Polyamines Impair Immunity to Helicobacter pylori by Inhibiting L-Arginine Uptake Required for Nitric Oxide Production

GASTROENTEROLOGY 2010;139:1686 –1698

Polyamines Impair Immunity to Helicobacter pylori by Inhibiting L-Arginine Uptake Required for Nitric Oxide Production RUPESH CHATURVEDI,*,‡ MOHAMMAD ASIM,*,‡ SVEA HOGE,*,§ NURUDDEEN D. LEWIS,*,储 KSHIPRA SINGH,*,‡ DANIEL P. BARRY,* THIBAUT DE SABLET,*,‡ M. BLANCA PIAZUELO,* ADITYA R. SARVARIA,* YULAN CHENG,¶ ELLEN I. CLOSS,# ROBERT A. CASERO Jr,** ALAIN P. GOBERT,*,‡‡ and KEITH T. WILSON*,‡,储 *Division of Gastroenterology, Hepatology, and Nutrition, Department of Medicine, and 储Department of Cancer Biology, Vanderbilt University Medical Center, Nashville, Tennessee; ‡Veterans Affairs Tennessee Valley Healthcare System, Nashville, Tennessee; §Department of General, Abdominal and Vascular Surgery, Otto-von-Guericke University, Magdeburg, Germany; ¶Division of Gastroenterology, University of Maryland School of Medicine, Baltimore, Maryland; #Department of Pharmacology, Johannes Gutenberg University, Mainz, Germany; **Department of Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine, Baltimore, Maryland; and ‡‡Institut National de la Recherche Agronomique, Unité de Microbiologie UR454, Saint-Genès-Champanelle, France

BASIC– ALIMENTARY TRACT

BACKGROUND & AIMS: Helicobacter pylori–induced immune responses fail to eradicate the bacterium. Nitric oxide (NO) can kill H pylori. However, translation of inducible NO synthase (iNOS) and NO generation by H pylori–stimulated macrophages is inhibited by the polyamine spermine derived from ornithine decarboxylase (ODC), and is dependent on availability of the iNOS substrate L-arginine (LArg). We determined if spermine inhibits iNOS-mediated immunity by reducing L-Arg uptake into macrophages. METHODS: Levels of the inducible cationic amino acid transporter (CAT)2, ODC, and iNOS were measured in macrophages and H pylori gastritis tissues. L-Arg uptake, iNOS expression, and NO levels were assessed in cells with small interfering RNA knockdown of CAT2 or ODC, and in gastric macrophages. The ODC inhibitor, ␣-difluoromethylornithine, was administered to H pylori–infected mice for 4 months after inoculation. RESULTS: H pylori induced CAT2 and uptake of L-Arg in RAW 264.7 or primary macrophages. Addition of spermine or knockdown of CAT2 inhibited L-Arg uptake, NO production, and iNOS protein levels, whereas knockdown of ODC had the opposite effect. CAT2 and ODC were increased in mouse and human H pylori gastritis tissues and localized to macrophages. Gastric macrophages from H pylori–infected mice showed increased ODC expression, and attenuated iNOS and NO levels upon ex vivo H pylori stimulation versus cells from uninfected mice. ␣-Difluoromethylornithine treatment of infected mice restored L-Arg uptake, iNOS protein expression, and NO production in gastric macrophages, and significantly reduced both H pylori colonization levels and gastritis severity. CONCLUSIONS: Up-regulation of ODC in gastric macrophages impairs host defense against H pylori by suppressing iNOS-derived NO production. Keywords: Spermine; Macrophages; Gastritis.

H

elicobacter pylori is a gram-negative, microaerophilic bacterium that colonizes the human gastric mucosa. It infects about half of the world’s population, and approximately 30%– 40% in the United States.1,2 H pylori induces an innate immune response in neutrophils, monocytes, and macrophages,3–7 and a

chronic lymphocytic response.8 Although H pylori is classically considered a noninvasive pathogen, the bacterium disrupts epithelial integrity and both the bacterium itself 9 and its antigens3 are present in the lamina propria, and are in direct contact with immune cells where they can impact responses in macrophages and other antigen-presenting cells.8,9 However, the overall immune response is ineffective, because the bacterium generally persists for the life of the host, leading to chronic gastritis, peptic ulcers, lymphoma, and gastric adenocarcinoma, the second leading cause of cancer death worldwide.1 Therefore, gaining insight into the failure of the immune response could be of great importance in developing new strategies for chemoprevention or disease management. Nitric oxide is a fundamental component of innate immunity with well-characterized antimicrobial and antitumor effector functions.6,10 –13 High-output production of NO derives from the substrate L-arginine (L-Arg) by inducible NO synthase (iNOS, NOS2), which defines the classic pathway of activation in macrophages by microbes and their products, and is of critical importance to host defense.6,10,14 We have shown that H pylori induces iNOS and NO production in macrophages in vitro and that this effect is contactindependent, occurring with H pylori lysates, soluble surface proteins, culture supernatants, and recombinant urease.4 – 6,15,16 In addition, we have reported that NO generation in H pylori–stimulated macrophages is strongly dependent on L-Arg availability in culture medium and requires concentrations well above the circulating physiologic level of L-Arg in mice and human beings of 0.1 mmol/L to maximize response; this effect was mediated primarily by an enhancement of Abbreviations used in this paper: CAT, cationic amino acid transporter; DFMO, ␣-difluoromethylornithine; iNOS, inducible nitric oxide synthase; L-Arg, L-arginine; L-Lys, L-lysine; NO, nitric oxide; ODC, ornithine decarboxylase; RT-PCR, reverse-transcription polymerase chain reaction; siRNA, small interfering RNA; Th, T helper. © 2010 by the AGA Institute 0016-5085/$36.00 doi:10.1053/j.gastro.2010.06.060

iNOS translation with increasing L-Arg levels.16 Further, in our studies, macrophages kill H pylori in an NO-dependent manner, but this requires concentrations of L-Arg of 0.4 mmol/L and greater.6,15,16 Consistent with our findings, it has been shown that L-Arg can increase NO generation by iNOS, even at concentrations greatly above the Michaelis constant of the enzyme,10 which is in the range of 10 ␮mol/L.11 iNOS is expressed in gastritis tissues of H pylori– infected human beings, and localizes to inflammatory cells, particularly macrophages.17,18 However, iNOS is unable to generate an antimicrobial benefit for the infected host. The adaptive immune response is also insufficient because enhancement of T helper (Th)1/ Th17 responses in adoptive transfer and vaccine studies reduces H pylori colonization.8,19,20 We have reported that one mechanism that restricts H pylori– stimulated NO generation in macrophages is the simultaneous generation of polyamines within these cells by up-regulation of ornithine decarboxylase (ODC).5,15,21,22 ODC converts L-ornithine into putrescine, which is metabolized into the polyamines spermidine and spermine by constitutively expressed synthases. We have reported that spermine acts to inhibit NO production and killing of H pylori by reducing stimulated iNOS protein expression through an effect on iNOS translation; spermidine had a more modest inhibitory effect on iNOS, whereas putrescine had no such effect.15 As with iNOS,23 the induction of ODC occurs by soluble factors and is attenuated by isogenic deletion of UreA and recapitulated by recombinant urease.5,15,22 L-Arg uptake in macrophages has been attributed to the y⫹ transport system.12 The cationic amino acid transporters (CAT) include CAT1, which is constitutively expressed and involved in uptake of L-Arg for basic metabolism;13 and CAT2, which includes the alternatively spliced isoforms CAT2A, a low-affinity transporter primarily in liver, and CAT2B, the high-affinity L-Arg transporter in macrophages.12,13,24,25 The CAT2 protein is part of a larger family of solute carriers, and thus also is known as solute carrier 7A2.24 CAT3 and CAT4 also have been described.14,24 CAT3 is found in brain and thymus, and CAT4’s function is unknown at this time.14,24 We report that H pylori–induced CAT2 expression mediates L-Arg uptake, iNOS protein synthesis, and NO production, and that spermine inhibits iNOS by blocking L-Arg transport into macrophages. CAT2 and ODC are both up-regulated in macrophages within H pylori gastritis tissues. Administration of an ODC inhibitor to H pylori– infected mice enhanced L-Arg uptake, iNOS protein levels, and NO production in gastric macrophages, and resulted in significant attenuation of both H pylori colonization and gastritis. Therefore, modulation of polyamine generation in the stomach could represent a novel approach to enhancing immunity to H pylori.

