantial increase in bacteremia and

bacterial burden in some organs, including liver and spleenand a strikingly decreased survival compared to the wild-type mice [23], and also in response to pulmonaryKlebsiella pneumoniae infection [22].

Iron is fundamental for the pathogenesis of H. pylori.The ability of H. pylori to acquire iron is consideredimportant for colonization, persistence, and virulence ofthis microorganism in the gastric mucosa [24–27]. Expres-sion of NGAL/Lcn2 in the stomach has been reported bothat the protein and mRNA level [14, 28]. In a mouse model,immunoreactivity for the NGAL-homologous protein, 24p3,was observed in epithelial, endothelial, and infiltratinginflammatory cells in the stomach in response to variousforms of gastrointestinal injury [29]. Recently, increasedexpression of NGAL/Lcn2 was reported in rhesus macaqueschallenged with H. pylori [30]. These observations suggestthat NGAL/Lcn2 is expressed in gastric mucosa in responseto inflammation/infection. Here, we study the expression ofNGAL/Lcn2 in human gastric mucosa and show for thefirst time that NGAL/Lcn2 is upregulated in response toH. pylori infection and inflammation.

Material and methods

Tissue samples

Formalin-fixed and paraffin-embedded tissue samples wereobtained from patients undergoing surgery for gastriccancer in two hospitals in Costa Rica (Max Peralta Hospitaland Rafael Angel Calderón Guardia Hospital) and one inBergen, Norway (Haukeland University Hospital). Thesamples encompassed non-neoplastic mucosa adjacent to themalignant tissue (n=45; 37 from Costa Rica and eight fromNorway) and neoplastic lesions (n=38; 32 from Costa Ricaand six from Norway). Three to four micrometer thickparaffin-embedded tissue sections were cut from the non-neoplastic material, stained with H&E, and evaluated by apathologist (ODL). Among the 45 non-neoplastic sections,there were foci characterized as normal mucosa (with somedegree of inflammation; n=6), gastritis (n=30), intestinalmetaplasia (IM; n=17), and dysplasia (n=11). Informationregarding the histopathological subtype of gastric cancer wascollected for all 38 neoplastic lesions. The histopathologicalclassification was according to Laurén’s classification system(Norwegian cases) and Japanese classification system(Costa Rican cases). For the purposes of this study, CostaRican cases were reclassified according to Laurén’s classi-fication system following established criteria given by theJapanese Gastric Cancer Association [31]. Of the 38 cases,20 were classified as intestinal subtype and 18 as diffusesubtype, and they were all non-cardia cancers (from corpusand antrum regions of the stomach). The study was

226 Virchows Arch (2009) 455:225–233

approved by the ethics committees and institutional reviewboards of each institution (Costa Rica: VI 742-94-571, VI742-99-340; Norway: REK 053228) and performed inaccordance with the World Medical Association Declara-tion of Helsinki, 1996.

Antibodies

Affinity purified mouse monoclonal antibody (mAb)against human NGAL/Lcn2 (clone 211.1) has been de-scribed previously [32]. Rat mAb against human NGAL/Lcn2 was purchased from R&D systems (Minneapolis,USA). mAbs against cytokeratins (CKs, clones AE1/AE3),neutrophil elastase (clone NP57), and rabbit polyclonalantibodies (pAbs) against H. pylori (code no. B0471),fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG and nonimmune rabbit IgG were purchasedfrom Dako (Glostrup, Denmark). Cy3-conjugated goatanti-rabbit was obtained from Jackson Immunoresearch(West Grove, PA, USA). Monoclonal antibody directedagainst trinitrophenyl hapten (TNP, IgG1) was previouslydescribed [33].

Immunoperoxidase staining

Three micrometer thick paraffin-embedded tissue sectionswere deparaffinized with xylene and hydrated in gradualseries of ethanol–water dilutions. For NGAL immunohis-tochemistry with mouse and rat mAbs, sections were heat-treated in a T/T micromed microwave processor (Milestone,Sorisol, Italy) at 98°C for 30 min in target retrieval solutionpH 6,0 (code no. S1699, Dako). For immunohistochemistrywith antibodies against neutrophil elastase, CKs, H. pylori,and TNP, sections were pretreated with proteinase K(10 μg/mL) for 20–25 min at 37°C. Endogenous peroxidaseactivity was blocked by incubation in 1% hydrogenperoxide solution for 15 min and washed briefly in TrisBuffered Saline (TBS; 50 mM Tris, 150 mM NaCl, pH 7.6)containing 0.5% triton X-100. The primary antibodies werediluted in antibody diluent (Dako) and incubated for 2 h inShandon racks (Thermo Shandon, Pittsburg, PA, USA) at thefollowing dilutions: NGAL mouse mAb 1:200 (5.0 μg/mL),NGAL rat mAb 1:350 (1.4 μg/mL), Neutrophil Elastase1:200, CKs 1:300, and anti-H. pylori 1:150. Subsequently,the primary antibodies were detected with EnVision reagentusing either anti-mouse IgG or anti-rabbit IgG horseradishperoxidase-conjugated polymers (Dako) and polyclonalrabbit anti-rat immunoglobulins/HRP (code no. P0450,Dako). The reactions were visualized by incubating thesections with NovaRED (Vector Laboratories, BurlingameCA, USA) or DAB chromogen (Dako; for H. pylori pAbs)according to manufacturer’s instructions and counterstainedwith Mayer´s haematoxylin.

