Microbiol Immunol 2008; 52: 197–208doi:10.1111/j.1348.0421.2008.00030.x

ORIGINAL ARTICLE

Nitric oxide produced in Peyer’s patches exhibitsantiapoptotic activity contributing to an antimicrobialeffect in murine salmonellosisMohammad S. Alam∗, Mohammad H. Zaki∗, Tomohiro Sawa, Sabrina Islam, Khandaker A. Ahmed,Shigemoto Fujii, Tatsuya Okamoto and Takaaki Akaike

Department of Microbiology, Graduate School of Medical Sciences, Kumamoto University, 1-1-1 Honjo, Kumamoto 860-8556, Japan

CorrespondenceTakaaki Akaike, Department of Microbiology,Graduate School of Medical Sciences,Kumamoto University, 1-1-1 Honjo,Kumamoto 860-8556, Japan.Tel: +81 96 373 5100;fax: +81 96 362 8362;email: takakaik@gpo.kumamoto-u.ac.jp

∗Equal contribution.

Received 10 December 2007; accepted 22January 2008

List of Abbreviations: APC,antigen-presenting cells; CFU, colony formingunits; FAE, follicle-associated epithelium; FITC,fluorescein isothiocyanate; iNOS, inducibleisoform of nitric oxide synthase; iNOS+/−,heterozygous iNOS-deficient mice; iNOS−/−,homozygous iNOS-deficient mice; i.p.,intraperitoneal; MHC, major histocompatibilitycomplex; NO, nitric oxide; p.o., oral; PBS,0.01 M sodium phosphate-buffered 0.15 Msaline; PP, Peyer’s patches; SED, subepithelialdome; SipB, salmonella invasion protein;TUNEL, terminal deoxynucleotide transferase(TdT)-mediated dUTP-biotin nick end-labeling

Key wordsantiapoptosis, nitric oxide, Peyer’s patches,salmonellosis.

ABSTRACTSalmonella species normally infect hosts via the oral-fecal route. We pre-viously reported that NO had potent host defense functions in murinesalmonellosis, not only via a direct antibacterial effect but also because it wascytoprotective for infected host cells. Here, we used an oral route to infectiNOS-deficient mice infected with S. enterica serovar Typhimurium to fur-ther investigate the cytoprotective role of NO in preventing damage causedby Salmonella organisms in PP. Oral bacterial challenge (2 × 105 CFU, or>100 LD50) produced a more severe infection and greater lethality in iNOS-deficient mice than in iNOS-competent mice. We used specific antibodies toS. enterica Typhimurium, neutrophils, iNOS, nitrotyrosine, and dendriticcells (CD11c-positive) in histochemical and immunohistochemical studiesto examine infected PP tissues. S. enterica Typhimurium colonization in PPfrom iNOS-deficient mice was significantly higher than that in wild-typemice. Histochemical assays showed extensive cellular damage in PP. We thenexamined PP tissues for apoptosis by means of in situ TUNEL analysis andby measuring caspase-3 specific activity in tissue homogenates. Increasednumbers of TUNEL-positive cells and severe granulomatous inflammationwith increased infiltration of neutrophils and macrophages were observedduring infection in iNOS-deficient mice compared with wild-type mice.iNOS-deficient mice had increased numbers of dendritic cells and signifi-cantly higher caspase-3-specific activity in PP. These data confirm that NOexerts its protective function not only through direct antibacterial action,but also by preventing apoptosis and thereby contributing to antimicrobialdefense during salmonellosis.

Salmonella species cause various disease syndromes, rang-ing from gastroenteritis to systemic infections such as ty-phoid fever. S. enterica serovar Typhimurium, a majorcause of bacterial gastroenteritis, can also produce murinesalmonellosis, which is characterized by progressive for-mation of multiple microabscesses and septicemia (1, 2).After Salmonella organisms are ingested, they reach thesmall intestine where they invade the mucosa by cross-

ing the epithelial barrier through M cells overlying lym-phoid follicles of PP (3, 4). PP contain a dome, underlyingfollicles (B cell zone with germinal centers), and parafol-licular regions enriched with T cells. Among otherimportant cells distributed in PP are macrophages anddendritic (CD11c) cells, both of which essentially workas MHC class II+ APC (5). Salmonella invasion re-sults in characteristic tissue changes, including infiltration

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 197

M. S. Alam et al.

by neutrophils and monocytes, which contribute toerosion of the intestinal mucosa (3). After the bacte-ria cross the epithelial barrier, they encounter tissuemacrophages and other host cells in the lamina pro-pria. The bacteria replicate in PP and then enter thelymphatic system and bloodstream and spread to deepertissues (6).

NO generated by iNOS has been demonstrated tohave a beneficial effect on host defense against variouspathogenic bacteria and protozoa (1, 2, 7–10). iNOS ex-pression has been assumed to produce antimicrobial ac-tivity as a result of formation of reactive nitrogen oxidesderived from NO (11–15). For example, peroxynitrite, apotent oxidant formed from NO and a superoxide radical,is microbicidal for various bacteria including Salmonella(1, 2, 7, 11–13); and nitrosothiols, thiol adducts of one-electron oxidized derivatives of NO, have potent bacterio-static activity against Salmonella (11, 12, 14, 16).

Salmonella-infected host cells such as macrophages,hepatocytes, and intestinal epithelial cells undergo apop-tosis during infection, which in turn may accelerate bacte-rial invasion and dissemination and lead to septicemia(17–20). Several studies have shown that endogenousiNOS expression or exposure to low doses of exogenousNO donors can inhibit apoptosis in human B lymphocytesin vivo and in in vitro cultures of splenocytes, eosinophils,endothelial cells, macrophages, hepatocytes, and cell lines(21). Strong antiapoptotic activity has been documentedfor NO and especially for its derivative nitrosothiols, thepossible mechanism being inhibition of a cascade of intra-cellular cysteine proteases known as caspases, which mayinvolve S-nitrosylation of the cysteine in the active site ofthe enzymes (21–23). Indeed, we have found that iNOS-derived NO can function as an antiapoptotic moleculeand thus protect host cells and macrophages from severedamage caused by bacterial invasion (1, 2, 7).

