Suppression of cell division-associated genes by Helicobacter pylori attenuates proliferation of RAW264.7 monocytic macrophage cells

Helicobacter pylori at multiplicity of infection (MOI ≥ 50) have been shown to cause apoptosis in RAW264.7 monocytic macrophage cells. Because chronic gastric infection by H. pylori results in the persistence of macrophages in the host’s gut, it is likely that H. pylori is present at low to moderate, rather than high numbers in the infected host. At present, the effect of low-MOI H. pylori infection on macrophage has not been fully elucidated. In this study, we investigated the genome-wide transcriptional regulation of H. pylori-infected RAW264.7 cells at MOI 1, 5 and 10 in the absence of cellular apoptosis. Microarray data revealed up- and down-regulation of 1341 and 1591 genes, respectively. The expression of genes encoding for DNA replication and cell cycle-associated molecules, including Aurora-B kinase (AurkB) were down-regulated. Immunoblot analysis verified the decreased expression of AurkB and downstream phosphorylation of Cdk1 caused by H. pylori infection. Consistently, we observed that H. pylori infection inhibited cell proliferation and progression through the G1/S and G2/M checkpoints. In summary, we suggest that H. pylori disrupts expression of cell cycle-associated genes, thereby impeding proliferation of RAW264.7 cells, and such disruption may be an immunoevasive strategy utilized by H. pylori.

patients, indicating the importance of macrophages in host defense against H. pylori 9,10 . Macrophages respond to H. pylori infection by increasing surface expression of CD80, CD86 and HLA-DR accompanied by elevated secretion of cytokines including IL-12p70 and IL-23 that stimulate T H 1 and T H 17 responses, respectively 9 .
To maintain persistent infection of the host, H. pylori develops various immune evasion strategies to resist elimination by the host immune system, one of which is through delaying the macrophage-mediated phagocytosis 11,12 . Besides, chronic exposure to H. pylori impairs antigen presentation by macrophages, thus inhibiting development of T H 1 cells and IFN-γ secretion 13 . Several studies have reported that at high MOIs, H. pylori causes abrupt cell death of monocytes 14 and macrophages through activation of Erk- 15 , arginase II- 16,17 , or mitochondrial-dependent 18,19 pathways. H. pylori is also reported to initiate cell death through autophagic mechanism 20 . Despite these data showing H. pylori induces monocytes and macrophage cell death, in vivo examination of patient samples detected a large number of these cells at the infection site 9,10 . We therefore hypothesize that H. pylori is most likely present in the stomach at levels that are not sufficient to trigger apoptosis in host macrophages and may instead be protective, as H. pylori at low MOIs reduces apoptotic cell death in B lymphocytes 21 . The crosstalk of macrophages and H. pylori at low MOIs, which at present has not been fully described, is important for understanding the host defense against H. pylori, particularly during initial and chronic infection stages.
In this study, we performed microarray analysis to investigate genome-wide gene expression by RAW264.7 monocytic macrophages infected with H. pylori at MOI 10. Our report showed that H. pylori suppressed the expression of genes that encode for DNA synthesis and cell cycle-associated molecules that functionally translated to disrupted proliferation and cell cycle progression in these H. pylori-infected RAW264.7 cells.

