If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Medicine, Division of Gastroenterology and theVeterans Affairs Maryland Health Care System, Baltimore, Maryland, 21201, and theGreenebaum Cancer Center, University of Maryland School of Medicine, Baltimore, Maryland 21201, the
* This work was supported by National Institutes of Health Grants DK53620 and DK63626 (to K. T. W.) and CA51085 and CA98454 (to R. A. C.), the Office of Medical Research, Department of Veterans of Affairs (to K. T. W.), and the Crohn's and Colitis Foundation of America (to K. T. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ¶ Present address: Unité de Microbiologie, Institut National de la Recherche Agronomique de Clermont-Ferrand-Theix, 63122 Saint-Genès-Champanelle, France.
Helicobacter pylori infects the human stomach by escaping the host immune response. One mechanism of bacterial survival and mucosal damage is induction of macrophage apoptosis, which we have reported to be dependent on polyamine synthesis by arginase and ornithine decarboxylase. During metabolic back-conversion, polyamines are oxidized and release H2O2, which can cause apoptosis by mitochondrial membrane depolarization. We hypothesized that this mechanism is induced by H. pylori in macrophages. Polyamine oxidation can occur by acetylation of spermine or spermidine by spermidine/spermine N1-acetyltransferase prior to back-conversion by acetylpolyamine oxidase, but recently direct conversion of spermine to spermidine by the human polyamine oxidase h1, also called spermine oxidase, has been demonstrated. H. pylori induced expression and activity of the mouse homologue of this enzyme (polyamine oxidase 1 (PAO1)) by 6 h in parallel with ornithine decarboxylase, consistent with the onset of apoptosis, while spermidine/spermine N1-acetyltransferase activity was delayed until 18 h when late stage apoptosis had already peaked. Inhibition of PAO1 by MDL 72527 or by PAO1 small interfering RNA significantly attenuated H. pylori-induced apoptosis. Inhibition of PAO1 also significantly reduced H2O2 generation, mitochondrial membrane depolarization, cytochrome c release, and caspase-3 activation. Overexpression of PAO1 by transient transfection induced macrophage apoptosis. The importance of H2O2 was confirmed by inhibition of apoptosis with catalase. These studies demonstrate a new mechanism for pathogen-induced oxidative stress in macrophages in which activation of PAO1 leads to H2O2 release and apoptosis by a mitochondrial-dependent cell death pathway, contributing to deficiencies in host defense in diseases such as H. pylori infection.
Helicobacter pylori is a Gram-negative, microaerophilic bacterium that selectively colonizes the mammalian stomach and causes gastritis, peptic ulcers, and gastric cancer. Intriguingly the human host mounts a vigorous innate and adaptive immune response, yet this results only in lifelong gastritis without eradication of the organism. H. pylori has evolved several strategies to enhance its own survival in the face of this immune response. For example, we have shown that while the host produces NO derived from inducible nitric-oxide synthase in response to soluble products of H. pylori (
which produces putrescine from l-ornithine, that is then converted to the polyamines spermidine and spermine. H. pylori-induced macrophage apoptosis is NO-independent, but is abrogated by inhibition of either arginase or ODC and restored by addition of spermidine or spermine (
These studies raise the question as to the mechanism of polyamine-driven macrophage apoptosis. While ODC is the rate-limiting enzyme for polyamine synthesis, there are several pathways of polyamine metabolism that are relevant to their biological function in cells. Spermine and spermidine can be metabolized by an oxidative process that results in the release of H2O2. Specifically this can occur in two ways. In the originally identified back-conversion pathway, spermine or spermidine is metabolized by the enzyme spermidine/spermine N1-acetyltransferase (SSAT) to acetylspermine or acetylspermidine, respectively (
). We hypothesized that oxidation of polyamines generated by H. pylori was the key step in causing macrophage apoptosis. We therefore sought to determine the respective roles of the SSAT-APAO versus the PAO1 pathway, the ability of H. pylori to induce these enzymes, and the relationship to apoptosis. This is the first report of any microbe inducing polyamine oxidation specifically via induction of PAO1 expression and activity in macrophages. We will show that a cascade of events ensues in which oxidation of spermine results in H2O2 release, mitochondrial membrane depolarization, cytochrome c release, caspase-3 activation, and apoptosis.
All reagents for cell culture, RNA extraction, and RT-PCR were from Invitrogen. MDL 72527, a PAO inhibitor, was a gift from N. Seiler (Strasbourg, France). All other chemicals were purchased from Sigma.
), was used. Bacteria were passaged on Brucella agar plates containing 10% sheep blood and were maintained under microaerobic conditions. For experiments, H. pylori were harvested from plates, washed twice, and suspended in PBS. H. pylori lysate (HPL) was prepared with a French press in PBS (
), we used HPL for all of the studies presented herein. Concentrations of bacteria were determined by optical density at 600 nm.