Negative controls

The sections were pretreated in the same way as describedabove for the NGAL antibody. The mouse NGAL mono-clonal antibody was substituted with anti-TNP mAb incu-bated at the same concentrations as that for NGAL.

Double immunofluorescence staining

The sections were processed as described above for theNGAL mouse mAb using heat-induced retrieval in targetretrieval solution pH 6.0 (code no. S1699, Dako). The anti-NGAL mouse mAb (1:150, 6.7 μg/mL) was diluted inantibody diluent (Dako) together with anti-H. Pylori pAbs(1:150) and incubated in the tissue sections for 2 h at roomtemperature. The antibodies were subsequently detectedwith FITC-conjugated goat-anti-mouse IgG, 1:200, andCy3-conjugated goat anti-rabbit IgG, 1:200, respectively.After brief rinses with TBS, the sections were mountedwith ProLong Gold antifade (Molecular Probes, Eugene,Oregon).

Confocal microscopy

The double stained sections were analyzed using a confocallaser-scanning microscope, LSM 510 META (Carl Zeiss,Jena, Germany), equipped with a 488 nm argon laser and a543 nm HeNe1 laser, as previously described [34].

Scoring for the immunoperoxidase stainings

The tissue sections stained with NGAL mouse mAb fromboth non-neoplastic gastric mucosa adjacent to cancertissue and from gastric cancer lesions were evaluated bytwo independent investigators (WAA and ODL). Neutro-phils present in the microvasculature and within the tissueserved as internal positive control for NGAL/Lcn2 expres-sion. NGAL/Lcn2 immunoreactivity was scored, accordingto the estimated percentage of crypts showing NGAL/Lcn2-positive epithelial cells, in the following four types of non-neoplastic gastric mucosa: normal mucosa, gastritis, IM,and dysplasia. The scoring was as follows: 0, less than 10%positive crypts; 1+, between 10% and 30% positive crypts;2+, between 30% and 60% positive crypts; and 3+, morethan 60% positive crypts. In gastric cancer lesions, NGAL/Lcn2 immunoreactivity was scored based on the estimatedpercentage of positive cancer cells seen in the wholesection. Thus, the percentages of positive cancer cells weregrouped into the following categories: 0, less than 5%NGAL/Lcn2-positive cancer cells detected; 1, between 5%and 30% positively stained cells; 2, between 30% and 60%positive tumor cells; and 3, when more than 60% of cancercells were positive.

Virchows Arch (2009) 455:225–233 227

H. pylori positivity was scored based on the density ofbacteria and the number of crypts containing bacteria intothe following categories: −, no evidence of H. pylori on thesection; +, less than three crypts with small clusters ofbacteria; ++, either less than three crypts with dense clustersof bacteria or more than three crypts with small clusters;and +++, three or more crypts with dense clusters ofbacteria.

Statistical analysis

χ2 analysis was performed to evaluate the differencesregarding the frequency of cases having NGAL/Lcn2-positive cancer cells in intestinal versus diffuse subtypesof gastric cancer. χ2 analysis and Fisher’s exact test wereused to assess the correlation between NGAL/Lcn2expression in gastric epithelial cells and H. pylori infectionin normal-like and gastritis-affected mucosa. p≤0.05 wasconsidered statistically significant in all cases.

Results

NGAL/Lcn2 expression in non-neoplastic and neoplasticgastric mucosa

Immunoperoxidase staining for NGAL/Lcn2 in non-neo-plastic and neoplastic tissue sections demonstrated NGAL/Lcn2 staining in all the specimens studied. We separatelycharacterized the pattern of expression of NGAL/Lcn2 foreach of the different states of the gastric mucosa: normalmucosa, gastritis, IM, dysplasia, and cancer. In general, theNGAL/Lcn2-positive cells observed in the non-neoplasticand neoplastic tissue samples included neutrophils, epithelialcells, and cancer cells. The NGAL/Lcn2 immunoreactivity ofthe normal mucosa (with some degree of inflammation) andgastritis-affected mucosa was intense (high frequency ofsections scored as 3+), while that of IM, dysplasia, and cancerwas low (high frequency of the sections scored as 0–1+;Table 1). This suggests that the expression of NGAL/Lcn2 isupregulated locally in the early steps of the gastric carcino-genesis that are characterized by chronic inflammation ofthe gastric mucosa.