L–ARGININE UPTAKE IN H PYLORI INFECTION

1687

Materials and Methods Reagents Reagents for cell culture, RNA extraction, and reverse-transcription polymerase chain reaction (RTPCR) were from Invitrogen (Carlsbad, CA). Other chemicals were from Sigma (St Louis, MO). ␣-Difluoromethylornithine (DFMO) was from ILEX Oncology (San Antonio, TX).

Bacteria H pylori SS1, Citrobacter rodentium, and Campylobacter jejuni were used as described in the Supplementary Materials and Methods section.

Mice, Cells, and Culture Conditions The murine macrophage RAW 264.7 cell line was maintained as described.4,5,15,21 In C57BL/6 mice, peritoneal macrophages were isolated,15 gastric infection was performed,5,26 and colonization and gastritis were assessed (Supplementary Materials and Methods). Some mice were treated with DFMO (1% wt/vol) in the drinking water.27,28

Human Subjects Tissues from endoscopic biopsies were used as described.17

Measurement of NO The concentration of nitrite (NO2⫺), the oxidized metabolite of NO, was assessed by the Griess reaction.4-6,15

RT-PCR RNA was extracted using TRIzol (Invitrogen) and complementary DNA (cDNA) synthesized.5,16 Primer sequences are described in Supplementary Table 1 and PCR methods are described in the Supplementary Materials and Methods section.

L-Arg Transport See the Supplementary Materials and Methods section for details.

Immunoblotting for iNOS and CAT2 Western blotting for iNOS and ␤-actin was performed as described.15,16 CAT2 was detected with a rabbit polyclonal antibody to the C-terminal 70 amino acids of mouse CAT2.12

Transient Transfection of CAT2 and ODC Small Interfering RNA Transfection with small interfering RNA (siRNA) duplexes was performed as described15,21; sequences are

BASIC– ALIMENTARY TRACT

November 2010

1688

CHATURVEDI ET AL

as described in the Supplementary Materials and Methods section.

ODC Activity ODC enzyme activity was measured by radiochemical assay.27

GASTROENTEROLOGY Vol. 139, No. 5

Immunohistochemistry for ODC and CAT2, and Immunofluorescence Staining for CAT2 and the Macrophage Marker F4/80 See the Supplementary Materials and Methods section for details.

BASIC– ALIMENTARY TRACT

Figure 1. Induction of CAT2 and L-Arg uptake in RAW 264.7 macrophages and effect of spermine and L-lysine. Macrophages were incubated with H pylori lysate (HPL) at a multiplicity of infection of 100. The mRNA expression for CAT1 and CAT2 was assessed at 6 hours after HPL stimulation ⫾ the inhibitor (Inhib) spermine (Spm; 12.5 ␮mol/L) or L-lysine (Lys; 20 mmol/L) by (A) real-time PCR and (B) semiquantitative RT-PCR. (C) Western blot analysis for CAT2 (80 kilodaltons), iNOS (130 kilodaltons), and ␤-actin (42 kilodaltons) 24 hours after stimulation. (D) L-Arg transport assessed as uptake of 14C-L-Arg into cells at the times indicated. (E and F) NO levels measured as concentration of NO2⫺, at 24 hours after stimulation. (F) L-Arg was added to L-Arg-free, serum-free medium ⫾ Spm. (A–F) Data are from 3 experiments performed in duplicate. (A and E) **P ⬍ .01 versus control; (D) *P ⬍ .05, **P ⬍ .01 versus control, HPL ⫹ Spm, and HPL ⫹ Lys; (E and F) §§P ⬍ .01, §§§P ⬍ .001 versus HPL alone.

L–ARGININE UPTAKE IN H PYLORI INFECTION

Polyamine Measurement Polyamines were measured as described.

27

Gastric Macrophages Cells were isolated from glandular stomach by positive selection with F4/80 antibody.16,29 RNA isola-

1689

tion, real-time PCR, and flow cytometry were performed as described.16,29

Statistical Analysis Quantitative data are shown as the mean ⫾ standard error. Analyses are shown in the Supplementary Materials and Methods section.

Results H pylori Stimulates L-Arg Uptake in Macrophages That Are Inhibited by Spermine When RAW 264.7 cells were stimulated with H pylori lysate, CAT2 messenger RNA (mRNA) levels were increased by more than 50-fold when assessed by realtime PCR, whereas CAT1 was not induced (Figure 1A). Neither spermine nor L-lysine (L-Lys), a known competitive inhibitor of L-Arg uptake,25 had any effect on stimulated CAT2 mRNA levels (Figure 1A and B). Similarly, H pylori up-regulated CAT2 protein levels, as assessed by Western blotting, and spermine and L-Lys did not affect CAT2 protein expression (Figure 1C). In contrast, both spermine and L-Lys markedly attenuated iNOS protein levels (Figure 1C), consistent with our previous reports.15,16 Because uptake of extracellular L-Arg has been shown to be required for iNOS-derived NO generation in macrophages,13,16,25 we assessed L-Arg transport from the extracellular medium into macrophages. H pylori stimulation resulted in a 4-fold increase in L-Arg uptake by RAW 264.7 cells, which was blocked by addition of either spermine or L-Lys (Figure 1D). Spermine and L-Lys each inhibited H pylori–stimulated NO production (Figure 1E) in a manner that paralleled their effects on L-Arg uptake. These findings raised the possibility that spermine could inhibit L-Arg uptake by competing for the same transporter. However, addition of increasing levels of L-Arg did not overcome the inhibitory effect of spermine on NO production, even at 6.4 mmol/L (Figure 1F), which is 64-fold greater than the circulating level of 0.1 mmol/L in mice and human beings.27,30 When primary mouse peritoneal macrophages were used, H pylori–stimulated cells also showed induction of CAT2, but not CAT1 mRNA levels (Supplementary Figure 1A), and a significant increase in L-Arg uptake with H pylori stimulation that was abolished by spermine (Sup-

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Figure 2. Effect of transfection of CAT2 siRNA on H pylori–stimulated iNOS expression and NO production. RAW 264.7 cells were transiently transfected with either scrambled siRNA or siRNA to CAT2. (A) mRNA expression assessed by RT-PCR at 6 hours after stimulation with H pylori lysate (HPL). (B) Western blot analysis at 24 hours after stimulation. (A and B) n ⫽ 3. (C) L-Arg uptake and (D) NO2⫺ levels, measured 24 hours after stimulation. **P ⬍ .01 versus control macrophages transfected with scrambled siRNA; §§P ⬍ .01 versus HPL-stimulated macrophages transfected with scrambled siRNA (n ⫽ 6). Scr, scrambled.

BASIC– ALIMENTARY TRACT

November 2010

1690

CHATURVEDI ET AL

plementary Figure 1B). These data are consistent with the inhibitory effect of spermine on NO generation in peritoneal macrophages that we have reported.15

H pylori–Stimulated iNOS Protein Expression and NO Production in Macrophages Is Dependent on Induction of CAT2 The role of CAT2 in NO production in response to extracellular bacterial stimuli is not known. Therefore, CAT2 knockdown by siRNA was performed in H pylori– stimulated cells (Figure 2A). Although this knockdown had no effect on iNOS mRNA levels (Figure 2A), it resulted in a marked inhibition of iNOS protein expres-

GASTROENTEROLOGY Vol. 139, No. 5

sion (Figure 2B), which was paralleled by reduced L-Arg uptake (Figure 2C) and NO generation (Figure 2D). Consistent with our report that L-Arg availability regulates iNOS translation and not protein stability,16 inhibition of iNOS protein degradation with the proteosomal inhibitor lactacystin did not restore iNOS protein levels in the presence of CAT2 siRNA at time points from 6 to 24 hours (Supplementary Figure 2).

Knockdown of ODC Increases iNOS-Derived NO Production by Causing Sustained L-Arg Uptake We next determined the effect of endogenous spermine generation on CAT2, L-Arg uptake, and

BASIC– ALIMENTARY TRACT

Figure 3. Transfection of ODC siRNA enhances L-Arg uptake and iNOS protein expression in RAW 264.7 macrophages. (A) Real-time PCR for ODC with stimulation with lysates of C rodentium (C rod), C jejuni (C jej), and H pylori lysate (HPL) all at a multiplicity of infection of 100. (B) Top panel, RT-PCR for CAT2; bottom panel, Western blot analysis for CAT2 and ␤-actin. (C) L-Arg transport. Scr, scrambled. *P ⬍ .05, **P ⬍ .01 versus control macrophages transfected with scrambled siRNA; §§P ⬍ .01 versus HPL-stimulated macrophages transfected with scrambled siRNA. (D) Western blot analysis for iNOS and ␤-actin. (E) Densitometry of iNOS Western blots, standardized to ␤-actin. (F) NO2⫺ levels measured at 24 hours after stimulation. (E and F) **P ⬍ .01; ***P ⬍ .001 versus macrophages transfected with scrambled siRNA. n ⫽ 3– 4 for all experiments.