Cells and Culture Conditions
The murine macrophage cell line RAW 264.7 was maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, 1 mm sodium pyruvate, and 10 mm HEPES at 37 °C in a humidified 5% CO2 atmosphere. For experiments, cells were washed, and the same medium without penicillin-streptomycin was added 2 h before stimulation. Multiplicity of infection (m.o.i.) was determined as the ratio of bacteria/eukaryotic cells. In the experiments with HPL, the m.o.i. was based on the number of bacteria prior to lysis (
RAW 264.7 macrophages were seeded at 1 × 106/well in 6-well plates. After stimulation, total RNA was isolated using TRIzol reagent. Subsequently 2 μg of RNA from each sample was reverse transcribed using 50 units of Superscript II reverse transcriptase. PCR was conducted using 2 μl of cDNA and 1 unit of Taq DNA polymerase. For ODC, PAO1, APAO, and SSAT 15 pmol each of sense and antisense primers were used with 3 pmol each of β-actin primers in a multiplex reaction (
). One PCR cycle consisted of the following: 94 °C for 1 min, 60 °C for 1.5 min, and 72 °C for 1.8 min. The total numbers of cycles were 30 for PAO1 and SSAT, 32 for APAO, and 20 for ODC. A final elongation step of 10 min at 72 °C was used for each reaction. Sense and antisense primer sequences and PCR product sizes were as follows: murine PAO1, 5′-CACGTGATTGTGACCGTTTC-3′ and 5′-TGGGTAGGTGAGGGTACAGTC-3′, 222 bp; murine APAO, 5′-CTTTTCCAGGGGAGACCTTC-3′ and 5′-CACACCACCTGGATGAACTG-3′, 250 bp; murine SSAT, 5′-GACCCCTGAAGGACATAGCA-3′ and 5′-CCGAAGCACCTCTTCTTTTG-3′, 248 bp; murine ODC, 5′-CAGCAGGCTTCTCTTGGAAC-3′ and 5′-CATGCATTTCAGGCAGGTTA-3′, 602 bp; and murine/human β-actin, 5′-CCAGAGCAAGAGAGGTATCC-3′ and 5′-CTGTGGTGGTGAAGCTGTAG-3′, 436 bp. PCR products were run on 2% agarose gels with 0.4 μg/ml ethidium bromide. Stained bands were visualized under UV light and photographed with a digital gel documentation system (EDAS 290 and 1D software, Eastman Kodak Co.).
Real Time PCR
2 μg of RNA from each sample was reverse transcribed using 50 units 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 nm primers for murine PAO1, APAO, SSAT, or ODC and β-actin as listed above. 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 (CT) of target and reference genes.
Assay for ODC Activity
ODC activity was determined by a radiometric analysis in which the amount of 14CO2 liberated from l-[14C]ornithine was measured as described previously (
5 × 105 RAW 264.7 cells in 24-well plates were stimulated with HPL at m.o.i. of 100. Enzyme activities were assayed in cell homogenates by a modification of a chemiluminescence assay as described previously (
). Luminol-dependent chemiluminescence was determined using a Monolight 3010 luminometer with two reagent injectors. Luminol was prepared as a 100 mm stock solution in Me2SO and diluted to 100 nm with H2O immediately prior to use. Macrophages were scraped into 500 μl of 80 mm borate buffer, pH 9.0, and homogenized with an Ultra-Turrax (IKA Works, Wilmington, NC). For the assay of PAO1 activity, 100 μl of cell lysate was added to 100 μl of 80 mm borate buffer, pH 9.0, containing 570 milliunits of horseradish peroxidase and 250 μm spermine. All reagents were combined and incubated for 2 min at 37 °C, and then the tube was transferred to the luminometer. Luminol (5 nmol) was added, and the resulting chemiluminescence was integrated over 20 s. The integral values were calibrated against standards containing known concentrations of H2O2, and the activities were expressed as nmol of H2O2/mg of protein/min. The assay was linear in the range of 3-89 nmol of H2O2. All of the samples were diluted appropriately to be assayed in the linear range.
For APAO activity, cell homogenates were prepared in borate buffer, pH 9.0, exactly as above. 100 μl of the cell lysate was added to 100 μl of 80 mm borate buffer, pH 9.0, containing 570 milliunits of horseradish peroxidase and 250 μmN1-acetylspermine (Fluka Chemie, Buchs, Switzerland) as substrate. This assay mixture was incubated for 2 min at 37 °C, and then the tube was transferred to the luminometer. Luminol (5 nmol) was added, and the resulting chemiluminescence was integrated over 20 s. Integral values were calibrated and expressed exactly as for PAO1 above with the same linear range.
Determination of SSAT Activity
SSAT activity was determined in RAW 264.7 cell extracts by an assay that measures the formation of l-[14C]acetylspermidine from l-[14C]acetyl-CoA (PerkinElmer Life Sciences) in 10 min at 30 °C as described previously (
). The assay mixture contained 50 mm Tris-HCl (pH 7.8), 3 mm spermidine, and 12.7 μm (specific activity, 63 mCi/mmol) l-[14C]acetyl-CoA in a total volume of 100 μl. The reaction was stopped by adding 10 μl of 1 m NH2-OH·HCl and boiling for 3 min. The resulting samples were spotted onto P-81 phosphocellulose discs and counted in a liquid scintillation counter. Enzyme activity was expressed as nmol of l-[14C]acetylspermidine formed/min/mg of protein.