The pattern of expression of NGAL/Lcn2 varied sub-stantially according to the level of affection of the gastricmucosa. In normal and gastritis affected mucosa, NGAL/Lcn2 immunoreactivity was mainly seen in epithelial cells(Fig. 1a–d). NGAL/Lcn2 was, however, differentiallyexpressed depending on the location of the epithelial cellsin the crypts, with the most intense signal being observed inepithelial cells located at the neck and base of the cryptsand weak or negative immunoreactivity at the surfaceepithelial cells (Fig. 1d). The expression of NGAL/Lcn2

was absent at foci with IM, where scattered NGAL/Lcn2-positive neutrophils were seen in the lamina propria(Fig. 1c–e). In dysplastic mucosa, NGAL/Lcn2 immuno-reactivity was seen in epithelial cells of the glands locateddeeper into the gastric mucosa (Fig. 1f). In gastric cancerlesions, NGAL/Lcn2 staining was seen in 53% (20 out of38) of the cases in cancer cells widespread within themalignant growth, with similar frequency in intestinal(Fig. 1g) and diffuse (Fig. 1h) histological subtypes (Table 1;p=0.84; χ2=0.04). NGAL/Lcn2 immunoreactivity in neo-plasia was heterogeneous, being intense in some areas of thetumor growth and weak and/or negative in other areas (notshown).

The specificity of NGAL/Lcn2 immunoreactivity wasverified by analysis of a series of positive and negativecontrols. Similar NGAL/Lcn2 staining was obtained withmouse and rat mAbs against NGAL/Lcn2 in both non-neoplastic (Fig. 1i, j) and neoplastic tissue (not shown) inall of 15 cases tested. As negative control, we substitutedthe NGAL mouse mAb with a mAb against TNP. Nospecific staining was obtained with this antibody whenincubated at similar immunoglobulin concentration as therespective anti-NGAL antibody preparation (Fig. 1k, l).

H. pylori infection associates with the inductionof NGAL/Lcn2 in gastric mucosa

We observed, as described above, a clear and consistentNGAL/Lcn2 immunoreactivity in epithelial cells of normalmucosa and from gastritis (Table 1, Fig. 2b). Since it is knownthat H. pylori infection induces a chronic inflammatory

Table 1 NGAL/Lcn2 immunoreactivity in non-neoplastic and neo-plastic tissue sections according to the different states of the gastricmucosa and histological subtypes of gastric cancer

Lesion n NGAL/Lcn2 scoringa

0 1+ 2+ 3+

Normal-appearingb 6 2 0 0 4

Gastritisb 30 3 1 1 25

IMc 17 15 2 0 0

Dysplasia 11 4 5 1 1

Cancerd 38 18 11 5 4

Intestinal 20 10 5 2 3

Diffuse 18 8 6 3 1

NGAL neutrophil gelatinase-associated lipocalin, Lcn2 lipocalin-2,IM intestinal metaplasiaa Scoring method is described in Material and methods sectionb Scoring according to Helicobacter pylori status is shown in Table 2cχ2 analysis for normal and gastritis vs intestinal metaplasia anddysplasia, p<0.0001dχ2 analysis for intestinal versus diffuse, p=0.84

228 Virchows Arch (2009) 455:225–233

response in gastric mucosa [35] and that NGAL/Lcn2expression is upregulated in response to bacterial infectionand inflammation [36], we explored the possibility that theexpression of NGAL/Lcn2 in epithelial cells of the gastricmucosa is associated with H. pylori infection. Tissue samplesfrom non-neoplastic mucosa, with foci of normal and gastritisaffected mucosa, were selected and stained for H. pylori.

Immunoperoxidase staining showed H. pylori in 72% (26 outof 36) of the cases. H. pylori clusters were observed at theluminal edge of the crypts, in deeper areas of the crypts, andinside some of the mucosal glands, with the highest densityof bacteria in gastritis-affected mucosa (Fig. 2c). We nextscored the NGAL/Lcn2 immunoreactivity in epithelial cellsof the normal and gastritis-affected mucosa from adjacent