H. pylori at MOI 10 activates monocytic macrophages cells.
To ascertain whether monocytic macrophages will be activated by H. pylori, we infected RAW264.7 cells with H. pylori Sydney strain 1 (SS1) at MOI 10. H. pylori SS1 is employed in this study as it is a well-established mouse-adapted pathogenic strain and its infectivity has been confirmed in RAW264.7 cells 16 . At 24 hours post infection (hpi), RAW264.7 cells were grossly enlarged (Fig. 1a), and increased intensities of forward scatter (FSC) and side scatter (SSC) parameters detected via flow cytometry verified the augmented cell size and complexity in the infected RAW264.7 cells (Fig. 1b). Besides, we observed that upon infection, RAW264.7 cells increased surface expression of macrophage markers F4/80 and CD11b, suggesting monocyte-to-macrophage differentiation. Uninfected controls were composed of undifferentiated monocytic macrophages displaying F4/80 low and CD11b (Mac-1) medium/high phenotypes whereas infected cells exhibited F4/80 high and CD11b high expression. Further, we observed no sign of apoptotic events within the infected macrophage population at MOI 1 to 10 (Supplementary Figure S1), providing support that H. pylori at these MOIs was capable of activating cells, but inadequate of inducing apoptotic cell death in RAW264.7 cells. On the contrary, at MOI of 100, H. pylori induced apoptosis (annexin + ) in approximately 30% of RAW264.7 cells at 24 hpi.

H. pylori infection causes dysregulation of gene transcription in RAW264.7 cells.
We then compared the transcriptional milieu between uninfected and infected monocytic macrophages through a genome-wide microarray analysis. Two replicates of uninfected and H. pylori (MOI 10)-infected RAW264.7 cells for 24 h were prepared independently and analyzed on an Agilent SurePrint G3 Human GE 8 × 60k microarray platform which comprised 55,821 probes. Scatter plot was generated based on normalized (Log 2 ) expression levels of total probes (Fig. 2a), and the total data were further filtered with fold changes (FC) > 2 or FC < -2 ( * P > 0.05) to select significant differentially expressed probes (Fig. 2b). A total number of 2471 probes (1341 genes) and 2651 probes (1591 genes) were significantly up-and down-regulated, respectively. Using these significant probes, hierarchical clustering (HCL) was executed with Pearson Correlation distance metric and average linkage. Heat map generated showed two separate clusters (Fig. 2c), indicating that H. pylori infection influences the regulation of an array of genes in RAW264.7 cells in both upward and downward trends. DNA replication pathway is vitally subdued in H. pylori-infected RAW264.7 cells. Next, we performed pathway analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database on the genes whose expression was significantly altered by H. pylori infection and the fold enrichment (FE) score for each pathway was calculated. In total, 25 biological pathways with changes more than 2-fold (P < 0.05) were identified (Fig. 2d). Nine of these pathways were significantly enriched, among which lysosome pathway (mmu04142) showed the highest score (FE = 3.94, P < 0.0001) (Supplementary Table S1), followed by cytokine-cytokine receptor interaction pathway (mmu04060) (FE = 1.8, P = 0.0016). On the contrary, 16 pathways were significantly suppressed (Fig. 2d). Interestingly, DNA replication pathway (mmu03030) was the most suppressed pathway at FE = -8.27 (P < 0.0001) (Supplementary Table S1). Both mismatch (mmu03430) and nuclear excision (mmu03420) DNA repair mechanisms were identified at FE = -7.30 (P < 0.0001) and FE = -3.49 (P < 0.0001), respectively, consistent with previous studies reporting increased DNA damage and compromised DNA repair mechanism during H. pylori infection 22,23 . Notably, H. pylori infection significantly disrupted terpenoid backbone (mmu00900) and steroid (mmu00100) biosynthesis pathways (FE = -7.66, P < 0.0001, and FE = -6.93, P < 0.0001, respectively). Terpenoid precursors in eukaryotes through sterol biosynthetic pathways are converted to cholesterol, ergosterol and phytosterol that are crucial for antimicrobial activity 24,25 and for cellular attachment by H. pylori and other intracellular pathogens 26 .   Table 1). The expression of the top 10 upand down-regulated genes were verified by quantitative real-time-polymerase chain reaction (qRT-PCR) analysis (Supplementary Figure S2). Of the top 10 up-regulated genes, five of them were immune response-related genes that encode for cytokines Csf1, Csf3, and Il-1β and chemokines Ccl7 and Cxcl2. Csf1 and Csf3 were significantly induced at 208-fold (P = 0.0002) and 771-fold (P = 0.0031), whereas the pro-inflammatory cytokine Il-1β was increased at 257-fold (P = 0.0028), relative to the uninfected control. Csf1 (also known as macrophage-colony stimulating factor) and Csf3 (also known as granulocyte-colony stimulating factor) are important factor for hematopoietic cells differentiation into macrophages and granulocytes, respectively 27 . H. pylori infection also elevated the transcription of chemokine genes Ccl7 and Cxcl2 by 682-fold (P = 0.0003) and 496-fold (P = 0.0008), respectively. Ccl7 [previously known as Monocyte-specific chemokine 3 (Mcp3)] is a chemoattractant for monocyte and macrophage 28 , whereas Cxcl2 [previously named as Macrophage inflammatory protein 2α (MIP2α)] is chemotactic for polymorphonuclear leukocytes 29 . Thus, excessive expression of these cytokines and chemokines can intensify the immune responses at the infection site.