Polyamine levels were determined by precolumn dansylation reverse phase high performance liquid chromatography as reported previously (
Annexin V-FITC Staining—RAW 264.7 cells (0.5 × 106 cells/well) were cultured in 24-well plates in the presence of HPL for 6-24 h. In some experiments RAW 264.7 cells were cultured in the presence or absence of the H2O2-detoxifying enzyme catalase (250-1000 units/ml) or polyethylene glycol (PEG)-catalase (25-250 units/ml) for 24 h. Apoptosis was assayed using an annexin V-FITC apoptosis detection kit (Oncogene Research Products, San Diego, CA). Cells were washed with PBS and resuspended in binding medium and stained with annexin V-FITC. Cells were incubated for 30 min at room temperature in the dark and counterstained with 10 μl of propidium iodide (PI; 30 μg/ml). 1 × 104 cells were analyzed with a flow cytometer (FACSCalibur, BD Biosciences). Spectral overlap was electronically compensated using single color control cells stained with PI or FITC. Analysis of the multivariate data was performed with CELLQuest™ software (BD Biosciences). The upper right (annexin V+/PI+) quadrant represents late apoptotic cells, and the lower right (annexin V+/PI-) quadrant represents early apoptotic cells, while the upper left (annexin V-/PI+) and lower left (annexin V-/PI-) quadrants represent necrotic and viable cell populations, respectively.
Apoptosis by DNA Histogram Analysis—RAW 264.7 cells or peritoneal macrophages (0.5 × 106 cells/well in 24-well plates) were stimulated with HPL for 24 h in the presence and absence of MDL 72527, transfected PAO1 siRNA, scrambled siRNA, or transfected PAO1 cDNA. Both adherent and floating cells were collected and fixed in 70% ethanol. Low molecular weight DNA was extracted with 0.2 m phosphate citrate buffer (pH 7.8), and the remaining cells were stained with PI (
) using a cell cycle analysis kit (Roche Applied Science). Cells were analyzed by flow cytometry. 1.5 × 105 cells were acquired and analyzed with ModFit LT™ software (BD Biosciences) using the sub-G0/G1 peak as the apoptotic population (
DNA Nick-end Labeling of Cells—Macrophages (5 × 104/well) were cultured in 4-well plastic chambered slides (Nunc, Naperville, IL) with or without HPL (m.o.i. of 30) and inhibitors for 24 h. Cells were then washed with PBS and fixed with 4% formaldehyde. DNA fragmentation was analyzed by TUNEL assay as described previously (
) using an in situ apoptosis detection kit (Trevigen, Gaithersburg, MD). The percentage of apoptotic cells was determined after counting of 10 microscope fields at a magnification of 400×.
Determination of Cell Viability
Viability of RAW 264.7 cells was estimated by a colorimetric assay with the cell proliferation kit II (Roche Applied Science). The tetrazolium salt 2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-2H-tetrazolium-5-carboxyaniline (XTT), which is metabolized to formazan by intact mitochondrial dehydrogenases, and electron coupling reagent were added after 20 h and left in the medium for the remaining 4 h of the stimulation period. The viability of cells was estimated on the basis of formazan formed, which was detected spectrophotometrically.
Transient Transfection of PAO1 siRNA in Macrophages
siRNA duplexes were utilized that targeted mouse PAO1 nucleotides 467-487, numbered from the start codon (sense, 5′-GGACGUGGUUGAGGAAUUC-3′; antisense, 5′-CCUGCACCAACUCCUUAAG-3′). Scrambled control siRNA that has no sequence homology to any known genes was used as the control. 20 μl of the 20 μm stock duplex siPAO1 or control scrambled siRNA was mixed with 100 μl of optiMEM medium (Invitrogen). This mixture was gently added to a solution containing 5 μl of LipofectAMINE 2000 (Invitrogen) in 100 μl of optiMEM. The solution was incubated for 30 min at room temperature and gently overlaid onto 90% confluent RAW 264.7 cells in 1 ml of optiMEM for 18 h. Medium was changed, and cells were incubated for 6 h in Dulbecco's modified Eagle's medium. Cells were treated with HPL for 6-24 h. After 6 h RNA was isolated, and PAO1 mRNA expression was assessed by RT-PCR using primers that flank the target sequence as follows: sense, 5′-CAATGGCCTTTTGGAAGAGA-3′; and antisense, 5′-TTACCATGCCGGAAGAACTC-3′. Apoptosis and PAO1 enzyme activity assays were performed after 24 h of stimulation with or without HPL.
Transient Transfection of PAO1 in Macrophages
RAW 264.7 cells were cotransfected with 400 ng of pSV-β-galactosidase and 200 ng of pcDNA3.1-PAO1 using LipofectAMINE Plus (Invitrogen) and optiMEM medium. Cell culture medium was changed 6 h after transfection to complete Dulbecco's modified Eagle's medium, and cells were stimulated with HPL for 24 h. Transfected cells were stained with PI and analyzed for apoptosis and PAO1 activity as described above. Transfection efficiency was calculated by measuring the β-galactosidase activity in transfected cells.