Fig. 1 Immunoperoxidase staining for NGAL/Lcn2 in non-neoplasticand neoplastic gastric tissue. Paraffin sections from non-neoplastic (a–f, i–l) and neoplastic (g–h) gastric mucosa were incubated with mousemAb against NGAL/Lcn2 (a–i, k), rat mAb against NGAL/Lcn2 (j),and a mAb against TNP (l). Panels i–j and k–l represent adjacentsections. In gastritis-affected mucosa (a–d), NGAL/lcn2 is expressedby epithelial cells of gastric mucosa and scattered neutrophils in thelamina propia (b; neutrophils (arrows)). NGAL/Lcn2 is seen inepithelial cells located at the neck and base of the crypts moreevidently than in surface epithelial cells (d). NGAL/Lcn2 expressionis absent in intestinal metaplasia (c–e), next to gastritis foci showingintense immunoreactivity (c and d; goblet cells (Gc)), where only

scattered NGAL/Lcn2-positive neutrophils are observed in the laminapropia (e; arrows). In dysplasia (f), NGAL/Lcn2 is observed inepithelial cells of glands located deep into the mucosa. In the gastriccancer lesions, NGAL/Lcn2 immunoreactivity is observed in cancercells that are widespread within the tumor growth in both intestinal (g;stroma (St), cancer (Ca)) and diffuse subtype (h). A similar stainingpattern is seen with mouse mAb (i) and rat mAb (j). As negativecontrol, the mAb against NGAL/Lcn2 (k) was substituted with a mAbagainst TNP (i), and no specific staining is seen. Sections werecounterstained with hematoxylin. Scale bars: ∼100 μm (a, c, i, j),∼50 μm (d–h, k, l), ∼25 μm (b)

Virchows Arch (2009) 455:225–233 229

tissue sections, as described in Material and methods, andnoticed that the majority of the H. pylori-positive cases hadan elevated expression of NGAL/Lcn2 (scores 2–3+) ascompared to H. pylori-negative cases (Table 2). It has beenshown that NGAL/Lcn2 is expressed at a basal low level in anumber of human epithelial tissues such as the respiratorytract, the gastrointestinal and urinary tracts that are exposed tothe external environment and thereby prone to infections[14, 18, 28], indicating that a low expression level is to beexpected in mucosal surfaces. We, therefore, dichotomizedthe expression of NGAL/Lcn2 into NGAL/Lcn2 low immu-noreactivity (scores 0, 1+) and NGAL/Lcn2 high immunore-activity (scores 2+, 3+; Table 2) for a statistical analysis.Using this classification, we found a significant associationbetween the expression of NGAL/Lcn2 in epithelial cells andH. pylori infection (p=0.02; χ2=5.43; df=1; Fisher exact testp=0.04). Double immunofluorescence analysis for NGAL/Lcn2 combined with H. pylori in non-neoplastic normal andgastritis-affected mucosa from 24 of the gastric cancerpatients (18 H. pylori-positive and 6 H. pylori-negative),indeed showed intense and consistent NGAL/Lcn2 immuno-reactivity in crypts where epithelial cells were in directcontact with H. pylori bacteria in the H. pylori-positive cases(Fig. 2d, f). In contrast, in H. pylori-negative cases, theNGAL/Lcn2 signal was weak in epithelial cells at similarlocations (not shown).

Discussion

We have investigated the expression of NGAL/Lcn2 ineach of the pre-neoplastic stages leading to gastric cancer,focusing on the potential induction of this protein inresponse to H. pylori infection. By performing immunohis-tochemistry and immunofluorescence analysis, we showthat NGAL/Lcn2 is upregulated in epithelial cells of normaland gastritis-affected mucosa infected with H. pylori. Theseare typical sites of robust inflammatory activity [7].Interestingly, we also observed that the expression ofNGAL/Lcn2 is absent at foci with IM adjacent to areas ofgastritis. At IM, H. pylori infection is spontaneouslyeradicated and inflammation is partially resolved [2, 37].Our results, therefore, support the role of NGAL/Lcn2 as acomponent of the innate immunity and suggest that H.pylori infection is causatively associated with upregulationof this molecule in gastric mucosa.

In the present study, we found a significant higherNGAL/Lcn2 immunoreactivity in gastric epithelial cells ofH. pylori-infected individuals than in those non-infected,which was substantiated by observing H. pylori clusters inclose proximity to NGAL/Lcn2-positive epithelial cells inour double immunofluorescence analysis. NGAL/Lcn2 hasbeen shown to be constitutively expressed in epithelial cells

of a number of tissues that are often challenged bymicroorganisms such as the respiratory tract and the urinaryand gastrointestinal tracts [14, 28]. NGAL/Lcn2 mRNAand protein levels are greatly increased in response topneumococcal and mycobacterial infections in the respira-tory tract and urinary tract infections, in both animal modelsand humans [22, 36, 38, 39]. Taking together these obser-vations and our results, we suggest that the constitutivebasal expression of NGAL/Lcn2 that is observed in thegastric epithelium ([14, 28] and this paper) is substantiallyupregulated upon H. pylori infection. A recent study inrhesus macaques, showing that NGAL/Lcn2 is induced ingastric mucosa of monkeys challenged with certain H. pyloristrains [30], is in agreement with our observations andsuggests that the upregulation of NGAL/Lcn2 by H. pylorimay be related to specific bacterial virulence factors.