November 2010

1691

(Figure 3A) that H pylori lysate induced a 9-fold increase in ODC mRNA levels in cells transfected with scrambled siRNA, whereas lysates of the enteric pathogens C jejuni and C rodentium failed to induce ODC; in

BASIC– ALIMENTARY TRACT

iNOS. We have reported that transient transfection of ODC siRNA results in effective decreases in ODC expression, activity, and spermine levels in H pylori– stimulated RAW 264.7 macrophages.15 We verified

L–ARGININE UPTAKE IN H PYLORI INFECTION

Figure 4. Induction of CAT2 in mouse and human H pylori gastritis tissues. (A) RT-PCR for CAT2 in mouse tissues from uninfected and H pylori–infected mice at 4-months postinoculation. (B) Real-time PCR for CAT2 in mouse tissues, n ⫽ 4 control and 9 infected. **P ⬍ .01 versus uninfected tissues. (C) CAT2 expression by real-time PCR in human tissues from uninfected patients with normal histology (n ⫽ 4), H pylori–negative gastritis (n ⫽ 4), and H pylori–positive gastritis (n ⫽ 6). **P ⬍ .01 versus normal; §§P ⬍ .01 versus H pylori–negative gastritis tissues. (D) Immunofluorescent detection of CAT2 that localizes to macrophages. Mouse tissues were stained for CAT2 detected with fluorescein isothiocyanate (FITC; green) and for the macrophage marker F4/80, detected with tetramethyl rhodamine isothiocyanate (red). Nuclei were stained with 4=,6-diamidino-2-phenylindole (blue). Co-localization is shown in merged images by yellow color. Magnification for each image is 200⫻, plus 600⫻ for inset. (E) Immunohistochemistry for CAT2 in human biopsies, magnification is 200⫻ in left and center panels, and 600⫻ in right panel.

1692

CHATURVEDI ET AL

GASTROENTEROLOGY Vol. 139, No. 5

BASIC– ALIMENTARY TRACT

Figure 5. Induction of ODC in mouse and human H pylori gastritis tissues. At 4 months after inoculation in mice, ODC was assessed by (A) RT-PCR, (B) real-time PCR, and (C) enzyme activity. (B and C) **P ⬍ .01 versus uninfected. (B) n ⫽ 4 uninfected and 9 infected and (C) n ⫽ 5 uninfected and 5 infected. In human beings, ODC was assessed by (D) RT-PCR, (E) real-time PCR, and (F) immunohistochemistry. (E) **P ⬍ .01 for H pylori–positive gastritis (n ⫽ 6) versus uninfected normal tissues (n ⫽ 4); §§P ⬍ .01 versus H pylori–negative gastritis tissues (n ⫽ 4). (F) Magnifications are 200⫻ for both normal and H pylori–positive gastritis, and 600⫻ in the inset.

addition, ODC siRNA abolished H pylori–stimulated ODC expression. ODC siRNA transfection did not affect induction of CAT2 mRNA or protein levels (Figure 3B). However, with ODC knockdown the H pylori– stimulated L-Arg uptake at 6 hours persisted out to 24 hours, whereas cells transfected with scrambled siRNA had attenuation of L-Arg uptake beyond the first 6 hours after activation (Figure 3C). There was more iNOS protein expression detected at 12–24 hours after H pylori stimulation in the presence of ODC siRNA than in cells transfected with scrambled siRNA (Figure 3D). These findings were confirmed by densitometry (Figure 3E). In parallel, ODC knockdown increased

Figure 6. Deficient iNOS protein expression and NO production in gastric macrophages from H pylori–infected mice. Mice were inoculated with H pylori or broth control, and after 4 months gastric macrophages were isolated by positive selection. Cells from both uninfected (light bars) and H pylori–infected (black bars) mice then were treated ex vivo with vehicle control (Ctrl) or H pylori lysate (HPL) for 24 hours. (A–C) Real-time PCR analysis of CAT2, ODC, and iNOS mRNA levels, respectively. (D) iNOS protein levels determined by flow cytometry; data are composite results in relative fluorescence units. (E) Representative histogram tracings for the conditions shown in panel D. (A–D) *P ⬍ .05, **P ⬍ .01 versus unstimulated control cells from uninfected mice; #P ⬍ .05, ##P ⬍ .01 for HPL-stimulated cells from infected mice versus unstimulated cells from infected mice; §P ⬍ .05 for HPL-stimulated cells from infected mice versus HPL-stimulated cells from uninfected mice (n ⫽ 6). (F) NO2⫺ levels (left panel) and L-Arg uptake (right panel). Spermine (Spm; 12.5 ␮mol/L) was added at the start of the ex vivo stimulation. *P ⬍ .05, **P ⬍ .01 versus unstimulated control cells from uninfected mice; §§P ⬍ .01 for HPL-stimulated cells from infected mice versus HPL-stimulated cells from uninfected mice; ##P ⬍ .01 for stimulation with HPL ⫹ Spm versus HPL only in cells from uninfected mice (n ⫽ 6).

NO generation from 12 to 24 hours after activation (Figure 3F). ODC knockdown also enhanced L-Arg uptake and NO production in macrophages activated with live H pylori (Supplementary Figure 3). This effect was not observed when cells were activated with C jejuni or C rodentium (Supplementary Figure 4).

H pylori Infection Causes Induction of CAT2 and ODC In Vivo To provide further biological relevance, we determined the expression of CAT2 and ODC in H pylori gastritis tissues. There was a consistent up-regulation of CAT2 mRNA levels (Figure 4A and B) in mouse chronic gastritis tissues. We also found an 8 –fold increase in CAT2 mRNA levels in human H pylori gastritis tissues compared with normal gastric tissues, whereas CAT2 was not up-regulated in gastritis tissues from subjects without H pylori infection (Figure 4C). CAT2 protein levels were increased in infected mouse tissues when assessed by immunofluorescence, with co-localization to gastric macrophages in the mucosa and submucosa (Figure 4D). Immunoperoxidase staining showed increased CAT2 levels in human H pylori gastritis tissues that localized to lamina propria mononuclear cells, with sporadic staining of some gastric epithelial cells (Figure 4E). We also detected a significant increase in ODC mRNA levels in gastritis tissues from H pylori–infected mice compared with uninfected control tissues (Figure 5A and B). ODC enzyme activity was increased in parallel in the H pylori gastritis tissues (Figure 5C). There was also a significant increase in ODC mRNA expression in human H pylori gastritis tissues (Figure 5D and E), which was not observed with H pylori– negative gastritis tissues. By immunohistochemistry, ODC was expressed primarily in mononuclear cells of the lamina propria in H pylori–infected human gastritis tissues (Figure 5F). Taken together, these data indicate CAT2 and ODC are up-regulated in parallel in H pylori gastritis.

Gastric Macrophages Isolated From H pylori–Infected Mice Show Impaired Uptake of L-Arg, iNOS Protein Expression, and NO Production To assess the interplay between ODC, CAT2, and iNOS in vivo, we studied macrophages isolated from mice chronically infected with H pylori at 4 months postinoculation versus uninfected controls. These cells then were cultured in the presence or absence of H pylori lysate ex vivo, to determine if ODC induction during gastritis regulates iNOS. CAT2 mRNA levels were increased in cells from infected mice (Figure 6A), and upon ex vivo activation with H pylori lysate there was a further increase in CAT2 levels in cells from either uninfected or H pylori– infected mice. In the cases of ODC (Figure 6B) and iNOS (Figure 6C), there was an increase in mRNA levels in

L–ARGININE UPTAKE IN H PYLORI INFECTION

1693

macrophages from infected mice when compared with cells from uninfected mice, and with ex vivo stimulation with H pylori lysate, the cells from infected mice showed a potentiation of mRNA levels. Because there are fewer than 105 macrophages per stomach,29 flow cytometry was used to analyze protein levels; this showed that iNOS was not increased significantly in cells from H pylori–infected mice, and upon ex vivo stimulation these cells expressed less iNOS protein than cells from naive uninfected mice (Figure 6D and E). In parallel, there was a lack of significant increase in NO release from gastric macrophages from H pylori–infected mice, and a decrease in NO production with ex vivo H pylori stimulation when compared with cells from uninfected mice (Figure 6F). When exogenous spermine was added, this significantly attenuated ex vivo–stimulated NO production in cells from uninfected mice, but there was little effect on cells from infected mice. L-Arg uptake (Figure 6F) was increased modestly in cells from infected mice and was decreased substantially in gastric macrophages from infected mice versus uninfected mice upon H pylori stimulation ex vivo, consistent with the higher levels of ODC in these cells (Figure 6B). Spermine reduced ex vivo H pylori–stimulated L-Arg uptake to basal levels in the cells from uninfected mice, but did not further decrease L-Arg uptake in cells from infected mice. These data suggest that induction of ODC in macrophages during chronic H pylori infection may explain a defective NO response to H pylori owing to limited L-Arg uptake and iNOS protein expression.