Flow Cytometric Detection of H2O2 by CM-H2DCFDA
To measure intracellular H2O2 we used the cell-permeable redox-sensitive dye CM-H2DCFDA (Molecular Probes); the nonfluorescent reduced form is converted to the fluorescent form when oxidized, allowing detection by flow cytometry (
). RAW 264.7 cells were treated for 1-6 h with HPL with or without MDL 72527, catalase, or PEG-catalase, washed with PBS, and treated with 10 μm CM-H2DCFDA for 20 min at 37 °C. 1 × 105 cells were analyzed on a BD Biosciences FACScan for changes in fluorescence (
5 × 105 cells were plated in 24-well plates. Cells were stimulated with HPL for 6 h, washed with cold PBS, and incubated with a reaction mixture containing 50 μm Amplex Red reagent (Molecular Probes) and 0.1 unit/ml horseradish peroxidase in Krebs-Ringer phosphate buffer for 10 min. Plates were read with a fluorescence microplate reader with excitation of 530 nm and emission detection at 590 nm, and the H2O2 level was determined by using a standard curve with varying dilutions of H2O2 (
The electron gradient across the mitochondrial membrane space during normal respiration is Δψm. Loss of Δψm (depolarization) was measured by flow cytometry after staining with a MitoCapture™ kit (Calbiochem) according to the manufacturer's protocol. The cationic dye fluoresces red as it aggregates inside healthy mitochondria. In apoptotic cells, if the Δψm collapses, the dye stays as a monomer in the cytoplasm and emits green light.
Immunocytochemistry for Cytochrome c
RAW 264.7 cells were plated on slide chambers (1 × 104 cells/well) and stimulated with HPL with or without MDL 72527 for 18 h. Cells were fixed with 4% paraformaldehyde for 15 min and permeabilized with 100% chilled methanol for 4 min on ice. Slides were blocked with 5% goat serum in PBS for 60 min at room temperature and then incubated with a monoclonal anti-mouse cytochrome c antibody (Oncogene) overnight at 4 °C. A rabbit anti-mouse secondary antibody conjugated to FITC was used, and staining was visualized with a Nikon ECLIPSE E 800 microscope.
Immunoblotting for Cytochrome c
RAW 264.7 cells were treated with HPL with or without MDL 72527 and incubated for 18 h. Mitochondrial and cytoplasmic fractions were prepared with a cytosolic/mitochondrial fractionation kit (Oncogene). Protein concentration was determined by the method of Bradford (
), and cytosolic and mitochondrial fraction proteins (40 μg/well) were separated on 16% SDS Tris-HCl gels and transferred to polyvinylidene difluoride membranes (Bio-Rad) by semidry electrotransfer. Membranes were blocked for 2 h at room temperature with PBS containing 0.1% Tween and 5% nonfat dry milk and incubated overnight with anti-cytochrome c antibody at 1:200 dilution. This was followed by a rabbit anti-mouse polyclonal antibody conjugated to horseradish peroxidase (1:5000 dilution). Chemiluminescence detection was performed using the SuperSignal West Pico chemiluminescent substrate (Pierce) and exposure to Hyperfilm ECL (Amersham Biosciences).
Measurement of Caspase-3 Activity
Caspase-3 activity was measured by the cleavage of the chromogenic tetrapeptide (Ac-DEVD-p-nitroanilide) using a kit from Calbiochem. In brief, 1 × 106 cells were lysed, combined with substrate in the caspase reaction mixture, and incubated at 37 °C in the incubation chamber of a SPECTRAmax® PLUS microplate reader (Amersham Biosciences). Absorbance was read at 405 nm for 3 h at intervals of 10 min. The conversion factor for the microplate reader was calculated with 100 μl of 50 μmp-nitroaniline. Caspase-3 activity was expressed as pmol/min/mg of protein.
For comparisons between multiple groups, the Student-Newman-Keuls test was used, and for single comparisons between two groups, Student's t test was used. Statview version 5.01 (SAS Institute, Cary, NC) for the Macintosh was used.
HPL Induces Gene Expression and Enzyme Activity of ODC, PAO1, and SSAT but Not APAO in Macrophages—We used murine RAW 264.7 macrophages that we have used previously to demonstrate H. pylori-induced apoptosis (
). Macrophages were stimulated with HPL at m.o.i. of 100 for 6 h, and mRNA expression was assessed by RT-PCR. As shown in Fig. 1A, levels of ODC, PAO1, and SSAT were up-regulated, while APAO was not induced. This pattern was confirmed by real time PCR analysis (Fig. 1B) with -fold increases of 11.9 ± 2.8 (p < 0.01), 3.7 ± 0.3 (p < 0.001), and 3.8 ± 0.3 (p < 0.001) for ODC, PAO1, and SSAT, respectively, compared with unstimulated macrophages. In contrast, APAO levels were slightly decreased by 0.8 ± 0.2-fold.