We reported earlier that pharmacological inhibition ofeither NO or superoxide production results in markedenhancement of Salmonella growth and increased mor-tality in murine salmonellosis, which suggests that bothNO and superoxide contribute in critical ways to host de-fenses against Salmonella (2). Other previous studies withiNOS-deficient mice have clearly shown the pivotal roleof NO in macrophage defense against Salmonella (24). Inaddition, we have reported that iNOS-deficient mice aremore susceptible to Salmonella infection, and have pro-posed that NO has significant host defense functions notonly via its direct antibacterial effects but also via cyto-protection of infected host cells (1, 7).

In most earlier Salmonella studies, experiments in-volved the i.p. route or in vitro work with macrophages.The p.o. route (the natural route of infection) has rarelybeen employed, although interferon-γ-induced apopto-

sis has previously been reported to occur in T cells inPP after p.o. infection with Toxoplasma gondii (25). Ourpresent studies investigated the outcome of p.o. infectionwith Salmonella, because this route mimics natural infec-tion in humans. Also, to obtain a clear understanding ofthe pathogenesis of Salmonella infections, including ty-phoid fever, it is important to establish NO-dependentantimicrobial and protective mechanisms in the host atthe first host defense barrier, such as PP in the intestine.For these investigations, we used iNOS-deficient and wild-type mice infected p.o. with a virulent S. enterica serovarTyphimurium strain. We report here that lack of iNOScaused increased growth of Salmonella within PP, and ledto increased apoptosis in cells in PP, which in turn pro-duced severe systemic infection.

MATERIALS AND METHODS

Animals

Male wild-type C57BL/6 (B6) and C57BL/6-NOS2tm1Lau

(iNOS−/−) (26) mice, 5 or 7 weeks old, were produced byJackson Laboratories (West Grove, PA, USA). Heterozy-gous iNOS-deficient mice (iNOS+/−) were produced bymating homozygous iNOS−/− mice with their wild-typecounterparts (iNOS+/+) in our laboratory, in a facility atthe Center for Animal Resources and Development, Ku-mamoto University. iNOS−/− littermates bred from thesame iNOS+/− parents were used throughout the studies.All experiments were carried out according to the ethicalguidelines in the Laboratory Protocol of Animal Handling,Kumamoto University.

Bacteria and culture medium

S. enterica serovar Typhimurium, a virulent Gifu 12142strain (provided by Prof. Takayuki Ezaki, Gifu University),had been isolated from a patient with septicemia caused byserovar Typhimurium. Bacteria were grown overnight inbrain-heart infusion broth (Difco Laboratories, Detroit,MI, USA) at 37oC for routine culture.

Salmonella infection in mice

S. enterica serovar Typhimurium Gifu 12142 strain in0.2 ml of PBS at pH 7.4 was administered p.o. to themice. A control group of mice were mock-inoculatedwith PBS. Mice were fasted for 24 hr before the oral bac-terial challenge. At various time points after infectionthe mice were killed and the PP tissues removed. Thesmall intestine was separated and freed from surround-ing mesentery. In a routine preliminary washing proce-dure, 10 ml of PBS was passed through the small intestine.

198 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd

Nitric oxide and murine salmonellosis

PP tissues were excised from the duodenum or ileumwall in such a way that they remained free of intestinalcontents.

Bacterial count with PP tissues

The number of bacteria present in the PP tissues wasquantified using the colony formation assay as describedearlier (7). Briefly, all PP tissues collected from a singlemouse were weighed and homogenized in ice-cold PBS us-ing a Polytron homogenizer (Kinematica GmbH, Lucerne,Switzerland). The resultant homogenate was then seriallydiluted and subjected to colony formation assay on brain-heart infusion agar for 18 hr at 37oC.

Histological examination

PP were fixed with 10% buffered neutral formalin solu-tion, embedded in paraffin, and cut into 3-μm sections.Sections were stained with hematoxylin-eosin.

Immunohistochemistry

For immunohistochemical analysis, tissues were processedin 2% periodate-lysine-paraformaldehyde fixative at 4oCfor 4 hr. After 12 hr of successive washings with PBS con-taining 10%, 15%, and 20% sucrose, tissues were em-bedded in Tissue-Tek OCT compound (Miles, Elkhart,IN, USA), frozen in dry ice-acetone, and kept at −80oCuntil use. Sections 6 μm thick were prepared with a cryo-stat, and cryosections were air-dried overnight. Sectionswere then stained by means of the indirect immunoper-oxidase method (1), with a specific antibody for iNOS(1:100; Santa Cruz Biotechnology, Santa Cruz, CA, USA),anti-neutrophil monoclonal antibody (1:100; Serotec,Raleigh, NC, USA), or a polyclonal anti-nitrotyrosineantibody (1:500; Upstate Biotechnology, Lake Placid,NY, USA) as primary antibody. Tissue-bound peroxi-dase activity was visualized via reaction with the sub-strate 3,3′-diaminobenzidine; hematoxylin was used fornuclear staining. PP tissues were also stained with ham-ster anti-mouse CD11c antibody (N418, dendritic cells)(1:40; Serotec) and FITC-labeled anti-Salmonella anti-body (1:10; KPL, Guildford, UK). The number of den-dritic cells was counted by using 100 photographs foreach group taken at a magnification of ×40 and expressedper square millimeter of PP sections. Localization andnumber of Salmonella organisms in the PP were deter-mined and examined with a fluorescence microscope ora confocal laser scanning microscope (Olympus, Tokyo,Japan). Images were further processed by using OlympusFLUOVIEW 500 software.