In Vivo Inhibition of ODC Restores NO-Mediated Immunity to H pylori and Ameliorates Gastritis To directly determine if spermine accumulation attenuates NO-dependent innate immune response to H pylori, we administered the ODC inhibitor DFMO to mice (1% wt/vol in the drinking water) for the 4-month course of infection, based on studies that have established chronic use of this agent for chemoprevention of colon tumors in mice.31 With H pylori infection, there was a 2-fold increase in polyamine levels in gastric tissues that were reduced to basal levels with DFMO treatment (Figure 7A). In freshly isolated gastric macrophages from H pylori–infected mice, the levels of L-Arg uptake (Figure 7B), iNOS protein (Figure 7C), and NO (Figure 7D) were enhanced significantly in cells from mice treated with DFMO. Importantly, direct assessment of gastric colonization with H pylori revealed that DFMO treatment of mice resulted in a significant, 2-log-order reduction in bacterial levels, when assessed by culture or quantitative PCR (Figure 7E). In parallel, there was a significant reduction in gastric inflammation, with no development of gastric atrophy, as shown

BASIC– ALIMENTARY TRACT

November 2010

1694

CHATURVEDI ET AL

GASTROENTEROLOGY Vol. 139, No. 5

BASIC– ALIMENTARY TRACT

by histologic gastritis scores (Figure 7E) and representative histologic sections (Figure 7F).

Discussion In experimental H pylori infection there is a rapid innate immune response with influx of macrophages and neutrophils within 48 hours of inoculation.32 However, persistence of the bacterium has been attributed to defective adaptive immune responses because decreased colonization has been shown with exaggerated Th1 responses in adoptive transfer models19 or enhancement of Th1/Th17 responses in vaccine models.20 The ineffectiveness of the adaptive response has been related to enhanced regulatory T-cell responses, which may protect the host from uncontrolled inflammation, but lead to survival of H pylori.33 We determined if the function of macrophages in the innate response to H pylori infection is compromised by impairment of L-Arg uptake that is essential for M1 response.14 We show that in H pylori infection, the inducible L-Arg transporter CAT2 is up-regulated, but its ability to transport L-Arg in macrophages is inhibited by ODC, resulting in loss of NO-derived immune defense against H pylori (summarized in Supplementary Figure 5). Our data indicate that CAT2 is the primary L-Arg transporter in H pylori–activated macrophages because there was a large induction of CAT2, and not CAT1, and knockdown of CAT2 prevented stimulated L-Arg uptake. Mycobacterium bovis similarly induces L-Arg transport in RAW 264.7 macrophages that are associated with CAT2, but not CAT1 expression.34 In macrophages from CAT2-deficient mice there is deficient NO production and L-Arg transport in response to interferon-␥ plus lipopolysaccharide.13,14 We found that spermine inhibited L-Arg uptake in H pylori–stimulated RAW 264.7 or primary peritoneal macrophages, without altering CAT2 mRNA or protein levels. There is a report of inhibition of CAT2 mRNA expression by spermine in lipopolysaccharide-stimulated alveolar macrophages, but the partial inhibition of CAT2 mRNA did not correlate with a more substantial effect of spermine on L-Arg uptake.35 There is also evidence of spermine inhibiting transport activity of a potassium channel without altering its expression.36

L–ARGININE UPTAKE IN H PYLORI INFECTION

1695

We have reported that H pylori–induced NO production depends on availability of extracellular L-Arg, which enhances iNOS translation and thus iNOS protein synthesis.16 Extracellular spermine can be transported into macrophages within 2 hours of stimulation by lipopolysaccharide.37 However, our data show that spermine is not a competitive inhibitor of L-Arg transport, in contrast to L-Lys.25 Spermine is able to bind to and inactivate proteins,38 raising the possibility that a similar effect could be occurring with CAT2 in H pylori–stimulated cells. We observed a biphasic effect of H pylori on macrophage uptake of L-Arg, with increases at 6 and 24 hours after stimulation, but no increase at 12 or 18 hours. Because we also show that spermine inhibits L-Arg transport, this biphasic time course of L-Arg uptake is consistent with our previous report that ODC expression and enzyme activity peak 6 hours after H pylori stimulation, and spermine levels peak at 12 hours, followed by a decrease out to 24 hours.21,22 We now show that ODC knockdown resulted in a sustained increase in L-Arg uptake without this biphasic pattern, resulting in a concomitant increase in iNOS protein expression and NO production at time points from 12 to 24 hours. These data indicate that endogenous synthesis of spermine by the induction of ODC impairs macrophage L-Arg transport in H pylori infection. Spermine also may regulate its own production by inhibiting L-Arg uptake because L-Arg can be rate-limiting for synthesis of the ODC substrate Lornithine by arginase.14 These effects were specific to H pylori because C jejuni and C rodentium did not induce ODC, and ODC knockdown had no effect on L-Arg transport or NO production induced by these pathogens. An important consideration is the in vivo significance of CAT2 and ODC in the infected stomach. CAT2 mRNA expression is increased in models of lung injury.39 Our data show up-regulation of CAT2 expression and macrophage localization in H pylori gastritis tissues. It has been reported that H pylori eradication decreases ODC mRNA expression and activity,40 but the cellular source of ODC has not been identified. We now show that ODC expression is increased specifically in human H pylori infection compared with H pylori–

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Figure 7. Inhibition of polyamine synthesis restores L-Arg uptake and iNOS-derived NO production and reduces H pylori colonization and gastritis. Mice were infected with H pylori or broth control and received DFMO (1% wt/vol) in the drinking water or water alone for 4 months, beginning the day after inoculation with H pylori. (A) Polyamine levels in stomach tissues. (B–D) Gastric macrophages were isolated from the groups indicated. (B) L-Arg uptake, (C) iNOS protein expression by flow cytometry, and (D) NO2⫺ levels were assessed. (A–D) *P ⬍ .05, **P ⬍ .01, ***P ⬍ .001 versus uninfected control; §§P ⬍ .01, §§§P ⬍ .001 versus H pylori–infected without DFMO. (A) n ⫽ 6 mice per group, (B) n ⫽ 6 –13 mice per group; each symbol represents a different mouse. (E) H pylori colonization and level of gastritis. Left panel, colonization by serial dilution and culture; center panel, quantitative PCR for ureB; right panel, histology scores for gastritis severity. Histologic scores in uninfected mice were 0.8 ⫾ 0.3 (n ⫽ 14) and 0.7 ⫾ 0.3 (n ⫽ 5), in mice receiving water alone or DFMO, respectively. §§P ⬍ .01, §§§P ⬍ .001 versus HP without DFMO. (F) Representative H&E staining of gastritis tissues. Arrowheads indicate crypt abscesses.