Activity of each enzyme was measured 6-24 h after stimulation with HPL. There was a biphasic increase in activity of ODC (Fig. 2A) with a rapid peak at 6 h (5.9 ± 0.3-fold increase) followed by a decline at 12 h and a second peak at 18 h. PAO1 activity (Fig. 2B) was significantly increased at 6 and 12 h with 3.9 ± 0.9- and 6.1 ± 0.1-fold increases, respectively, with peak activity at 18 h (11.7 ± 0.2-fold increase). In contrast, SSAT was not increased until 18 h (5.3 ± 1.2-fold increase), and APAO was not induced at any of the time points.
To further explore the significance of the H. pylori-induced enzyme activation, we determined the putrescine, spermidine, and spermine levels at 6-24 h after HPL activation (Fig. 2C). There was no detectable level of putrescine up to 12 h, but there was a marked increase at 18 h that was then significantly reduced from this level at 24 h. In contrast, there was an increase in spermidine at 6 and 12 h and an increase in spermine at 12 h followed by a clear decline in these two polyamines that was inversely proportional to the increase in putrescine from 12-24 h. These polyamine data are consistent with the back-conversion of spermine to spermidine that is mediated by the induction of PAO1 from 6-18 h. The subsequent increase in putrescine is likely due to the increase in SSAT activity and subsequent back-conversion of acetylated polyamines by APAO.
Time Course for HPL-induced Macrophage Apoptosis Correlates with Activation of ODC and PAO1 Activities—Because polyamine synthesis and oxidative catabolism have been implicated in the apoptosis of epithelial cell lines, we compared the time course of ODC and PAO1 activation to that of apoptosis in H. pylori-stimulated macrophages. By using annexin V plus PI labeling of live cells, we measured both early and late apoptosis. Fig. 3A indicates that early apoptosis, representing annexin V+/PI- cells (Fig. 3B, right lower quadrants in density plots) was significantly increased by 2.1 ± 0.2-fold at 6 h and by 5.6 ± 0.3-fold at 12 h after HPL stimulation as compared with control. As the early apoptosis decreased from 12 to 24 h, there was a concomitant increase in late apoptosis that peaked at 18 h, an indication that early apoptosis was truly representative of a progressive apoptotic process. The presence of apoptosis beginning at 6 h is consistent with the initial spike in ODC activity at 6 h and the increase in PAO1 activity at this time point and argues against an important role of the SSAT-APAO pathway in the apoptosis since SSAT activity did not increase until 18 h after stimulation (Fig. 2B).
Macrophage Apoptosis Is Dependent on Polyamine Oxidation—Because the time course of induction of gene expression and enzyme activity of PAO1 and polyamine back-conversion correlated with apoptosis, we determined the effect of an inhibitor of PAO, MDL 72527 (
). To quantify the end point of apoptosis we used the sub-G0/G1 peak of PI-stained fixed cells analyzed by flow cytometry to measure the apoptosis in these experiments. As shown in Fig. 4, A and C, there was a 14.6 ± 1.7-fold increase in apoptosis at 24 h after HPL stimulation, and MDL 72527 inhibited this apoptosis by 35.5 ± 8.8% at 25 μm and by 83.5 ± 3.7% at 250 μm. To verify that the apoptosis analysis correlated with cell death, we measured cell viability (Fig. 4B) and found that HPL reduced macrophage survival by 54.3 ± 1.3%, and MDL 72527 restored cell viability in a concentration-dependent manner.
Because we had found that the increase in early apoptosis in Fig. 3 paralleled the increase in PAO1 activity in Fig. 2B, we further assessed this correlation by determining the effect of MDL 72527 on the early apoptosis between 6 and 24 h after stimulation with HPL. As shown in Fig. 4D, the increase in early apoptosis (annexin V+/Pi- cells) at 6-24 h was significantly attenuated by MDL 72527. Since SSAT is not induced until 18 h, the inhibition of the apoptosis by PAO inhibition at 6 and 12 h must be due to inhibition of PAO1. This provides additional evidence that the induction of macrophage apoptosis in response to H. pylori is due to PAO1 rather than SSAT-APAO.
We next determined whether the H. pylori results in the RAW 264.7 cell line occurred in non-transformed macrophages. As shown in Fig. 4E, when freshly isolated mouse peritoneal macrophages were stimulated with HPL, there was a 5.7 ± 1.1-fold increase in apoptosis that was inhibited by 74.5 ± 3.1% with MDL 72527, indicating that the same mechanism of apoptosis is occurring in these cells.
Finally we sought to directly address whether PAO1 is essential to H. pylori-induced macrophage apoptosis. We therefore specifically inhibited PAO1 by transiently transfecting RAW 264.7 cells with PAO1 siRNA and compared results to a scrambled control siRNA. As shown in Fig. 5, PAO1 siRNA markedly decreased the HPL-stimulated PAO1 mRNA expression (Fig. 5A) and completely blocked the induction of PAO1 enzyme activity (Fig. 5B). When apoptosis was assessed (Fig. 5C), the PAO1 siRNA completely abolished the HPL-induced apoptosis and restored the cell viability to above unstimulated control levels (Fig. 5D). Taken together, these data show that knock-down of stimulated PAO1 expression eliminated HPL-induced macrophage apoptosis. Since the inhibition of apoptosis was complete, we have provided direct evidence that PAO1, specifically, is responsible for the apoptosis.