Identification of apoptotic change in theliver during infection

Apoptotic change in PP occurring during Salmonella in-fection was analyzed by use of the TUNEL assay, with anin situ apoptosis detection kit (TACS; Trevigen, Gaithers-burg, MD, USA), according to the manufacturer’s in-structions. After sections were prepared as describedabove, the TUNEL procedure was performed with 6 μm-thick sections, followed by a streptavidin-labeled per-oxidase reaction with TACS Blue Label to produce ablue color. The sections were examined by means oflight microscopy, and TUNEL-positive cells were counted,the numbers being expressed per square millimeter ofPP sections.

Assays of caspase-3 in PP

PP tissues were kept in liquid nitrogen until process-ing. Caspases are believed to be among the primary ef-fector molecules of apoptosis. Once activated, caspasescleave various proteins, thereby disabling important cel-lular structural, functional, and repair processes (27).In the present experiment, caspase-3 amidolytic activ-ity was assessed in the PP tissues. Cell lysates were pre-pared by Polytron homogenization (Kinematica GmbH)of PP tissue in 500 μl of cold lysis buffer (20 mM TrispH 8.0, plus 137 mM NaCl, 10% glycerol, 1% Triton X-100, 1 mM phenylmethylsulfonylfluoride, and 10 μg/mlaprotinin). Tissue homogenate supernatant was obtainedby centrifugation (20000 × g at 4oC). The protein con-tent was adjusted to 0.5 mg in 800 μl of reaction buffer(20 mM Tris-HCl, pH 7.5, plus 1% Triton X-100, 137 mMNaCl, and 10% glycerol), and the sample was incubated at37oC for 2 hr with 10 μM caspase-3 synthetic fluorescentsubstrate (MOCAc-Asp-Glu-Val-Asp-Ala-Pro-Lys(Dnp)-NH2) (Peptide Institute, Osaka, Japan). During this assay,10 μM specific caspase-3 inhibitor (Ac-DMQD-CHO)(Peptide Institute) was added to the reaction mixture.Caspase activity was measured fluorometrically at ex-citation and emission wavelengths of 328 and 392 nm,respectively. Specific caspase-3 activity was determinedby subtracting values obtained in the presence of itsinhibitor (28).

Statistical analysis

All data are expressed as means ± SEM. Statistical analysiswas performed by the two-tailed unpaired t-test. A P-valueof < 0.05 was considered statistically significant.

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 199

M. S. Alam et al.

Fig. 1. Localization and number of Salmonella organisms in infectedPP tissue. (a) Immunostaining for Salmonella in control and Salmonella-infected PP from wild-type and iNOS−/− mice. Both wild-type and iNOS-deficient mice received p.o. doses of 2 × 105 CFU/mouse of S. entericaserovar Typhimurium Gifu 12142 strain. PP were immunostained by usingFITC-labeled antibody specific for Salmonella. PP from mock-inoculatedmice were used as controls. (b) Immunofluorescence image illustratingthe localization of Salmonella in the SED area of PP from iNOS−/− mice(arrows). L, lumen. (c) Salmonella count in PP of both wild-type and

iNOS−/− mice. The number of stained bacteria was counted by usingeither fluorescence or confocal laser scanning microscopy. Tissue sec-tions from different PP (n = 6–10) obtained from at least three micewere examined. Columns and error bars indicate means ± SEM (n = 3).∗P < 0.05 vs. wild-type controls. (d) Bacterial count determined by colonyformation assay with PP from wild-type and iNOS-deficient mice on day5 after infection. Data are means ± SEM (n = 6). ∗P < 0.01 vs. wild-typecontrols.

RESULTS

Bacterial yield in PP of wild-type and iNOS−/−mice infected with Salmonella

The natural route of infection of enteric Salmonellais known to be oral, with Salmonella species crossingintestinal mucosal M cells to reach PP lymphoid folli-cles, where they proliferate and then spread to deepertissues via lymphatic and blood vessels (3, 6). To exam-ine the role of iNOS in host defense against Salmonella atthe site at which infection begins, that is, intestinal mu-cosa, wild-type and iNOS-deficient mice were infected p.o.with S. enterica serovar Typhimurium Gifu 12142 strain at

2 × 105 CFU/mouse. At various times after infection, PPwere separated and processed for immunostaining withFITC-labeled anti-Salmonella antibody. Stained bacteriain PP tissues were examined and counted using a flu-orescence microscope or confocal-laser scanning micro-scope. A substantial fluorescence response as shown inFigure 1a indicated the presence and location ofSalmonella in PP, whereas no appreciable fluorescencewas observed with mock-infected PP from either wild-type or iNOS−/− mice. These immunofluorescence datathus support that the observed fluorescence is Salmonellaspecific and not due to nonspecific fluorescence of tissues.A confocal microscopic image of the Salmonella-infected

200 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd

Nitric oxide and murine salmonellosis

iNOS−/− PP more clearly shows the presence and exactlocation of Salmonella in the PP SED region (Fig. 1b). Toexamine the effect of NO on Salmonella colonization inPP, the number of fluorescent bacteria was counted inPP tissues of wild-type and iNOS−/− mice. Figure 1c pro-vides a morphometric analysis of Salmonella localized inPP from both wild-type and iNOS−/− mice during infec-tion. In the early phase of infection (day one), both groupsof mice had almost the same degree of bacterial infiltra-tion on the luminal side of the intestinal mucosa (Fig. 1c),for example, in the FAE and the SED of the PP region (cf.Fig. 1b). At a later stage of infection (day five) however,widespread and significantly greater Salmonella infiltra-tion and growth were observed in the PP from iNOS−/−

mice compared with PP from wild-type mice (Fig. 1c).To verify this immunohistochemical evidence of

Salmonella growth in PP, the total number of viable bacte-ria recovered from PP was quantified with the colony for-mation assay. PP tissue was collected from wild-type andiNOS−/− mice on day five after Salmonella infection andhomogenized. Appropriately diluted PP homogenate wasthen plated on brain-heart infusion agar and cultured toallow colony formation. Consistent with the immunoflu-orescence counting of bacteria, PP of iNOS−/− mice hadsignificantly higher bacterial growth than that of wild-type(Fig. 1d).