BASIC– ALIMENTARY TRACT

November 2010

1696

CHATURVEDI ET AL

BASIC– ALIMENTARY TRACT

negative gastritis or normal tissues and that experimental infection induces ODC expression. Immunostaining in human tissues revealed staining of lamina propria mononuclear cells, and we have shown increased expression of ODC in gastric macrophages from infected mice. It has been reported that there are increased polyamine levels in gastric tissues from H pylori–infected human subjects41 and that spermine levels were increased in intestinal-type tumors, typical of H pylori–induced cancer, compared with diffuse gastric tumors.42 We have found that in isolated mouse gastric macrophages from H pylori–infected mice there is a minimal increase in iNOS protein and NO production compared with cells from uninfected mice. There was also loss of induction of iNOS protein and NO synthesis, and diminished L-Arg uptake upon ex vivo stimulation with H pylori in macrophages from chronically infected mice. These data suggest a form of tolerance in innate immune response, with failure to mount an iNOS-mediated antimicrobial defense. This provides new insight into findings that iNOS-deficient mice do not show an exacerbation of H pylori colonization and that serum nitrate/nitrate is not increased in infected wild-type mice.43 Consistent with our data implicating ODC as an inhibitor of iNOS synthesis in vitro, we show that depletion of endogenous polyamine synthesis by DFMO enhanced the levels of L-Arg uptake, iNOS protein, and NO in gastric macrophages from H pylori–infected mice. The importance of these data was confirmed by the ability of DFMO to reduce H pylori colonization, as well as gastritis. This differs from reports that reduction in colonization requires increased gastritis.19,20,33 Our findings raise the possibility that DFMO could have a direct effect on the bacterium itself. However, H pylori does not possess ODC,44 making it unlikely that the effect of DFMO occurs by altering H pylori polyamine synthesis. Furthermore, we have found the following: (1) H pylori does not show ODC enzymatic activity, in contrast to Escherichia coli; (2) DFMO treatment of H pylori in liquid culture does not impair bacterial viability; and (3) the beneficial effect of DFMO on H pylori colonization and gastritis is effectively attenuated in iNOS knockout mice (data not shown). The reduction in colonization with DFMO is consistent with our findings that killing of extracellular H pylori by macrophages is dependent on NO and regulated by L-Arg availability.6,15,16 However, it is possible that polyamine depletion may have additional effects on immune responses, which is an area of active investigation in our laboratory. The use of the pharmacologic agent DFMO to inhibit ODC has potential limitations compared with a genetic approach. We have used DFMO in our studies because homozygous deletion of ODC in mice is lethal.28

GASTROENTEROLOGY Vol. 139, No. 5

ODC⫹/⫺ mice do survive to adulthood, and reduced lymphomagenesis when crossed to a c-Myc-overexpressing mouse has been shown; these results also occurred with chronic oral DFMO treatment,28 further validating the approach we used in the current study. Moreover, we have shown that DFMO effectively reduced gastric polyamine levels in H pylori–infected mice to the same levels as in uninfected mice. We have obtained the ODC⫹/⫺ mice to determine the effect of ODC heterozygosity on gastric polyamine levels and the immunopathogenesis of H pylori infection. Deleterious effects of DFMO have been reported in acute gastric injury models,45 and DFMO impairs wound repair of intestinal epithelial cells in vitro.46 However, our data show a benefit of decreased H pylori colonization that was accompanied by a decrease in gastric inflammation, and DFMO had no effect on histology in uninfected mice. There is substantial interest in DFMO as a chemoprevention agent in colon carcinogenesis because it inhibits intestinal tumor growth in mouse models31 and was effective in the prevention of recurrent colonic adenomas in combination with sulindac in a human trial.47 Ultimately, DFMO could prove to be a useful adjunctive treatment for H pylori, especially considering problems of antibiotic resistance and issues of reinfection after antibiotic treatment in areas of very high prevalence.

Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at www.gastrojournal.org, and at doi: 10.1053/j.gastro.2010.06.060. References 1. Fox JG, Wang TC. Inflammation, atrophy, and gastric cancer. J Clin Invest 2007;117:60 – 69. 2. Varanasi RV, Fantry GT, Wilson KT. Decreased prevalence of Helicobacter pylori infection in gastroesophageal reflux disease. Helicobacter 1998;3:188 –194. 3. Mai UE, Perez-Perez GI, Allen JB, et al. Surface proteins from Helicobacter pylori exhibit chemotactic activity for human leukocytes and are present in gastric mucosa. J Exp Med 1992;175: 517–525. 4. Wilson KT, Ramanujam KS, Mobley HLT, et al. Helicobacter pylori stimulates inducible nitric oxide synthase expression and activity in a murine macrophage cell line. Gastroenterology 1996;111: 1524 –1533. 5. Gobert AP, Cheng Y, Wang JY, et al. Helicobacter pylori induces macrophage apoptosis by activation of arginase II. J Immunol 2002;168:4692– 4700. 6. Gobert AP, McGee DJ, Akhtar M, et al. Helicobacter pylori arginase inhibits nitric oxide production by eukaryotic cells: a strategy for bacterial survival. Proc Natl Acad Sci U S A 2001;98:13844 – 13849. 7. Meyer F, Ramanujam KS, Gobert AP, et al. Cutting edge: cyclooxygenase-2 activation suppresses Th1 polarization in response to Helicobacter pylori. J Immunol 2003;171:3913–3917.

8. Wilson KT, Crabtree JE. Immunology of Helicobacter pylori: insights into the failure of the immune response and perspectives on vaccine studies. Gastroenterology 2007;133:288 –308. 9. Necchi V, Candusso ME, Tava F, et al. Intracellular, intercellular, and stromal invasion of gastric mucosa, preneoplastic lesions, and cancer by Helicobacter pylori. Gastroenterology 2007;132: 1009 –1023. 10. Forstermann U, Closs EI, Pollock JS, et al. Nitric oxide synthase isozymes. Characterization, purification, molecular cloning, and functions. Hypertension 1994;23:1121–1131. 11. Stuehr DJ, Cho HJ, Kwon NS, et al. Purification and characterization of the cytokine-induced macrophage nitric oxide synthase: an FAD- and FMN-containing flavoprotein. Proc Natl Acad Sci U S A 1991;88:7773–7777. 12. Kakuda DK, Sweet MJ, Mac Leod CL, et al. CAT2-mediated L-arginine transport and nitric oxide production in activated macrophages. Biochem J 1999;340:549 –553. 13. Nicholson B, Manner CK, Kleeman J, et al. Sustained nitric oxide production in macrophages requires the arginine transporter CAT2. J Biol Chem 2001;276:15881–15885. 14. Yeramian A, Martin L, Serrat N, et al. Arginine transport via cationic amino acid transporter 2 plays a critical regulatory role in classical or alternative activation of macrophages. J Immunol 2006;176:5918 –5924. 15. Bussiere FI, Chaturvedi R, Cheng Y, et al. Spermine causes loss of innate immune response to Helicobacter pylori by inhibition of inducible nitric-oxide synthase translation. J Biol Chem 2005; 280:2409 –2412. 16. Chaturvedi R, Asim M, Lewis ND, et al. L-Arginine availability regulates inducible nitric oxide synthase-dependent host defense against Helicobacter pylori. Infect Immun 2007;75:4305– 4315. 17. Fu S, Ramanujam KS, Wong A, et al. Increased expression and cellular localization of inducible nitric oxide synthase and cyclooxygenase 2 in Helicobacter pylori gastritis. Gastroenterology 1999;116:1319 –1329. 18. Mannick EE, Bravo LE, Zarama G, et al. Inducible nitric oxide synthase, nitrotyrosine, and apoptosis in Helicobacter pylori gastritis: effect of antibiotics and antioxidants. Cancer Res 1996; 56:3238 –3243. 19. Eaton KA, Mefford M, Thevenot T. The role of T cell subsets and cytokines in the pathogenesis of Helicobacter pylori gastritis in mice. J Immunol 2001;166:7456 –7461. 20. DeLyria ES, Redline RW, Blanchard TG. Vaccination of mice against H pylori induces a strong Th-17 response and immunity that is neutrophil dependent. Gastroenterology 2009;136:247– 256. 21. Chaturvedi R, Cheng Y, Asim M, et al. Induction of polyamine oxidase 1 by Helicobacter pylori causes macrophage apoptosis by hydrogen peroxide release and mitochondrial membrane depolarization. J Biol Chem 2004;279:40161– 40173. 22. Cheng Y, Chaturvedi R, Asim M, et al. Helicobacter pylori-induced macrophage apoptosis requires activation of ornithine decarboxylase by c-Myc. J Biol Chem 2005;280:22492–22496. 23. Gobert AP, Mersey BD, Cheng Y, et al. Cutting edge: urease release by Helicobacter pylori stimulates macrophage inducible nitric oxide synthase. J Immunol 2002;168:6002– 6006. 24. Verrey F, Closs EI, Wagner CA, et al. CATs and HATs: the SLC7 family of amino acid transporters. Pflugers Arch 2004;447:532– 542. 25. Closs EI, Basha FZ, Habermeier A, et al. Interference of L-arginine analogues with L-arginine transport mediated by the y⫹ carrier hCAT-2B. Nitric Oxide 1997;1:65–73. 26. Bussiere FI, Chaturvedi R, Asim M, et al. Low multiplicity of infection of Helicobacter pylori suppresses apoptosis of B lymphocytes. Cancer Res 2006;66:6834 – 6842. 27. Gobert AP, Cheng Y, Akhtar M, et al. Protective role of arginase in a mouse model of colitis. J Immunol 2004;173:2109 –2117.