Overexpression of PAO1 Induces Apoptosis in Macrophages—To provide further proof of principle concerning the causal role of PAO1 in macrophage apoptosis, we transiently transfected RAW 264.7 cells with a full-length cDNA for PAO1. When compared with cells transfected with vector alone, there was a 7.3 ± 0.9-fold increase in PAO1 activity in transfected cells that was nearly identical to the 7.4 ± 1.3-fold increase with HPL stimulation of neotransfected cells (Fig. 6A). When apoptosis was assessed by DNA histogram analysis in these cells (Fig. 6B), there was a 7.2 ± 0.5-fold increase in apoptosis with PAO1 transfection that exactly paralleled the increase in enzyme activity and was comparable to the increase in apoptosis with HPL stimulation of neotransfected cells.
H. pylori-induced PAO1 Results in H2O2 Generation in Macrophages and Apoptosis—Since we had demonstrated that the time course of activation of PAO1 paralleled that of the onset of apoptosis, inhibition of PAO1 by a pharmacologic inhibitor or by PAO1 siRNA markedly attenuated apoptosis, and PAO1 overexpression caused apoptosis, we sought to directly determine whether H2O2 released by PAO1 activation was responsible for the apoptosis. First we measured intracellular H2O2 levels in response to HPL by flow cytometry (Fig. 7) in the presence of MDL 72527 or the H2O2-detoxifying enzyme catalase. We assessed time points from 0-6 h because we had identified that there was significant induction of PAO1 and early apoptosis at 6 h, and later time points may have nonspecific release of H2O2 from apoptosis. As shown in Fig. 7, A and B, there was significant inhibition of HPL-stimulated H2O2 generation at 4 and 6 h by MDL 72527. Additionally there was significant attenuation of H2O2 generation by catalase at 6 h. Because the CM-H2DCFDA may also detect other oxyradicals, we performed the Amplex Red assay that specifically measures H2O2 in solution. As shown in Fig. 7C, supernatant levels of H2O2 were increased by 2-fold by HPL stimulation, and this increase was inhibited by 96.2 ± 2.4% by MDL 72527 and by 86.9 ± 3.6% by catalase. The increased efficiency of H2O2 reduction by catalase in the extracellular assay compared with the intracellular assay data in Fig. 7A most likely reflects the limited uptake of catalase by cells. To further address this, we have used cell-permeable PEG-catalase and found that there was an 80% inhibition of HPL-stimulated intracellular H2O2 generation in the flow cytometric assay (data not shown). Taken together, these data indicate that polyamine oxidation is a major source of oxidative stress and H2O2 generation specifically in H. pylori-stimulated macrophages.
To directly implicate the H2O2 generation in the apoptosis, we measured the effect of catalase on HPL-stimulated apoptosis. To avoid the confounding effects of H2O2 released nonspecifically during the late phase of apoptosis, we assayed apoptosis with annexin V and PI in live cells, as in Fig. 3, and used an m.o.i. of 30. As shown in Fig. 8, A and B, catalase inhibited apoptosis in a concentration-dependent manner with 24.9 ± 7.0, 52.6 ± 4.0, and 83.5 ± 3.1% inhibition for 250, 500, and 1000 units/ml, respectively. We also assessed apoptosis in the presence of cell-permeable PEG-catalase and found that there was again a concentration-dependent inhibition of apoptosis with 100% inhibition at 250 units/ml. These studies indicate that H2O2 generation is a major cause of HPL-induced apoptosis.
HPL-induced PAO1 Activity Results in Mitochondrial Membrane Depolarization, Cytochrome c Release from Mitochondria to Cytosol, and Caspase-3 Activation—Because depolarization of Δψm has been implicated in apoptosis, we determined the ability of H. pylori to cause this event. As shown in Fig. 9A, there was a significant decrease in Δψm at 12 h that peaked at 18 h. Depolarization of Δψm was significantly inhibited by MDL 72527 (by 81-100%) at the time points from 12 to 24 h (Fig. 9A). To confirm activation of the mitochondrial apoptosis pathway, release of cytochrome c from mitochondria to cytosol was assessed by immunoblotting. As shown in Fig. 9B, with HPL stimulation there was a significant decrease in mitochondrial cytochrome c and a concomitant increase in cytoplasmic cytochrome c, indicating translocation of cytochrome c from mitochondria to cytosol. In cells treated with MDL 72527, there was inhibition of both the decrease in mitochondrial levels and the increase in cytosolic levels of cytochrome c, indicating the prevention of the translocation of cytochrome c. These findings were confirmed by immunohistochemistry. As shown in Fig. 9C, unstimulated cells exhibited a pattern of punctate staining of perinuclear organelles (mitochondria), while HPL activation resulted in intense and diffuse cytosolic staining for cytochrome c (Fig. 9D). Activation in the presence of MDL 72527 (Fig. 9E) resulted in an appearance similar to the control cells with a restoration of punctate staining of perinuclear organelles indicating inhibition of cytochrome c release from mitochondria to cytosol. Cytochrome c released from mitochondria activates caspase-9, which ultimately activates caspase-3, a final step in activation of exonucleases and apoptosis (
). Upon activation with HPL, macrophage caspase-3 activity increased significantly by 8.3 ± 2.3- and 13.5 ± 1.5-fold at 18 and 24 h, respectively (Fig. 9F). As shown in Fig. 9G, MDL 72527 significantly reduced caspase-3 activity by 86.0 ± 5.6%.