H. pylori infection activates immune response-related genes in RAW264.7 cells.
To identify the biological functions of genes with pronounced expression changes subsequent to H. pylori infection, we categorized the differentially expressed genes according to their functional groups. As anticipated, majority of the immune-associated genes were markedly up-regulated by H. pylori infection (Fig. 3a). Colony stimulating factors Csf1, Csf2 and Csf3, were considerably up-regulated, along with pro-inflammatory cytokines Il1α, Il1β, and Tnfα, suggesting that H. pylori-infected RAW264.7 cells may trigger macrophage and granulocyte differentiation and promote robust inflammatory responses 27 . Further, up-regulation of Il23α in H. pylori-infected RAW264.7 cells may hasten the differentiation of T helper 17 cells to combat against H. pylori 47 . Activation markers including Cd44, Cd40, Cd86, and Cd274 were greatly up-regulated in H. pylori-infected macrophages. CD44 is a cell surface glycoprotein important for interaction and adhesion while CD40 and CD86 are receptor ligands for T cell CD40L, and CD28/CTLA4, respectively. Cd274 (also known as programmed cell death-1 ligand, PD-1L) binds to PD-1 receptor to modulate cell activation and inhibition 48 . Using flow cytometrical analyses, we verified that these transcriptional regulations were translated at protein levels (Fig. 3b). Shift of H. pylori-infected RAW264.7 cells from CD44 low/medium into CD44 medium/high population suggests cellular activation. Expression of CD86 was 13% higher in H. pylori-infected cells than in uninfected population, and 51% of the infected cells expressed CD274 compared to only 3.7% of the uninfected cells.
Conversely, the expression of Cd72, Cd97 and Cd101 were substantially down-regulated. CD72 and CD101 are known negative regulators of lymphocyte function. The cytoplasmic domain of CD72 consists of an immunoreceptor tyrosine inhibitor motif (ITIM) that suppresses B cell maturation and plasma cell differentiation 49,50 whereas CD101 (V7) inhibits T cell proliferation and T cell receptor signaling 51,52 . The function of CD97, an adhesion-linked G-protein-coupled receptor in immune cells remains poorly defined 53 . Besides, multiple genes encoding for chemokines/chemokine receptors (Ccl2, Ccl7, Ccr1, Cxcl12, etc.) and integrins (Itgβ3, Itgβ7, etc.) were greatly up-regulated. Together, these data suggest that H. pylori triggers robust transcription of genes that culminated in the activation of RAW264.7 cells.