These data illustrate that infection caused by S. entericaTyphimurium produces a much more extensive infectionin iNOS-deficient mice than in wild-type mice at the initialsite of infection, which indicates an important defensefunction for NO formed from iNOS in the local intestinalregion of primary Salmonella infection.

Pathological changes in PP of iNOS−/− andwild-type mice infected with Salmonella

Pathological changes in Salmonella-infected PP were eval-uated by histopathological and immunohistochemicalmeans. On day five after p.o. infection with the Gifu 12142strain at 2 × 105 CFU/mouse, histopathological studiesshows that PP from iNOS−/− mice had extensive necroticchanges (with many pyknotic cells) and severe septic in-flammation in the PP SED region, compared with PP fromwild-type mice (Fig. 2a). The inflammatory changes werecaused by infiltration of neutrophils and other inflamma-tory cells such as macrophages. In fact, more extensiveinfiltration of neutrophils was observed in PP tissues ofiNOS−/− mice than in wild-type mice (Fig. 2b), as evi-denced by immunohistochemistry for neutrophils. Thus,for comparison, on day five after infection serial sectionsof PP from wild-type mice were examined by immuno-histochemical methods for Salmonella colonization, iNOSexpression, nitrotyrosine formation, and neutrophil infil-

tration (Fig. 3). Staining for iNOS, neutrophils, and ni-trotyrosine was confined mainly to the area of Salmonellaproliferation. iNOS and nitrotyrosine were expressed insome infiltrating cells and surrounding septic areas aswell. These results indicate that NO may play a crucialrole in ameliorating Salmonella infection and its associ-ated pathological and inflammatory changes, as proposedin our earlier reports (1).

Because PP from iNOS−/− mice were found to sufferfrom severe pathological assault due to increased infiltra-tion of inflammatory cells and cellular damage, TUNELanalysis was performed with PP tissues from iNOS−/− andwild-type mice on day five after Salmonella infection. Asshown in Figures 4a and b, on day five of infection micedeficient in iNOS had extensive apoptotic changes in PPcompared with the wild-type. Such apoptotic changes cor-related well with the degree of neutrophil infiltration: theiNOS−/− PP had much remarkable immunostaining forneutrophils compared with the wild-type tissues (Fig. 4cand d). The localization of infiltrated neutrophils wasnot completely consistent with that of TUNEL-positivecells in PP, however. This finding indicates that duringinfection not only neutrophils but also other inflamma-tory cells such as macrophages and lymphocytes may un-dergo apoptosis, which is exacerbated by a lack of iNOSexpression.

Figure 5a provides results of a quantitative analysis ofTUNEL-positive cells in PP from both wild-type andiNOS−/− mice during infection. A significant increasein the number of apoptotic cells in PP from iNOS−/−

mice occurred on days three and five. In addition, onday one after Salmonella infection, specific caspase-3 ac-tivity was significantly higher in iNOS−/− mice than inwild-type animals (Fig. 5b). This data thus strongly sug-gests that NO not only restricts colonization and growthof Salmonella in PP, but also inhibits cellular damagecaused Salmonella-induced apoptosis, and further sup-ports our earlier report of NO-mediated cytoprotec-tion of mouse liver and macrophages during Salmonellainfection (1, 7).

Increase of dendritic cells in PP from iNOS−/−mice

Although this Salmonella species does infect macrophages(7), this cell population may not be the only one involvedin disseminating intracellular bacteria from mucosal sites.Dendritic cells, a population of APC present in PP alongwith macrophages and neutrophils, also seem to partic-ipate in bacterial dissemination. In fact, S. enterica Ty-phimurium has been reported to efficiently enter, and sur-vive within, murine PP dendritic cells (29). Therefore, ac-cumulation of dendritic (CD11c) cells in PP was examined

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 201

M. S. Alam et al.

Fig. 2. Histopathology of PP from wild-type and iNOS-deficient mice infected with Salmonella. PP sections obtained from wild-type and iNOS−/− miceat day 5 after p.o. infection (2 × 105 CFU/mouse) were stained with hematoxylin-eosin (a), or were immunostained for neutrophils (b). Arrows in (a)point to severe necrotic changes in the SED and arrow heads point to FAE. L, lumen.

by immunohistochemical methods (Fig. 6). The numberof CD11c-positive cells in PP after infection was counted:they were distributed sparsely in PP of both wild-type andiNOS−/− mice. A significant increase in the dendritic cellcount, however, was observed on day five after infection in

iNOS−/− mice compared with wild-type mice. Increaseddendritic cell populations were also noted in the luminalarea, outside the PP, in both wild-type and iNOS−/− mice,although this increase occurred to a greater degree in theiNOS−/− mice (data not shown).

202 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd

Nitric oxide and murine salmonellosis

Fig. 3. Immunohistochemical analysis of PP from wild-type mice on day 5 after infection. Mice were infected p.o. with 2 × 105 CFU of the Gifu 12142strain per mouse. Sequential sections were immunostained for Salmonella colonization (via FITC) (a), nitrotyrosine formation (b), iNOS expression (c),and neutrophil infiltration (d).