A particularly interesting finding of this study was theabrupt down-regulation of NGAL/Lcn2 expression ob-served at IM foci located next to normal and gastritisH. pylori-infected mucosa showing intense NGAL/Lcn2immunoreactivity. A number of histological and physio-logical changes occur in the gastric epithelium during thepre-neoplastic stages leading to gastric cancer, including afocal loss of glands and specialized cells (atrophic gastritis),followed by the transformation of the stomach mucosa intoan intestinal-like epithelium (IM) [8, 40]. These changes arenot well tolerated byH. pylori and results in its spontaneousdisappearance from the transformed epithelium [37, 41].Accordingly, we did not observe H. pylori bacteria at intes-tinal metaplastic glands. It is, therefore, tempting to suggestthat the abolition of NGAL/Lcn2 expression on IM isa consequence of the natural eradication of H. pylorioccurring at these foci. Thus, our findings in IM argue infavor of the potential link between H. pylori infection andthe upregulation of NGAL/Lcn2 in the gastric epithelium.

Iron is an essential element for the growth and virulenceof H. pylori, and the general consensus is that H. pyloriobtains this element from human molecules, mainlylactoferrin and heme [27, 42–44]. Despite one studysuggesting the production of siderophores by H. pylori[45], most investigations have been unable to reproducethat finding or to demonstrate the use of exogenoussiderophores as a source of iron by this bacterium [44, 46].Nevertheless, the induction of NGAL/Lcn2 is not restrictedto microorganisms that produce or uptake siderophores.

The induction of NGAL/Lcn2 has been consistentlylinked to inflammation. High NGAL/Lcn2 mRNA andprotein levels have been found in epithelial cells in inflam-matory colorectal diseases such as inflammatory boweldisease, appendicitis [18], as well as in other processesassociated with intense inflammatory activity [20, 47]. Aspart of the inflammatory response against H. pylori infection,a complex set of inflammatory mediators are released by

230 Virchows Arch (2009) 455:225–233

epithelial and infiltrating inflammatory cells into the gastricmucosa, including IL-1β and NF-κβ [48]. Certain H. pylorivirulence factors have been shown to potentiate thisinflammatory response [49]. It has been demonstrated thatIL-1β and NF-κβ (stabilized by its cofactor IκB-ζ) play acentral role in the induction of NGAL/Lcn2 in epithelial cells[19, 50]. Thus, it is likely that the upregulation of NGAL/

Lcn2 by H. pylori in gastric epithelial cells is driven bymediators of inflammation being released into the gastricepithelium in response to the bacterial infection.

It was recently demonstrated in animal models (rhesusmacaques and mice) that pathogenic strains of Salmonellaenterica induce NGAL/Lcn2 in colonic mucosa, whichconfers a competitive advantage to the bacterium forcolonizing and growing in the inflamed intestine [51]. Inagreement to this view, Hornsby et al. [30] hypothesizedthat the induction of a set of antimicrobial molecules,including NGAL/Lcn2, in rhesus monkeys infected withH. pylori may indeed serve to increase the competitiveadvantage of this bacterium. We, therefore, do not excludethe possibility that high expression of NGAL/Lcn2 ingastric mucosa may facilitate H. pylori infection.

Other functions have been suggested for NGAL/Lcn2 inaddition to its antimicrobial role [52]. NGAL/Lcn2 hasbeen shown to be expressed on neoplastic cells in severaltypes of adenocarcinomas including breast, pancreas, colon,and ovarian cancer [28, 53–56], and elevated protein levelshave been correlated with poor prognosis in some of thesemalignancies [56, 57]. In the current study, we observed

Fig. 2 Immunohistochemical analyses of NGAL/Lcn2 and H. pyloriin non-neoplastic gastric mucosa. Sections from samples with gastricmucosa adjacent to the neoplastic lesion were processed forimmunoperoxidase staining for NGAL/Lcn2 (a, b), H. pylori (c), orfor immunofluorescence analysis (d–f) using mAb against NGAL/Lcn2, pAbs against H. pylori, and NGAL mAb together with H. pyloripAbs, respectively. NGAL/Lcn2 immunoreactivity is weak, if observed,inH. pylori-negative mucosa (a). In H. pylori-infected mucosa, NGAL/Lcn2 is strongly expressed on epithelial cells (b). H. pylori clusters (redstars) can be observed at the luminal side in some of the crypts along

the gastric epithelium with the highest density in gastritis-affectedmucosa (c). In the double immunofluorescence analysis, mAb againstNGAL/Lcn2 together with pAbs against H. pylori (double staining ind–f) were detected with FITC-conjugated goat antimouse (greenfluorescence) and Cy3-conjugated goat antirabbit (red fluorescence),respectively. Intense NGAL/Lcn2 immunoreactivity is seen in epithe-lial cells that are in direct contact with H. pylori bacteria (d-f).Nomarsky DIC imaging technique was used in d–f. Scale bars:∼100 μm (a, b), ∼12 μm (c), ∼20 μm (d–f)