Confirmation by TUNEL Assay That Polyamine Oxidation Regulates H. pylori-stimulated Apoptosis—To directly visualize apoptosis, we performed TUNEL staining (Fig. 10). There was a significant increase in cellular changes consistent with apoptosis with HPL stimulation (Fig. 10B) when compared with the rare apoptotic cells in the unstimulated macrophages (Fig. 10A). In the presence of MDL 72527 (Fig. 10C) or catalase (Fig. 10D), there was a marked reduction in apoptosis of 68.8 ± 4.0 and 74.1 ± 5.7%, respectively (Fig. 10E), similar to our previously reported results with α-difluoromethylornithine in this assay (
). Intriguingly we had identified that addition of spermine or spermidine alone or to H. pylori-stimulated cells treated with the ODC inhibitor α-difluoromethylornithine caused apoptosis, but addition of putrescine had no such effect. This led us to speculate that a product of spermine or spermidine metabolism was required to cause apoptosis. Although regulation of polyamine catabolism was previously attributed to the inducible enzyme SSAT, the recently cloned, inducible enzyme PAOh1/spermine oxidase has now been shown to also be an important regulator of polyamine catabolism (
). Our current data directly implicate PAO1 induction by H. pylori and represent the first demonstration of the induction of polyamine oxidation as an important component of microbial pathogenesis. Furthermore we now demonstrate that induction of ODC alone is not sufficient to induce apoptosis in H. pylori-stimulated macrophages and that PAO1 activation is required.
H. pylori is considered a non-invasive organism because it lives in the mucus layer of the stomach where it can adhere to gastric epithelial cells and induce cytoskeletal rearrangements and signaling events such as NF-κB activation and proinflammatory interleukin-8 secretion (
). Because we have shown that contact of live bacteria is not required to induce macrophage apoptosis, we utilized H. pylori lysates in the current study to provide a standardized preparation for our studies and to mimic the exposure of mucosal macrophages to H. pylori components rather than intact bacteria.
We have previously demonstrated induction of ODC activity with H. pylori stimulation at 24 h (
), and we now conducted time course studies revealing that the activity actually peaks at an early time point of 6 h after stimulation. Additionally, by using real time PCR, we demonstrated a 10-fold increase in ODC mRNA expression. In preliminary studies using an ODC promoter construct (
) cloned into a luciferase reporter system, we have observed >8-fold increases in ODC promoter activity with H. pylori stimulation at 6 h, and we also have observed that there is a time-dependent, parallel increase in ODC mRNA and enzyme activity between 0 and 6 h (data not shown). Thus, there is an early induction of ODC that primes the system for induction of apoptosis by providing substrate for release of H2O2 by the polyamine oxidation that we have described in this report. A similar early induction of ODC has been reported in lipopolysaccharide-activated peritoneal macrophages (
). In our study, ODC activity transiently peaked at 6 h, declined at 12 h, and had a second peak at 18 h. It is likely that the biphasic pattern of the ODC activity is related to the variation in the intracellular polyamine levels that we observed. An important question that arises is why is there an increase in spermine from 6 to 12 h at the time that PAO1 is increasing. It is important to realize that at 6 h when ODC activity is highest there is no increase in spermine but an increase in spermidine and that the increase in spermine at 6-12 h is small. Both of these findings are indicative of back-conversion of spermine to spermidine by PAO1 because PAO1 is significantly increased at both 6 and 12 h above control levels, while SSAT is not increased until 18 h.
An additional issue that we addressed was the relative importance of APAO versus PAO1 in H. pylori-induced polyamine-mediated apoptosis. Our data clearly show that PAO1 mRNA levels were increased by 6 h, and there was a corresponding increase in enzyme activity at 6 and 12 h. Recently cloned splice variants of the human PAOh1 gene, now recognized as PAO1, have been found to be inducible by specific polyamine analogues in lung cancer cell lines (
). In contrast, neither APAO mRNA levels nor activity was induced by H. pylori, and while SSAT mRNA was up-regulated at 6 h, the enzyme activity was not increased until 18 h, suggesting a level of posttranscriptional inhibition. Superinduction of SSAT (>1000-fold increases in activity) by antitumor polyamine analogues has been shown to be cytotoxic in epithelial cell lines, and inhibition of SSAT with siRNA to 100-fold increases was recently shown to be sufficient to prevent apoptosis (
). Therefore, the 5-fold increase in SSAT we observed with H. pylori-stimulation in the absence of APAO induction is less likely to have a causative role in the apoptosis than the induction of PAO1 activity. When enzymatic activities were compared in macrophage lysates, PAO1 produced >50-fold more H2O2 than SSAT-APAO. Furthermore SSAT activation in the range we observed is more likely to be a protective mechanism in cells producing acetyl derivatives for export or for recycling when polyamines are present in excess (
). Although APAO is generally considered a constitutively expressed enzyme whose activity is rate-limited by the availability of its acetylated substrate, it can be induced by specific polyamine analogues (
). However, in the current studies, APAO was not up-regulated by H. pylori. Our data showing that MDL 72527 significantly inhibited early apoptosis at 6 and 12 h after stimulation, when PAO1 activity is increased but SSAT-APAO is not, provides additional evidence that PAO1 mediates the induction of macrophage apoptosis by H. pylori.