DISCUSSION

Previous studies have suggested that both NO and super-oxide contribute in critical ways to the host during S. enter-ica serovar Typhimurium infection in mice (2). More re-cently, we reported that iNOS−/− mice infected i.p. with anavirulent S. enterica serovar Typhimurium strain showedan increased bacterial load, mortality, and liver damagewith extensive hepatocellular apoptosis, as well as devel-opment of septicemia, compared with wild-type mice (1).Similar results were obtained from our recent observationof experimental infection with S. enterica serovar Typhiin vivo and in vitro (7). For previously reported workon murine salmonellosis, mice were infected with S. Ty-phimurium by means of i.p. inoculation; the p.o. route wasrarely employed. However, natural Salmonella infectionsare initiated through the p.o. route, after which bacteriainvade the intestinal mucosa, cross the epithelial barrierthrough M cells overlying PP, proliferate, and spread todeeper tissues via lymphatics and the bloodstream (3, 6).Because of the possible importance of the route of in-

fection, we decided to investigate the effect of NO dur-ing the early stages of infection, including its effect onSalmonella invasion of intestinal mucosa. For the presentstudy, therefore, we produced murine salmonellosis byp.o. administration of the virulent S. enterica serovar Ty-phimurium Gifu 12142 strain to wild-type and iNOS−/−

mice. Our present results show that a lack of iNOS ex-pression after p.o. infection exacerbates bacterial growthand increases infiltration of inflammatory cells, includingdendritic cells, in PP tissues, together with inducing severeapoptotic changes in cells in the lymphoid follicles.

Other studies on both mice and human subjects havedocumented induction of iNOS in the intestine after bac-terial infection (30, 31). NO, particularly NO produced inexcess by activated phagocytes expressing iNOS, has beenshown to function as a cytotoxic or cytostatic moleculeand inhibit the growth of pathogenic bacteria and pro-tozoa (8–10, 32). Overproduction of NO has also beenimplicated in the pathogenesis of sepsis, cerebral malaria,and some viral infections (9, 33). NO may be involved incarcinogenesis induced by parasites, viruses, and bacteria

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 203

M. S. Alam et al.

Fig. 4. Apoptosis and neutrophil infiltration in PP of wild-type andiNOS−/− mice during Salmonella infection. Mice were infected with theGifu 12142 strain in the same manner as described earlier. Sequentialsections of liver tissue obtained on day 5 after infection were examined

for apoptosis via the TUNEL method and for neutrophil infiltration by im-munohistochemical analysis, as described earlier. Apoptotic cells staineddeep blue (a and b), and neutrophils stained brown (c and d).

(e.g. Helicobacter pylori) by virtue of its mutagenic po-tential (33, 34). The pathological consequences of NOproduction for microbial pathogenesis seem to be deter-mined by delicate and complicated interactions betweenhost and pathogen. As an example, our present study ofmurine salmonellosis provides evidence of a potent hostdefense function of NO, which was most obvious in PP inintestines of Salmonella-infected mice.

It is now well known that NO itself is not a bacte-ricidal molecule (11, 15, 16). Rather, its cytotoxic effectis realized by derivative reactive nitrogen oxides (35).For example, NO reacts with superoxide in a diffusion-limited reaction to yield peroxynitrite (36, 37), whichis a strong oxidant and nitrating agent, to produce po-tent cytotoxic actions against various microbes by meansof disintegration and chemical modification of variousbiomolecules such as membrane lipids (35), nucleic acids(34, 38), and proteins (36, 39). Pathogenic intruders into

the host thus undergo oxidative stress caused by NO-meditated host defense (33). Resistance of Salmonella tooxidative and nitrative stress may allow this pathogen towithstand the NO- and oxygen radical-dependent killingmechanisms of phagocytic cells (7, 11). Peroxynitrite pro-duction can be identified indirectly by immunohisto-chemical detection of nitrotyrosine produced in cells andtissues (1, 40, 41). Although nitrotyrosine can be formedin a peroxynitrite-independent reaction, such as theNO−

2 -H2O2-peroxidase pathway (42), peroxynitrite ap-pears to be a major contributor to nitrotyrosine forma-tion in vivo. Intensive staining for nitrotyrosine was ob-served in local areas of microabscesses formed in PP fromSalmonella-infected wild-type mice; this staining colocal-ized with that for iNOS that was induced by Salmonellacolonization (Fig. 3). Furthermore, we previously showedthat pharmacological inhibition of either NO or superox-ide resulted in enhanced Salmonella growth in liver and

204 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd

Nitric oxide and murine salmonellosis

a

1 3 5

Days after infection

Wild-type mice

iNOS-/- mice

0

20

40

60

80

100

120N

o.

of T

UN

EL-p

ositiv

e c

ells

/mm

2

∗

∗

b

Day 1 after infection

0

100

200

300

400

500

600

∗

Caspase-3

activity/m

g p

rote

in

Wild-type mice

iNOS-/- mice

Fig. 5. Increased apoptosis in PP from iNOS−/− mice. (a) TUNEL analysis of PP tissues from both wild-type and iNOS−/− mice after infection. (b)Caspase-3 activity in infected PP tissues in the early phase (day 1) of infection. Mice were infected p.o. with 2 × 105 CFU of S.enterica serovarTyphimurium Gifu 12142 strain per mouse. Columns and error bars indicate means ± SEM (n = 3). ∗P < 0.05 vs. wild-type controls.

blood (2). It is therefore quite reasonable to expect thatperoxynitrite generated in PP has an antibacterial effectduring Salmonella infection in mice.

PP (lymphoid follicles) serve as the primary collectingsite for antigenic materials, including bacteria and virus,moving across the gut mucosa (43). PP are thus compo-nents of the mucosal immune system, and are importantin host defense against invasive enteric pathogens (44).Also, in the absence of iNOS, the leukocyte-endotheliuminteraction is markedly enhanced, which results in re-cruitment of more leukocytes and an exacerbated inflam-matory process (45). In the present study we observeda significant increase in Salmonella count, accompaniedby increased and widespread infiltration of neutrophils inPP from iNOS−/− mice, compared with wild-type mice.We also observed a marked increase of dendritic cells inPP and outside the intestinal epithelium (luminal area)during infection, which was consistent with the report ofRescigno et al. (46) which described similar Salmonella-induced dendritic cell recruitment. In addition, the num-ber of dendritic cells in PP from iNOS−/− mice was sig-nificantly increased compared with that in wild-type mice(Fig. 6). Dendritic cells are APC that have evolved to mon-itor the environment, detect pathogens, and trigger T cellactivation, thereby providing a link between the innateand adaptive immune systems (47). Therefore, elevatednumbers of dendritic cells in PP from iNOS−/− mice maybe part of the innate immune response to an increased

bacterial burden. It is also possible that increased num-ber of dendritic cells would allow faster dissemination ofSalmonella in iNOS−/− mice.