Table 2 NGAL/Lcn2 immunoreactivity in epithelial cells in relationto the Helicobacter pylori status in non-neoplastic tissue sectionsdiagnosed as normal-appearing and gastritis-affected mucosa

n NGAL/Lcn2 scoring

0 1+ 2+ 3+

H. pylori negativea 10 3 1 1 5

H. pylori positivea 26 2 0 0 24

Total 36 5 1 1 29

NGAL neutrophil gelatinase-associated lipocalin, Lcn2 lipocalin-2a Fisher’s exact test for H. pylori-positive versus H. pylori-negative,p=0.04

Virchows Arch (2009) 455:225–233 231

NGAL/Lcn2 immunoreactivity in gastric cancer cells in onehalf of the studied cases, with similar frequency in bothintestinal and diffuse gastric cancer histological subtypes. Arecent study found that NGAL/Lcn2 levels in tissue andurine from breast cancer patients are associated with invasionand metastasis and demonstrated, in cell lines transfectedwith NGAL/Lcn2, that this molecule promotes breast cancerprogression by inducing epithelial to mesenchymal transi-tion [58]. The molecular mechanisms explaining the role ofNGAL/Lcn2 in carcinogenesis remain entirely unknown,but it has been speculated that expression of NGAL/Lcn2provides the carcinoma cells the ability to acquire ironand thus, supporting their growth [52, 59]. Therefore, theexpression of NGAL/Lcn2 in gastric cancer in connectionto progression and prognosis of the disease should befurther investigated.

Acknowledgements We thank Ms. Öznur Turan and Ms. AgnieszkaIngvorsen for their excellent technical assistance and Mr. John Post forhis photographic assistance. This study was supported by the DanishCancer Society, the Lundbeck Foundation, Haukeland UniversityHospital (Helse-Vest), and Vicerrectoría de Investigación of theUniversity of Costa Rica.

Conflict of interest statement We declare that we have no conflictof interest.

References

1. Parkin DM, Bray FI, Devesa SS (2001) Cancer burden in the year2000. The global picture. Eur J Cancer 37(Suppl 8):S4–S66

2. Stemmermann GN, Fenoglio-Preiser C (2002) Gastric carcinomadistal to the cardia: a review of the epidemiological pathology ofthe precusors to a preventable cancer. Pathology 34:494–503

3. Correa P (1992) Human gastric carcinogenesis: a multistep andmultifactorial process—First American Cancer Society AwardLecture on cancer epidemiology and prevention. Cancer Res52:6735–6740

4. Uemura N, Okamoto S, Yamamoto S, Matsumura N, YamaguchiS et al (2001) Helicobacter pylori infection and the developmentof gastric cancer. N Engl J Med 345:784–789

5. Huang JQ, Zheng GF, Sumanac K, Irvine EJ, Hunt RH (2003)Meta-analysis of the relationship between cagA seropositivity andgastric cancer. Gastroenterology 125:1636–1644

6. Wilson KT, Crabtree JE (2007) Immunology of Helicobacterpylori: insights into the failure of the immune response andperspectives on vaccine studies. Gastroenterology 133:288–308

7. Robinson K, Argent RH, Atherton JC (2007) The inflammatoryand immune response to Helicobacter pylori infection. Best PractRes Clin Gastroenterol 21:237–259

8. Correa P, Houghton J (2007) Carcinogenesis of Helicobacterpylori. Gastroenterology 133:659–672

9. Fox JG, Wang TC (2007) Inflammation, atrophy, and gastriccancer. J Clin Invest 117:60–69

10. Neilands JB (1995) Siderophores: structure and function of microbialiron transport compounds. J Biol Chem 270:26723–26726

11. Braun V, Braun M (2002) Active transport of iron and siderophoreantibiotics. Curr Opin Microbiol 5:194–201

12. Krewulak KD, Vogel HJ (2008) Structural biology of bacterialiron uptake. Biochim Biophys Acta 1778:1781–1804

13. Kjeldsen L, Johnsen AH, Sengelov H, Borregaard N (1993)Isolation and primary structure of NGAL, a novel proteinassociated with human neutrophil gelatinase. J Biol Chem268:10425–10432

14. Cowland JB, Borregaard N (1997) Molecular characterizationand pattern of tissue expression of the gene for neutrophilgelatinase-associated lipocalin from humans. Genomics 45:17–23

15. Kjeldsen L, Cowland JB, Borregaard N (2000) Human neutrophilgelatinase-associated lipocalin and homologous proteins in rat andmouse. Biochim Biophys Acta 1482:272–283

16. Goetz DH, Holmes MA, Borregaard N, Bluhm ME, Raymond KNet al (2002) The neutrophil lipocalin NGAL is a bacteriostaticagent that interferes with siderophore-mediated iron acquisition.Mol Cell 10:1033–1043