An issue in our studies was that MDL 72527 is a specific inhibitor of polyamine oxidases but not selective for PAO1. MDL 72527 was originally utilized as an inhibitor of the form of PAO now termed APAO because it has a preference for acetylated polyamines as substrate (
). We interpret our findings of 35% inhibition of apoptosis at 25 μm and 75-83% inhibition at 250 μm to be consistent with the incomplete inhibition of PAO1 that has been reported with MDL 72527 in cell lines (
). We addressed the lack of specificity of MDL 72527 by conducting studies with an siRNA that effectively knocked down PAO1 mRNA expression and enzyme activity and found that the siRNA completely prevented H. pylori-stimulated apoptosis. Therefore, we conclude that PAO1, rather than SSAT-APAO, plays the major role in H. pylori-induced apoptosis. Further support for the role of PAO1 in apoptosis comes from our finding of induction of apoptosis with transfection of PAO1, indicating that PAO1 alone is sufficient to induce apoptosis in macrophages in the presence of sufficient spermine substrate. Intriguingly it has been reported that transfection of PAO1 increases ODC activity, but transfection of APAO does not have this effect (
). It is possible that induction of PAO1 by H. pylori contributes to the activation of ODC, potentially by inhibition of negative feedback of spermine on ODC. Ultimately this interaction contributes to the H2O2 generation and apoptosis.
We used a flow cytometric assay using CM-H2DCFDA dye to measure intracellular H2O2. Because H2O2 is also produced by dismutation of superoxide (
) it is possible that we could have measured other reactive oxygen species by this technique. Therefore, we also used the Amplex Red assay, which is highly specific for the detection of H2O2 in the supernatant, to confirm our observations with the CM-H2DCFDA. The fact that MDL 72527 blocked the H2O2 production in both assays strongly suggests that the major source of H2O2 is from polyamine oxidation and not from other sources. Similar to these findings, MDL 72527 has been shown to inhibit H2O2 generation in NCI H157 cells treated with polyamine analogues (
). Our results demonstrate that H. pylori-induced apoptosis was attenuated with catalase and abolished with cell-permeable PEG-catalase. Exogenous catalase has been shown to inhibit ceramide-induced reactive oxygen species generation and apoptosis in RAW 264.7 cells (
), H. pylori induced a significant depolarization of macrophage mitochondria. Translocation of cytochrome c into the cytosol from the mitochondria has been implicated in apoptosis because of its ability to activate the caspase cascade by binding to Apaf-1 (
). H. pylori induced translocation of cytochrome c from mitochondria to cytosol that was blocked by MDL72527, implicating this event in polyamine oxidation-induced apoptosis.
Since bacteria are known to produce polyamines and serum contains amine oxidases that could be capable of oxidizing polyamines, it is conceivable that H. pylori-derived polyamines could themselves contribute to the apoptosis we have observed. However, we have measured putrescine, spermidine, and spermine levels from amounts of H. pylori equivalent to those used in the present studies for stimulation of macrophages and found no detectable polyamine levels. Additionally we have reported that H. pylori readily induces apoptosis in macrophages in the absence of serum (
). Together these points indicate that macrophage apoptosis is due to oxidation of polyamines generated by the host cells and not the stimulating bacteria. Our findings are not limited to the murine system since we have observed that H. pylori induces loss of cell viability in human U937 monocytes that is blocked by MDL 72527 (data not shown). Additionally our findings may be specific to H. pylori since we have found that the Gram-negative enteric pathogen Citrobacter rodentium, which causes colitis in mice (
), stimulates apoptosis in intestinal epithelial cells and macrophages in vitro that is not attenuated by MDL 72527 (data not shown).
The persistence of H. pylori in the human stomach for the life of the host and the chronic gastritis and risk for gastric cancer that ensues all stem from an ineffective immune and inflammatory response in the mucosa. Our current studies provide new insight into the apoptosis of macrophages that contributes to the dysregulated immune response. Specifically we have shown that polyamine oxidation by the induction of PAO1 mediates the macrophage apoptosis. We have also observed that this same mechanism of apoptosis occurs in gastric epithelial cells
), but our present report provides new evidence that up-regulation of polyamine oxidation is likely to have deleterious effects in mucosal host defense in the setting of exposure to selected pathogens and perhaps other triggers of PAO1 activation. It will be important to determine the role of ODC and PAO1 in vivo, and to that end we have preliminary evidence that the expression of both enzymes is up-regulated in H. pylori gastritis tissues.