L–ARGININE UPTAKE IN H PYLORI INFECTION

1697

28. Nilsson JA, Keller UB, Baudino TA, et al. Targeting ornithine decarboxylase in Myc-induced lymphomagenesis prevents tumor formation. Cancer Cell 2005;7:433– 444. 29. Asim M, Chaturvedi R, Hoge S, et al. Helicobacter pylori induces ERK-dependent formation of a phospho-C-FOS/C-JUN AP-1 complex that causes apoptosis in macrophages. J Biol Chem 2010; 285:20343–20357. 30. Beaumier L, Castillo L, Ajami AM, et al. Urea cycle intermediate kinetics and nitrate excretion at normal and “therapeutic” intakes of arginine in humans. Am J Physiol 1995;269:E884 – E896. 31. Yerushalmi HF, Besselsen DG, Ignatenko NA, et al. Role of polyamines in arginine-dependent colon carcinogenesis in Apc(Min) (/⫹) mice. Mol Carcinog 2006;45:764 –773. 32. Scott Algood HM, Gallo-Romero J, Wilson KT, et al. Host response to Helicobacter pylori infection before initiation of the adaptive immune response. FEMS Immunol Med Microbiol 2007; 51:577–586. 33. Rad R, Brenner L, Bauer S, et al. CD25⫹/Foxp3⫹ T cells regulate gastric inflammation and Helicobacter pylori colonization in vivo. Gastroenterology 2006;131:525–537. 34. Peteroy-Kelly M, Venketaraman V, Connell ND. Effects of Mycobacterium bovis BCG infection on regulation of L-arginine uptake and synthesis of reactive nitrogen intermediates in J774.1 murine macrophages. Infect Immun 2001;69:5823–5831. 35. Mossner J, Hammermann R, Racke K. Concomitant down-regulation of L-arginine transport and nitric oxide (NO) synthesis in rat alveolar macrophages by the polyamine spermine. Pulm Pharmacol Ther 2001;14:297–305. 36. Lopatin AN, Makhina EN, Nichols CG. Potassium channel block by cytoplasmic polyamines as the mechanism of intrinsic rectification. Nature 1994;372:366 –369. 37. Zhang M, Borovikova LV, Wang H, et al. Spermine inhibition of monocyte activation and inflammation. Mol Med 1999;5:595– 605. 38. Moruzzi MS, Marverti G, Piccinini G, et al. The effect of spermine on calcium requirement for protein kinase C association with phospholipid vesicles. Int J Biochem Cell Biol 1995;27:783–788. 39. Endo M, Oyadomari S, Terasaki Y, et al. Induction of arginase I and II in bleomycin-induced fibrosis of mouse lung. Am J Physiol Lung Cell Mol Physiol 2003;285:L313–L321. 40. Konturek PC, Rembiasz K, Konturek SJ, et al. Gene expression of ornithine decarboxylase, cyclooxygenase-2, and gastrin in atrophic gastric mucosa infected with Helicobacter pylori before and after eradication therapy. Dig Dis Sci 2003;48:36 – 46. 41. Linsalata M, Russo F, Notarnicola M, et al. Polyamine profile in human gastric mucosa infected by Helicobacter pylori. Ital J Gastroenterol Hepatol 1998;30:484 – 489. 42. Russo F, Linsalata M, Giorgio I, et al. Polyamine levels and ODC activity in intestinal-type and diffuse-type gastric carcinoma. Dig Dis Sci 1997;42:576 –579. 43. Miyazawa M, Suzuki H, Masaoka T, et al. Suppressed apoptosis in the inflamed gastric mucosa of Helicobacter pylori-colonized iNOS-knockout mice. Free Radic Biol Med 2003;34:1621–1630. 44. Lee MJ, Huang CY, Sun YJ, et al. Cloning and characterization of spermidine synthase and its implication in polyamine biosynthesis in Helicobacter pylori strain 26695. Protein Expr Purif 2005; 43:140 –148. 45. Wang JY, Johnson LR. Luminal polyamines stimulate repair of gastric mucosal stress ulcers. Am J Physiol 1990;259:G584 – G592. 46. Rao JN, Li J, Li L, et al. Differentiated intestinal epithelial cells exhibit increased migration through polyamines and myosin II. Am J Physiol 1999;277:G1149 –G1158. 47. Meyskens FL, McLaren CE, Pelot D, et al. Difluoromethylornithine plus sulindac for the prevention of sporadic colorectal adenomas: a randomized placebo-controlled, double-blind trial. Cancer Prev Res 2008;1:32–38.

BASIC– ALIMENTARY TRACT

November 2010

1698

CHATURVEDI ET AL

Received January 19, 2010. Accepted June 24, 2010. Reprint requests Address requests for reprints to: Keith T. Wilson, MD, Division of Gastroenterology, Hepatology, and Nutrition, Vanderbilt University School of Medicine, 1030C MRB IV, 2215B Garland Avenue, Nashville, Tennessee 37232-0252. e-mail: keith.wilson@ vanderbilt.edu; fax: (615) 343-6229. Conflicts of interest The authors disclose no conflict.

GASTROENTEROLOGY Vol. 139, No. 5

Funding Supported by National Institutes of Health grants R01DK053620, R01AT004821, P01CA028842, and P01CA116087 (to K.T.W.), F31GM083500 and T32CA009592 (to N.D.L.), R01CA051085 and R01CA098454 (to R.A.C.), and T32DK007673 (to D.P.B.), the Flow Cytometry Core of the Vanderbilt University Digestive Disease Research Center grant (P30DK058404), a Merit Review Grant from the Office of Medical Research, Department of Veterans Affairs (to K.T.W.), and the Philippe Foundation (to T.S. and A.P.G.).

BASIC– ALIMENTARY TRACT

November 2010

L–ARGININE UPTAKE IN H PYLORI INFECTION

1698.e1

Supplementary Table 1. Primers Used for PCR Gene name iNOS ODC ODC CAT1 CAT2 CAT2 CAT2 (siRNA studies) ␤-actin 18S DNA UreB

Species

Forward primer (5= – 3=)

Reverse primer (5= – 3=)

Reference

Mouse Mouse Human Mouse Mouse Human Mouse

CACCTTGGAGTTCACCCAGT CAGCAGGCTTCTCTTGGAAC GCTGAGGTTGGTTTCAGCAT GGGTTTATGCCCTTTGGATT ATGGCTTTACAGGGACGTTG GTTGACTGCAGGGGTCATTT TTGCCGTGTGCCTTGTATTA

ACCACTCGTACTTGGGATGC CATGCATTTCAGGCAGGTTA GTTCTGGAATTGCTGCATGA TAAGGCATCATGAGCGTGAG GCGTTAAAGCTGCAGAAACC ACATTTGGGCTGGTCGTAAG CCGTTCTCAGAAGGTGGTTC

10 3,12 This study This study This study This study This study

Mouse Mouse H pylori

CCAGAGCAAGAGAGGTATCC GGACCAGAGCGAAAGCATTTGCC CAACAAATCCCTACAGCTTTTGC

CTGTGGTGGTGAAGCTGTAG TCAATCTCGGGTGGCTGAACGC CCATCAGCAGGGCCAGTT

3,5,6,9–12 This study This study

Supplementary Materials and Methods Bacteria H pylori SS1, a mouse-adapted human strain,1 was grown on Brucella blood agar plates under microaerobic conditions as described.2 Concentrations of bacteria were determined by optical density at 600 nm.2 For some studies, bacteria were lysed with a French press at 20,000 psi.2,3 A multiplicity of infection of 100,2 defined as the ratio of bacteria/eukaryotic cells, was used for the in vitro studies with the lysates. In some studies French-pressed lysates of C rodentium and C jejuni lysates also were used at a multiplicity of infection of 100 as described.4 Additional studies were conducted with live H pylori SS1 at a multiplicity of infection of 10 in direct contact with RAW 264.7 cells, and at a multiplicity of infection of 100 above Transwell (Corning, Inc, Corning, NY) filter supports to separate bacteria from cells.