17. Holmes MA, Paulsene W, Jide X, Ratledge C, Strong RK (2005)Siderocalin (Lcn 2) also binds carboxymycobactins, potentiallydefending against mycobacterial infections through iron seques-tration. Structure 13:29–41

18. Nielsen BS, Borregaard N, Bundgaard JR, Timshel S, Sehested Met al (1996) Induction of NGAL synthesis in epithelial cells ofhuman colorectal neoplasia and inflammatory bowel diseases. Gut38:414–420

19. Cowland JB, Sorensen OE, Sehested M, Borregaard N (2003)Neutrophil gelatinase-associated lipocalin is up-regulated inhuman epithelial cells by IL-1 beta, but not by TNF-alpha. JImmunol 171:6630–6639

20. Sorensen OE, Cowland JB, Theilgaard-Monch K, Liu L, Ganz Tet al (2003) Wound healing and expression of antimicrobialpeptides/polypeptides in human keratinocytes, a consequence ofcommon growth factors. J Immunol 170:5583–5589

21. Berger T, Togawa A, Duncan GS, Elia AJ, You-Ten A et al(2006) Lipocalin 2-deficient mice exhibit increased sensitivity toEscherichia coli infection but not to ischemia-reperfusion injury.Proc Natl Acad Sci U S A 103:1834–1839

22. Chan YR, Liu JS, Pociask DA, Zheng M, Mietzner TA et al(2009) Lipocalin 2 is required for pulmonary host defense againstKlebsiella infection. J Immunol 182:4947–4956

23. Flo TH, Smith KD, Sato S, Rodriguez DJ, Holmes MA et al(2004) Lipocalin 2 mediates an innate immune response tobacterial infection by sequestrating iron. Nature 432:917–921

24. van Amsterdam K, van Vliet AH, Kusters JG, van der Ende A(2006) Of microbe and man: determinants of Helicobacter pylori-related diseases. FEMS Microbiol Rev 30:131–156

25. Velayudhan J, Hughes NJ, McColm AA, Bagshaw J, Clayton CLet al (2000) Iron acquisition and virulence in Helicobacter pylori:a major role for FeoB, a high-affinity ferrous iron transporter. MolMicrobiol 37:274–286

26. Waidner B, Greiner S, Odenbreit S, Kavermann H, Velayudhan Jet al (2002) Essential role of ferritin Pfr in Helicobacter pyloriiron metabolism and gastric colonization. Infect Immun 70:3923–3929

27. Perez-Perez GI, Israel DA (2000) Role of iron in Helicobacterpylori: its influence in outer membrane protein expression and inpathogenicity. Eur J Gastroenterol Hepatol 12:1263–1265

28. Friedl A, Stoesz SP, Buckley P, Gould MN (1999) Neutrophilgelatinase-associated lipocalin in normal and neoplastic humantissues. Cell type-specific pattern of expression. Histochem J31:433–441

29. Playford RJ, Belo A, Poulsom R, Fitzgerald AJ, Harris K et al(2006) Effects of mouse and human lipocalin homologues 24p3/lcn2 and neutrophil gelatinase-associated lipocalin on gastroin-testinal mucosal integrity and repair. Gastroenterology 131:809–817

30. Hornsby MJ, Huff JL, Kays RJ, Canfield DR, Bevins CL et al(2008) Helicobacter pylori induces an antimicrobial response in

232 Virchows Arch (2009) 455:225–233

rhesus macaques in a cag pathogenicity island-dependent manner.Gastroenterology 134:1049–1057

31. Japanese Gastric Cancer A (1998) Japanese classification ofgastric carcinoma - 2nd english edition. Gastric Cancer 1:10–24

32. Kjeldsen L, Koch C, Arnljots K, Borregaard N (1996) Charac-terization of two ELISAs for NGAL, a newly described lipocalinin human neutrophils. J Immunol Methods 198:155–164

33. Boulianne GL, Hozumi N, Shulman MJ (1984) Production offunctional chimaeric mouse/human antibody. Nature 312:643–646

34. Nielsen BS, Rank F, Illemann M, Lund LR, Dano K (2007)Stromal cells associated with early invasive foci in humanmammary ductal carcinoma in situ coexpress urokinase andurokinase receptor. Int J Cancer 120:2086–2095

35. Atherton JC (2006) The pathogenesis of Helicobacter pylori-induced gastro-duodenal diseases. Annu Rev Pathol 1:63–96

36. Xu S, Venge P (2000) Lipocalins as biochemical markers ofdisease. Biochim Biophys Acta 1482:298–307

37. Valle J, Kekki M, Sipponen P, Ihamaki T, Siurala M (1996) Long-term course and consequences of Helicobacter pylori gastritis.Results of a 32-year follow-up study. Scand J Gastroenterol 31:546–550