Invasive enteropathogens (such as Shigella, Salmonella,and enteroinvasive Escherichia coli) trigger enhanced ex-pression of proinflammatory genes in intestinal epithelialcells, which subsequently causes a massive increase in theoccurrence of apoptotic cells in the epithelium and PP(29, 48). Previous studies have reported that enteric infec-tion induces apoptosis of macrophages, B cells, and CD4+

and CD8+ αβT cells that are present in PP (25, 48). Wehave found that Salmonella infection causes much moreextensive PP damage (microabscess formation and induc-tion of apoptosis) in NOS-deficient mice than in wild-type mice, and cells that underwent apoptosis were mostlyinfiltrating macrophages, neutrophils, and lymphocytes.Salmonella organisms can infect host cells and grow inmacrophages, hepatocytes (49), and CD11c-positive den-dritic cells (29), and can induce apoptosis of host cells viaactivation of a cascade of intracellular proteases knownas caspases (17–20). S. Typhimurium-derived SipB pro-tein, which is translocated into infected cells after biosyn-thesis by Salmonella, appears to be directly involved ininduction of apoptosis by binding and activating caspase-1 and -2 (17, 19, 50). Such apoptotic processes are thoughtto be essential for Salmonella to penetrate the epithelialbarrier and enter the systemic circulation, and therebycause typhoid-like diseases in mice.

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 205

M. S. Alam et al.

Fig. 6. Increase in number of dendritic cells in PPfrom iNOS−/− mice. (a) Immunostaining fordendritic cells (CD11c-positive) in PP of wild-typeand iNOS−/− mice. Mice were infected p.o. withthe Gifu 12142 strain at 2 × 105 CFU/mouse,and PP were immunostained with antibody todendritic (CD11c) cells on day 5 after infection.Mock-inoculated PP of wild-type and iNOS−/−

mice were used as control. Dendritic cells stainedbrown. (b) Morphometric analysis of CD11c cellsin PP during infection. Sections from different PP(n = 6–10) and at least three mice wereexamined and counted microscopically. Columnsand error bars indicate means ± SEM (n = 3).∗P < 0.05 vs. wild-type controls.

NO is now known to produce potent antiapoptoticeffects by influencing the caspase cascade (21–23). Forexample, NO can directly block caspase activity by S-nitrosylation of the active site cysteine in caspases (23).If Salmonella activates proapoptotic molecules, and if NOcounteracts apoptosis during Salmonella infections, a de-crease or eventual absence of NO production may resultin increased apoptosis (1, 7). Indeed, our TUNEL assayof infected PP from iNOS−/− mice showed a widespreaddistribution of TUNEL-positive cells in PP. Similarly, weobserved a higher level of caspase-3 specific activity in PPfrom iNOS−/− mice than in the PP from wild-type micein the early phase of infection, even when numbers of bac-teria in PP from both groups of mice were similar. Thiscaspase activation may also be caused by greater apoptoticresponses induced by the absence of NO. We have previ-ously reported extensive hepatocyte apoptosis in infected

iNOS−/− mice (1, 7). In our current study, we observedsimilar extensive apoptotic changes in PP from iNOS−/−

mice that may have resulted from the loss of potent anti-apoptotic effects of NO in the intestine. The antiapoptoticeffect of NO may thus contribute to host defense by pre-venting apoptosis of host cells and subsequent Salmonellainvasion and dissemination into deeper tissues from pri-mary septic foci. In this context, it is of great impor-tance that we could recently identify a novel NO signalingmolecule, 8-nitroguanosine 3′,5′-cyclic monophosphate,potentially contributing to the NO-mediated cytoprotec-tive effects (51).

NO is endowed with the unique ability to either initi-ate or block apoptosis, the effect depending on multiplevariables (21). An antitiapoptotic function of endogenousNO generated by iNOS has been observed in B cell lines,macrophages, and T lymphocytes (52), and the benefits

206 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd

Nitric oxide and murine salmonellosis

of a cell-mediated immune effector mechanism involvingcytokines are well known for salmonellosis (53). Together,the findings from the present study clearly demonstratethe positive effects of NO produced from iNOS as part ofa primary immune response that helps the host survivesalmonellosis.

ACKNOWLEDGMENTS

We thank Judith B. Gandy for her excellent editorial workon the manuscript. The work was supported by a grant-in-aid for Scientific Research from the Ministry of Education,Culture, Sports, Science and Technology and from theMinistry of Health, Labor and Welfare of Japan.

REFERENCES

1. Alam M.S., Akaike T., Okamoto S., Kubota T., Yoshitake J., Sawa T.,Miyamoto Y., Tamura F., Maeda H. (2002) Role of nitric oxide inhost defense in murine salmonellosis as a function of itsantibacterial and antiapoptotic activities. Infect Immun 70: 3130–42.

2. Umezawa K., Akaike T., Fujii S., Suga M., Setoguchi K., Ozawa A.,Maeda H. (1997) Induction of nitric oxide synthesis and xanthineoxidase and their roles in the antimicrobial mechanism againstSalmonella typhimurium infection in mice. Infect Immun 65:2932–40.

3. Darwin K.H., Miller V.L. (1999) Molecular basis of the interactionof Salmonella with the intestinal mucosa. Clin Microbiol Rev 12:405–28.

4. Savidge T.C., Smith M.W., James P.S., Aldred P. (1991)Salmonella-induced M-cell formation in germ-free mouse Peyer’spatch tissue. Am J Pathol 139: 177–84.