Mice, Cells, and Culture Conditions Experiments also were conducted with peritoneal macrophages isolated from male C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME). Cells were harvested 5 days after intraperitoneal injection of 0.5 mL Bio-Gel P-100 polyacrylamide beads (Bio-Rad, Hercules, CA).5 For mouse infection studies, C57BL/6 mice were inoculated with 5 ⫻ 108 colony-forming units of H pylori SS1 in 0.1 mL of Brucella broth, or broth vehicle control, every other day 3 times by oral gavage, and gastric tissues were harvested 4 months later.6 Colonization was assessed in the gastric body by homogenization of tissues in Brucella broth, followed by serial dilution and culture on tryptic soy blood agar plates containing antibiotics.7 Colonies were counted and expressed as colony-forming units/g tissue. Histologic scoring was determined by the modified Sydney System8 with acute and chronic inflammation each scored 0 –3 in the antrum and body, and the scores for antrum and body added together for a 0 –12 scale.

Human Subjects Tissues were obtained from human subjects undergoing routine endoscopic examinations under ap-

proved protocols as described.9 Subjects were excluded if they had a history of prior antibiotic treatment for H pylori or if they were taking nonsteroidal anti-inflammatory drugs or aspirin.9 H pylori status was determined by a combination of rapid urease testing and histology, as well as serologic testing in subjects with histologic gastritis without H pylori detection by tissue testing.9

Quantitative PCR for H pylori To assess the colonization by quantitative PCR, H pylori–specific ureB gene (target) and mouse 18S ribosomal DNA (reference) primers were used as listed in Supplementary Table 1. DNA was isolated using the DNeasy Blood & Tissue kit (Qiagen, Valencia, CA), and PCRs were performed using an iQ-5 iCycler (Bio-Rad, Hercules, CA) and iQ super SYBR Mix (Bio-Rad) with 200 nmol/L primers for H pylori–specific ureB and 100 nmol/L for murine 18S ribosomal DNA. Thermal cycling conditions included an initial denaturation step (94°C for 5 min) and 45 cycles (94°C for 30 s, 60°C for 30 s, and 72°C for 30 s). Absolute values were calculated from threshold values of target and reference genes using a standard curve generated with H pylori and mouse genomic DNA. In cases in which 18S ribosomal DNA could not be amplified reliably, these samples were excluded from the analysis.

RT-PCR RAW 264.7 cells were seeded at 2 ⫻ 106 cells per well, and after stimulation total RNA was isolated using TRIzol reagent (Invitrogen). RNA also was extracted from human and mouse tissues by homogenization in TRIzol. RNA extraction from isolated gastric macrophages was performed with a Versagene 96-well RNA purification kit (Gentra Systems, Minneapolis, MN) as described.10 Subsequently, 1 ␮g of RNA was reverse-transcribed using the iScript cDNA synthesis kit (Bio-Rad). Primer sequences are listed in Supplementary Table 1. Semiquantitative PCR for murine iNOS, murine CAT2, and murine and human ODC was conducted with 1 ␮L of cDNA and HotMasterMix (Eppendorf, Westbury, NY). Sense and antisense primers (15 pmol each) and 3 pmol of ␤-actin primers3,6 were used. One PCR cycle consisted of the following: 95°C for

1698.e2

CHATURVEDI ET AL

30 seconds, 60°C for 1 minute, and 72°C for 30 seconds. The total number of cycles was 35. A final elongation step of 7 minutes at 72°C was used for each reaction. PCR products were run on agarose gels containing ethidium bromide, and stained bands were visualized as described.3,6,9,11 Real-time PCR was performed as described.3,5,10,12 The primers sequences are listed in Supplementary Table 1 for mouse iNOS, CAT1, CAT2, and ODC, and human ODC and CAT2. In brief, 1 ␮g of RNA from each sample was reverse-transcribed using 50 units of Superscript II reverse transcriptase. PCRs were performed using an Opticon 2 thermal cycler (MJ Research, Cambridge, MA) and SYBR Green Master Mix (Molecular Probes, Eugene, OR) with 5.4 nmol/L primers for murine iNOS, CAT1, CAT2, and ODC, and human CAT2 and ODC and human/ murine ␤-actin. Thermal cycling conditions included an initial denaturation step (94°C for 2 min) and 40 cycles (94°C for 30 s, 60°C for 30 s, and 72°C for 30 s). Relative expression was calculated from threshold values of target and reference genes.

L-Arg Uptake Studies RAW 264.7 cells (5 ⫻ 105/well in 24-well plates), peritoneal macrophages (1 ⫻ 106/well in 24-well plates), or gastric macrophages (1 ⫻ 104/well in 96-well plates) were plated. Cells were activated with H pylori lysate for 6, 12, 18, or 24 hours, followed by washing with L-Arg–free medium. The uptake studies were initiated by adding 10 ␮L of [14C]-L-Arg (specific activity, 308 mCi/mmol) for 1 minute for RAW 264.7 cells, and for 5 minutes with peritoneal macrophages or gastric macrophages in L-Arg–free medium. After this incubation time, cells were washed with cold L-Arg–free medium supplemented with excess L-Arg (10 mmol/L). Cells were lysed with radioimmunoprecipitation assay buffer and supernatants were collected. Cell lysates were mixed with scintillation fluid and the [14C] content was determined in a scintillation counter. Protein was measured in cell lysates as described.3 L-Arg transport values were expressed as nmol [14C]-L-Arg●min⫺1●mg protein⫺1.

Transient Transfection of CAT2 and ODC siRNA in Macrophages Transfection with siRNA duplexes was performed as described.3,5,12,13 A total of 80 nmol/L CAT2 siRNA targeted to mouse CAT2 nucleotides 748 –766 from the start codon (sense, 5=-GCUAUUAAUAUCCUGGUCCTT3=; antisense, 5=-GGACCAGGAUAUUAAUAGCTG-3=) were transfected into RAW 264.7 cells with LipofectAMINE 2000 (Invitrogen). The same concentration of scrambled control siRNA also was transfected as a control. In some experiments, transfected cells were treated with lactacystin (Calbiochem, San Diego, CA), a proteosomal inhibitor, as described.10 CAT2 mRNA expression was measured 6 hours after activation with H pylori lysate or vehicle control, and RT-PCR was performed using

GASTROENTEROLOGY Vol. 139, No. 5

primers listed in Supplementary Table 1 that flanked the CAT2 gene in the region from 748 to 766 bp. iNOS expression was measured with the primers mentioned earlier. iNOS protein, L-Arg transport, and NO2⫺ were each measured as described earlier, after 24 hours of activation with H pylori lysate. ODC siRNA duplex targeting nucleotides 1980 –1998 were used exactly as described.5,12 RAW 264.7 cells transfected with ODC siRNA or scrambled siRNA were activated with H pylori lysate or vehicle control. After 6 hours, CAT2 expression was assayed by RT-PCR, using the primer set in Supplementary Table 1. NO2⫺, iNOS protein, and L-Arg transport were each measured after 24 hours of activation.

Immunohistochemistry for ODC and CAT2 Immunostaining was performed on human gastric tissues as described.9,14 Polyclonal mouse anti-human ODC antibody (1:400; Santa Cruz Biotechnology, Inc, Santa Cruz, CA) was used followed by goat biotinylated anti-mouse secondary antibody and peroxidase detection. For detection of CAT2, a polyclonal rabbit anti-human antibody (1:400; a gift from Friedrich Paulsen, Department of Anatomy and Cell Biology, Martin Luther University of Halle-Wittenberg, Halle, Germany) was used.15,16

Immunofluorescence Staining for CAT2 and F4/80 Immunofluorescence staining was performed on gastric tissues from mice 4 months after inoculation with H pylori SS1 or broth control. Antigen retrieval was performed as described.14 Tissue sections were incubated overnight at 4°C with an affinity-purified rabbit polyclonal antibody raised against the C-terminus of mouse CAT217 (1:400 dilution, provided by E.I.C.), and rat polyclonal antimouse F4/80 antibody (1:50 dilution; Invitrogen). CAT2 and F4/80 were detected with goat anti-rabbit secondary antibody conjugated to fluorescein isothiocyanate (1:2000 dilution; BD Biosciences, San Jose, CA), and rabbit anti-rat secondary antibody conjugated to tetramethyl rhodamine iso-thiocyanate (1:100 dilution; Invitrogen), respectively. Slides were imaged with a SPOT RT slider camera system (Diagnostics Instruments, Inc, Sterling Heights, MI) on a Nikon E800 microscope (Nickon, Inc, Melville, NY).

Statistical Analysis Results are expressed as means ⫾ standard error. Statistical analysis was performed with Statview version 5.01 (SAS Institute, Cary, NC) and Prism version 5.0b (GraphPad Software, Inc, San Diego, CA). Where 2 groups were compared, the Student t test was used. Data with more than 2 groups were analyzed by analysis of variance and the Student–Newman–Keuls post hoc multiple comparisons test. A P value of less than .05 was considered statistically significant.