38. Nelson AL, Barasch JM, Bunte RM, Weiser JN (2005) Bacterialcolonization of nasal mucosa induces expression of siderocalin, aniron-sequestering component of innate immunity. Cell Microbiol7:1404–1417

39. Saiga H, Nishimura J, Kuwata H, Okuyama M, Matsumoto S et al(2008) Lipocalin 2-dependent inhibition of mycobacterial growthin alveolar epithelium. J Immunol 181:8521–8527

40. Gutierrez-Gonzalez L, Wright NA (2008) Biology of intestinalmetaplasia in 2008: more than a simple phenotypic alteration. DigLiver Dis 40:510–522

41. Kokkola A, Kosunen TU, Puolakkainen P, Sipponen P, HarkonenM et al (2003) Spontaneous disappearance of Helicobacter pyloriantibodies in patients with advanced atrophic corpus gastritis.APMIS 111:619–624

42. Dhaenens L, Szczebara F, Husson MO (1997) Identification,characterization, and immunogenicity of the lactoferrin-bindingprotein from Helicobacter pylori. Infect Immun 65:514–518

43. Worst DJ, Otto BR, de Graaff J (1995) Iron-repressible outermembrane proteins of Helicobacter pylori involved in heme uptake.Infect Immun 63:4161–4165

44. Husson MO, Legrand D, Spik G, Leclerc H (1993) Iron acquisitionby Helicobacter pylori: importance of human lactoferrin. InfectImmun 61:2694–2697

45. Illingworth DS, Walter KS, Griffiths PL, Barclay R (1993)Siderophore production and iron-regulated envelope proteins ofHelicobacter pylori. Zentralbl Bakteriol 280:113–119

46. Dhaenens L, Szczebara F, Van Nieuwenhuyse S, Husson MO(1999) Comparison of iron uptake in different Helicobacterspecies. Res Microbiol 150:475–481

47. Sorensen OE, Thapa DR, Roupe KM, Valore EV, Sjobring U et al(2006) Injury-induced innate immune response in human skinmediated by transactivation of the epidermal growth factorreceptor. J Clin Invest 116:1878–1885

48. Algood HM, Cover TL (2006) Helicobacter pylori persistence: anoverview of interactions between H. pylori and host immunedefenses. Clin Microbiol Rev 19:597–613

49. Brandt S, Kwok T, Hartig R, Konig W, Backert S (2005) NF-kappaB activation and potentiation of proinflammatory responsesby the Helicobacter pylori CagA protein. Proc Natl Acad Sci U SA 102:9300–9305

50. Cowland JB, Muta T, Borregaard N (2006) IL-1beta-specific up-regulation of neutrophil gelatinase-associated lipocalin is con-trolled by IkappaB-zeta. J Immunol 176:5559–5566

51. RaffatelluM, GeorgeMD, AkiyamaY, HornsbyMJ, Nuccio SP et al(2009) Lipocalin-2 resistance confers an advantage to Salmonellaenterica serotype Typhimurium for growth and survival in theinflamed intestine. Cell Host Microbe 5:476–486

52. CliftonMC, Corrent C, Strong RK (2009) Siderocalins: siderophore-binding proteins of the innate immune system. Biometals 22:557–564

53. Stoesz SP, Friedl A, Haag JD, Lindstrom MJ, Clark GM et al(1998) Heterogeneous expression of the lipocalin NGAL inprimary breast cancers. Int J Cancer 79:565–572

54. Furutani M, Arii S, Mizumoto M, Kato M, Imamura M (1998)Identification of a neutrophil gelatinase-associated lipocalinmRNA in human pancreatic cancers using a modified signalsequence trap method. Cancer Lett 122:209–214

55. Lim R, Ahmed N, Borregaard N, Riley C, Wafai R et al (2007)Neutrophil gelatinase-associated lipocalin (NGAL) an early-screening biomarker for ovarian cancer: NGAL is associated withepidermal growth factor-induced epithelio-mesenchymal transi-tion. Int J Cancer 120:2426–2434

56. Bauer M, Eickhoff JC, Gould MN, Mundhenke C, Maass N et al(2008) Neutrophil gelatinase-associated lipocalin (NGAL) is apredictor of poor prognosis in human primary breast cancer.Breast Cancer Res Treat 108:389–397

57. Devarajan P (2007) Neutrophil gelatinase-associated lipocalin:new paths for an old shuttle. Cancer Ther 5:463–470

58. Yang J, Bielenberg DR, Rodig SJ, Doiron R, Clifton MC et al(2009) Lipocalin 2 promotes breast cancer progression. Proc NatlAcad Sci U S A 106:3913–3918

59. Devireddy LR, Gazin C, Zhu X, Green MR (2005) A cell-surfacereceptor for lipocalin 24p3 selectively mediates apoptosis and ironuptake. Cell 123:1293–1305

Virchows Arch (2009) 455:225–233 233