5. McGhee J.R., Kiyono H. (1999) The mucosal immune system. In:Paul W.A., ed. Fundamental Immunology. Philadelphia:Lippincott-Raven, pp. 909–44.

6. Jones B.D., Falkow S. (1996) Salmonellosis: host immune responsesand bacterial virulence determinants. Annu Rev Immunol 14:533–61.

7. Alam M.S., Zaki M.H., Yoshitake J., Akuta T., Ezaki T., Akaike T.(2006) Involvement of Salmonella enterica serovar Typhi RpoS inresistance to NO-mediated host defense against serovar Typhiinfection. Microb Pathog 40: 116–25.

8. Granger D.L., Hibbs J.B. Jr., Perfect J.R., Durack D.T. (1988)Specific amino acid (L-arginine) requirement for the microbiostaticactivity of murine macrophages. J Clin Invest 81: 1129–36.

9. James S.L. (1995) Role of nitric oxide in parasitic infections.Microbiol Rev 59: 533–47.

10. Nathan C., Shiloh M.U. (2000) Reactive oxygen and nitrogenintermediates in the relationship between mammalian hosts andmicrobial pathogens. Proc Natl Acad Sci USA 97: 8841–48.

11. DeGroote M.A., Fang F.C. (2000) Antimicrobial properties of nitricoxide. In: Fang F.C., ed. Nitric Oxide and Infection. New York:Kluwer Academic/Plenum, pp. 231–61.

12. DeGroote M.A., Granger D., Xu Y., Campbell G., Prince R., FangF.C. (1995) Genetic and redox determinants of nitric oxidecytotoxicity in a Salmonella typhimurium model. Proc Natl Acad SciUSA 92: 6399–403.

13. Kuwahara H., Miyamoto Y., Akaike T., Kubota T., Sawa T.,Okamoto S., Maeda H. (2000) Helicobacter pylori urease suppresses

bactericidal activity of peroxynitrite via carbon dioxide production.Infect Immun 68: 4378–83.

14. Miyamoto Y., Akaike T., Alam M.S., Inoue K., Hamamoto T., IkebeN., Yoshitake J., Okamoto T., Maeda H. (2000) Novel functions ofhuman α1-protease inhibitor after S-nitrosylation: inhibition ofcysteine protease and antibacterial activity. Biochem Biophys ResCommum 267: 918–23.

15. Yoshida K., Akaike T., Doi T., Sato K., Ijiri S., Suga M., Ando M.,Maeda H. (1993) Pronounced enhancement of NO-dependentantimicrobial action by an NO-oxidizing agent, imidazolineoxylN-oxide. Infect Immun 61: 3552–5.

16. Akaike T. (2000) Mechanisms of biological S-nitrosation and itsmeasurement. Free Radic Res 33: 461–9.

17. Jesenberger V., Procyk K.J., Yuan J., Reipert S., Baccarini M. (2000)Salmonella-induced caspase-2 activation in macrophages: a novelmechanism in pathogen-mediated apoptosis. J Exp Med 192:1035–46.

18. Kim J.M., Eckmann L., Savidge T.C., Lowe D.C., Witthoft T.,Kagnoff M.F. (1998) Apoptosis of human intestinal epithelial cellsafter bacterial invasion. J Clin Invest 102: 1815–23.

19. Monack D.M., Hersh D., Ghori N., Bouley D., Zychlinsky A.,Falkow S. (2000) Salmonella exploits caspase-1 to colonize Peyer’spatches in a murine typhoid model. J Exp Med 192: 249–58.

20. Monack D.M., Raupach B., Hromockyj A.E., Falkow S. (1996)Salmonella typhimurium invasion induces apoptosis in infectedmacrophages. Proc Natl Acad Sci USA 93: 9833–8.

21. Kim P.K, Zamora R., Petrosko P., Billiar T.R. (2001) Theregulatory role of nitric oxide in apoptosis. Int Immunopharmacol1: 1421–41.

22. Kim Y.M., Talanian R.V., Li J., Billiar T.R. (1998) Nitric oxideprevents IL-1β and IFN-γ-inducing factor (IL-18) release frommacrophages by inhibiting caspase-1 (IL-1β converting enzyme). JImmunol 161: 4122–8.

23. Mannick J.B., Miao X.Q., Stamler J.S. (1997) Nitric oxide inhibitsFas-induced apoptosis. J Biol Chem 272: 24125–8.

24. Vazquez-Torres A., Jones-Carson J., Mastroeni P., Ischiropoulos H.,Fang F.C. (2000) Antimicrobial actions of the NADPH phagocyteoxidase and inducible nitric oxide synthase in experimentalsalmonellosis. I. Effects of microbial killing by activated peritonealmacrophages in vitro. J Exp Med 192: 227–36.

25. Liesenfeld O., Kosek J.C., Suzuki Y. (1997) Gamma interferoninduces Fas-dependent apoptosis of Peyer’s patch T cells in micefollowing peroral infection with Toxoplasma gondii. Infect Immun65: 4682–9.

26. Laubach V.E., Shesely E.G., Smithies O., Sherman P.A. (1995) Micelacking inducible nitric oxide synthase are not resistant tolipopolysaccharide-induced death. Proc Natl Acad Sci USA 92:10688–92.

27. Heczko U., Carthy C.M., O’Brien B.A., Finlay B.B. (2001)Decreased apoptosis in the ileum and ileal Peyer’s patches: a featureafter infection with rabbit enteropathogenic Escherichia coli O103.Infect Immun 69: 4580–9.

28. Enari M., Talanian R.V., Wong W.W, Nagata S. (1996) Sequentialactivity of ICE-like and CPP-32 like proteases during Fas-mediatedapoptosis. Nature 380: 723–6.

29. Hopkins S.A., Niedergang F., Corthesy-Theulaz I.E., KraehenbuhlJ.P. (2000) A recombinant Salmonella typhimurium vaccine strain istaken up and survives within murine Peyer’s patch dendritic cells.Cell Microbiol 2: 59–68.