November 2010

L–ARGININE UPTAKE IN H PYLORI INFECTION

1698.e3

References 1. Lee A, O’Rourke J, De Ungria MC, et al. A standardized mouse model of Helicobacter pylori infection: introducing the Sydney strain. Gastroenterology 1997;112:1386 –1397. 2. Wilson KT, Ramanujam KS, Mobley HLT, et al. Helicobacter pylori stimulates inducible nitric oxide synthase expression and activity in a murine macrophage cell line. Gastroenterology 1996;111: 1524 –1533. 3. Chaturvedi R, Cheng Y, Asim M, et al. Induction of polyamine oxidase 1 by Helicobacter pylori causes macrophage apoptosis by hydrogen peroxide release and mitochondrial membrane depolarization. J Biol Chem 2004;279:40161– 40173. 4. Asim M, Chaturvedi R, Hoge S, et al. Helicobacter pylori induces ERK-dependent formation of a phospho-C-FOS/C-JUN AP-1 complex that causes apoptosis in macrophages. J Biol Chem 2010; 285:20343–20357. 5. Bussiere FI, Chaturvedi R, Cheng Y, et al. Spermine causes loss of innate immune response to Helicobacter pylori by inhibition of inducible nitric-oxide synthase translation. J Biol Chem 2005; 280:2409 –2412. 6. Gobert AP, Cheng Y, Wang JY, et al. Helicobacter pylori induces macrophage apoptosis by activation of arginase II. J Immunol 2002;168:4692– 4700. 7. Franco AT, Israel DA, Washington MK, et al. Activation of {beta}-catenin by carcinogenic Helicobacter pylori. Proc Natl Acad Sci U S A 2005;102:10646 –10651. 8. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 1996;20:1161–1181. 9. Fu S, Ramanujam KS, Wong A, et al. Increased expression and cellular localization of inducible nitric oxide synthase and cyclooxygenase 2 in Helicobacter pylori gastritis. Gastroenterology 1999;116:1319 –1329. 10. Chaturvedi R, Asim M, Lewis ND, et al. L-Arginine availability regulates inducible nitric oxide synthase-dependent host defense against Helicobacter pylori. Infect Immun 2007;75:4305– 4315. 11. Gobert AP, Mersey BD, Cheng Y, et al. Cutting edge: urease release by Helicobacter pylori stimulates macrophage inducible nitric oxide synthase. J Immunol 2002;168:6002– 6006. 12. Cheng Y, Chaturvedi R, Asim M, et al. Helicobacter pylori-induced macrophage apoptosis requires activation of ornithine decarboxylase by c-Myc. J Biol Chem 2005;280:22492–22496. 13. Xu H, Chaturvedi R, Cheng Y, et al. Spermine oxidation induced by Helicobacter pylori results in apoptosis and DNA damage: implications for gastric carcinogenesis. Cancer Res 2004;64: 8521– 8525. 14. Gobert AP, Cheng Y, Akhtar M, et al. Protective role of arginase in a mouse model of colitis. J Immunol 2004;173:2109 –2117. 15. Jaeger K, Paulsen F, Wohlrab J. Characterization of cationic amino acid transporters (hCATs) 1 and 2 in human skin. Histochem Cell Biol 2008;129:321–329. 16. Jager K, Bonisch U, Risch M, et al. Detection and regulation of cationic amino acid transporters in healthy and diseased ocular surface. Invest Ophthalmol Vis Sci 2009;50:1112–1121. 17. Closs EI, Albritton LM, Kim JW, et al. Identification of a low affinity, high capacity transporter of cationic amino acids in mouse liver. J Biol Chem 1993;268:7538 –7544.

Supplementary Figure 1. Induction of CAT2 and L-Arg uptake in murine peritoneal macrophages, and effect of spermine and L-lysine. Cells were elicited by intraperitoneal injection of Bio-Gel beads, and cells were harvested 5 days later. Macrophages were plated and stimulated with H pylori lysates at a multiplicity of infection (MOI) of 100. (A) The mRNA expression for CAT1 and CAT2 was assessed at 6 hours after H pylori lysate (HPL) stimulation ⫾ spermine (Spm; 12.5 mmol/L) or L-lysine (Lys; 20 mmol/L) by real-time PCR. (B) L-Arg transport assessed as uptake of 14C-L-Arg into cells at the times indicated. (A and B) Data are the mean ⫾ standard error (n ⫽ 3 experiments in duplicate). **P ⬍ .01 versus control; §P ⬍ .05 for HPL ⫹ Spm versus HPL alone.

1698.e4

CHATURVEDI ET AL

Supplementary Figure 2. Effect of inhibition of proteosomal degradation on iNOS protein levels in RAW 264.7 macrophages transfected with CAT2 siRNA. RAW 264.7 cells were stimulated with H pylori lysate (HPL) in the absence and presence of the proteosomal inhibitor, lactacystin (10 ␮mol/L), for the times indicated. (A) Western blot analysis for iNOS and ␤-actin in RAW 264.7 cells transfected with scrambled (Scr) siRNA. (B) Western blot analysis for iNOS and ␤-actin in RAW 264.7 cells transfected with CAT2 siRNA. Similar results were obtained in 2 experiments.

™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™3 Supplementary Figure 3. Effect of transfection of ODC siRNA on L-Arg uptake, and NO production in RAW 264.7 macrophages activated with live H pylori. (A) Cells transfected with scrambled siRNA or ODC siRNA were co-cultured with live H pylori (multiplicity of infection [MOI], 10) for the times indicated. L-Arg transport was measured as in Figure 1 for the conditions marked. *P ⬍ .05, **P ⬍ .01 versus H pylori–stimulated macrophages transfected with scrambled siRNA. LArg transport in unstimulated RAW 264.7 cells transfected with scrambled siRNA was 1.9 ⫾ 0.2 nmol 14C-L-Arg ● min⫺1 ● mg protein⫺1 and in unstimulated RAW 264.7 cells transfected with CAT2 siRNA was 2.2 ⫾ 0.8 nmol 14C-L-Arg ● min⫺1 ● mg protein⫺1. (B) NO2⫺ levels measured at 24 hours after co-culture with live H pylori (MOI, 10). *P ⬍ .05; ***P ⬍ .001 versus macrophages transfected with scrambled siRNA. (C) Live H pylori (MOI, 100) was placed above Transwell filter supports (0.4-mm pore size), and NO2⫺ levels were measured after 24 hours. **P ⬍ .01 versus macrophages transfected with scrambled siRNA. (A–C) Data are from 2 separate experiments performed in duplicate.

GASTROENTEROLOGY Vol. 139, No. 5

November 2010

L–ARGININE UPTAKE IN H PYLORI INFECTION

1698.e5

Supplementary Figure 4. Effect of transfection of ODC siRNA on L-Arg uptake, and NO production in RAW 264.7 macrophages activated with C jejuni or C rodentium. Cells were transfected with scrambled or ODC siRNA and activated with lysates of (A and B) C jejuni and (C and D) C rodentium for the times indicated at a multiplicity of infection (MOI) of 100. (A and C) L-Arg transport, measured as in Figure 1 and Supplementary Figure 1 for the conditions marked. (B and D) NO2⫺ levels. L-Arg uptake levels in unstimulated RAW 264.7 cells transfected with scrambled (Scr) siRNA and ODC siRNA were as listed in Supplementary Figure 3. (A–D) Data are from 2 separate experiments performed in duplicate.

1698.e6

CHATURVEDI ET AL

Supplementary Figure 5. Summary of alterations in host responses in macrophages owing to the effects of ODC-derived spermine on L-Arg uptake. CAT2 is the primary transporter of L-Arg into macrophages activated by H pylori. Extracellular L-Arg must be transported into stimulated macrophages to facilitate iNOS translation that is required to generate sufficient quantities of NO to kill extracellular H pylori. L-lysine acts as a competitive inhibitor that blocks L-Arg uptake and NO production in H pylori–stimulated macrophages. The induction of ODC by H pylori results in the generation of the polyamines putrescine, spermidine, and spermine. Spermine can be back-converted to spermidine, or when produced in large quantities by ODC it blocks the uptake of L-Arg by CAT2, without altering CAT2 mRNA or protein expression. Once in the cell, L-Arg also is used by arginase II to produce L-ornithine, which is used as the substrate for ODC to generate more polyamines. Inhibition of ODC with DFMO blocks the overproduction of polyamines, and, thus, restores iNOS synthesis, NO production, and antimicrobial defense against H pylori.

GASTROENTEROLOGY Vol. 139, No. 5