30. Cerquetti M.C., Goren N.B., Ropolo A.J., Grasso D.,Giacomodonato M.N., Vaccaro M.I. (2002) Nitric oxide andapoptosis induced in Peyer’s patches by attenuated strains ofSalmonella enterica serovar Enteritidis. Infect Immun 70: 964–9.

c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd 207

M. S. Alam et al.

31. Salzman A.L., Eaves-Pyles T., Linn S.C., Denenberg A.G., Szabo C.(1998) Bacterial induction of inducible nitric oxide synthase incultured human intestinal epithelial cells. Gastroenterology 114:93–102.

32. Doi T., Ando M., Akaike T., Suga M., Sato K., Maeda H. (1993)Resistance to nitric oxide in Mycobacterium avium complex and itsimplication in pathogenesis. Infect Immun 61: 1980–9.

33. Akaike T. (2001) Role of free radicals in viral pathogenesis andmutation. Rev Med Virol 11: 87–101.

34. Ohshima H., Bartsch H. (1994) Chronic infections andinflammatory processes as cancer risk factors: possible role of nitricoxide in carcinogenesis. Mutat Res 305: 253–64.

35. Rubbo H., Darley-Usmar V., Freeman B.A. (1996) Nitric oxideregulation of tissue free radical injury. Chem Res Toxicol 9: 809–20.

36. Beckman J.S., Koppenol W.H. (1996) Nitric oxide, superoxide, andperoxynitrite: the good, the bad, and ugly. Am J Physiol 271:C1424–37.

37. Beckman J.S., Beckman T.W., Chen J., Marshall P.A., Freeman B.A.(1990) Apparent hydroxyl radical production by peroxynitrite:implications for endothelial injury from nitric oxide andsuperoxide. Proc Natl Acad Sci USA 87: 1620–4.

38. Juedes M.J., Wogan G.N. (1996) Peroxynitrite-induced mutationspectra of pSP189 following replication in bacteria and in humancells. Mutat Res 349: 51–61.

39. Radi R., Beckman J.S., Bush K.M., Freeman B.A. (1991)Peroxynitrite oxidation of sulfhydryls. The cytotoxic potential ofsuperoxide and nitric oxide. J Biol Chem 266: 4244–50.

40. Kooy N.W, Royall J.A., Ye Y.Z., Kelly D.R., Beckman J.S. (1995)Evidence for in vivo peroxynitrite production in human acute lunginjury. Am J Respir Crit Care Med 151: 1250–4.

41. Szabo C., Salzman A.L., Ischiropoulos H. (1995) Endotoxin triggersthe expression of an inducible isoform of nitric oxide synthase andthe formation of peroxynitrite in the rat aorta in vivo. FEBS Lett363: 235–8.

42. Eiserich J.P., Hristova M., Cross C.E., Jones A.D., Freeman B.A.,Halliwell B., Van der Vliet A. (1998) Formation of nitricoxide-derived inflammatory oxidants by myeloperoxidase inneutrophils. Nature 391: 393–7.

43. Carter P.B., Collins F.M. (1974) The route of enteric infection innormal mice. J Exp Med 139: 1189–203.

44. Heel K.A., McCauley R.D., Papadimitriou J.M., Hall J.C. (1997)Peyer’s patches. J Gastroenterol Hepatol 12: 122–36.

45. Hickey M.J., Sharkey K.A., Sihota E.G., Reinhardt P.H.,Macmicking J.D., Nathan C., Kubes P. (1997) Inducible nitric oxidesynthase-deficient mice have enhanced leukocyte-endotheliuminteractions in endotoxemia. FASEB J 11: 955–64.

46. Rescigno M., Urbano M., Valzasina B., Francolini M., Rotta G.,Bonasio R., Granucci F.J., Kraehenbuhl P., Ricciardi-Castagnoli P.(2001) Dendritic cells express tight junction proteins and penetrategut epithelial monolayers to sample bacteria. Nat Immunol 2:361–7.

47. Rescigno M., Borrow P. (2001) The host-pathogen interaction: newthemes from dendritic cell biology. Cell 106: 267–70.

48. Zychlinsky A., Thirumalai K., Arondel J., Cantey J.R., AliprantisA.O., Sansonetti P.J. (1996) In vivo apoptosis in Shigella flexneriinfections. Infect Immun 64: 5357–65.

49. Mastroeni P., Skepper J.N., Hormaeche C.E. (1995) Effect ofanti-tumor necrosis factor alpha antibodies on histopathology ofprimary Salmonella infections. Infect Immun 63: 3674–82.

50. Hersh D., Monack D.M., Smith M.R., Ghori N., Falkow S.,Zychlinsky A. (1999) The Salmonella invasin SipB inducesmacrophage apoptosis by binding to caspase-1. Proc Natl Acad SciUSA 96: 2396–401.

51. Sawa T., Zaki M.H., Okamoto T., Akuta T., Tokutomi Y.,Kim-Mitsuyama S., Ihara H., Kobayashi A., Yamamoto M., Fujii S.,Arimoto H., Akaike T. (2007) Protein S-guanylation by thebiological signal 8-nitroguanosine 3′,5′-cyclic monophosphate. NatChem Biol 3: 727–35.

52. Bogdan C., Rollinghoff M., Diefenbach A. (2000) Reactive oxygenand reactive nitrogen intermediates in innate and specificimmunity. Curr Opin Immunol 12: 64–76.

53. Sztein M.B., Wasserman S.S., Tacket C.O., Edelman R., Hone D.,Lindberg A.A., Levine M.M. (1994) Cytokine production patternsand lymphoproliferative responses in volunteers orally immunizedwith attenuated vaccine strains of Salmonella typhi. J Infect Dis 170:1508–17.

208 c© 2008 The Societies and Blackwell Publishing Asia Pty Ltd