H. pylori infection suppresses transcription of genes encoding for DNA synthesis and cell cycle progress. In contrast to the enhanced expression of immune response-related genes, majority of genes involved in DNA replication such as members of Mcm (minichromosome maintenance), Pol (DNA polymerase), and Rfc (replication factor C) families were substantially down-regulated (Fig. 4a). Genes encoding for cyclins (Ccna1, Ccnb1, etc.), cyclin-dependent kinases (Cdk1 and Cdk2) and mitotic arrest deficient-like proteins (Mad1l1 and Mad1l2) were similarly down-regulated. The expression of selected genes that were altered consequent to H. pylori infection were verified by qRT-PCR (Fig. 4b). Consistent with the microarray data, Aurkb expression was reduced by 3.4 times following H. pylori infection. The mRNA levels of Ccnb1, Ccnb2, Ccne1, and Ccne2 were reduced by 201-, 17-fold, 2.2-and 9.4-fold, respectively. Likewise, Cdk1 and Cdk2 expression decreased by 112-and 27-fold, respectively, while Mad1l1 was reduced by 1.8-fold.
H. pylori blocks G1-to-S transition by suppressing Aurkb activity. Given that H. pylori infection of macrophages impaired multiple genes associated with DNA replication and cell cycle, we performed propidium iodide staining and flow cytometrical analysis of the infected RAW264.7 cells to examine whether dysregulated cell cycle gene transcriptions would impact cell cycle progress (Fig. 5a). Percentage of cells at the S phase were reduced by approximately 3-fold upon H. pylori infection (6.5 ± 0.1% versus 15.6 ± 0.9% in controls). Additionally, the proportion of infected cells at G2/M phase were reduced by almost 50% (8.7 ± 1.1% versus 13.9 ± 0.9% in controls). These observations were accompanied by an increased number of cells retained at the G0/G1 phase (from 70.4 ± 2.0% to 84.9 ± 0.9%).
During cell cycle, AurkB mediates S phase entry by interacting with Cdk1 (Cdc2) 54 . Through immunoblot analysis, we found that AurkB protein level was diminished in cells infected with H. pylori (Fig. 5b). Concurrently, AurkB-mediated downstream phosphorylation of Cdk1 was also abolished  (Fig. 5b). In addition to mediating S phase entry, AurkB initiates G2/M transition by activating kinetochore protein complexes including Cenp-a 37 . We observed that the mitotic protein Cenp-a was reduced in H. pylori-infected cells. The reduced levels of Cenp-a and other mitosis-related proteins (Espl1 and Zwilch) were also detected at transcriptional level (Supplemental Figure S2). In addition, expression of  Data were shown as mean ± SD from one experiment run in duplicate, and were representative data of two independent experiments. (b) Immunoblot analysis of cell lysates prepared from control and H. pyloriinfected RAW264.7 cells for 24 or 48 h. Antibodies against AurkB, Cenp-a, phospho-Cdk1 (Cdc2) or Cyclin D1 were used. β -actin was used as loading control. *All gels were run under same experimental condition. Images were cropped from full length blots (Supplementary Figure S4). Shown are representative data of two independent experiments. Cyclin D1 was suppressed in H. pylori-infected RAW264.7 cells. These data suggest that H. pylori could block G1/S and G2/M transitions by inhibiting formation of AurkB and cyclin/cdk complexes.
Both CagA+ or CagA-deficient H. pylori strains cause anti-proliferative effect in RAW264.7 cells. H. pylori produces cytotoxin CagA and VacA which can destroy the gastric epithelium and lead to ulcer formation 55,56 . H. pylori SS1 is a mouse-adapted strain 57 which is deficient in the function of Cag pathogenicity island (PAI) 58 and possesses VacA s2m2 genotype. Next, we examined two other H. pylori strains with functional CagPAI and different vacA genotype for ability to induce anti-proliferative effect in the RAW264.7 cells. To address this, we infected cells with two additional H. pylori strains, namely J99 and 298. Both H. pylori J99 (a standard strain) and 298 (a mice-adapted derivative from a local clinical isolate -UM032) have complete CagPAI, CagA and the more cytotoxic VacA s1m1 genotype 59,60 . Our results showed that infection of RAW264.7 cells with SS1, J99 or 298 demonstrated comparable cell proliferation (Fig. 6c) and cell cycle effects (Supplementary Figure S3), suggesting that H. pylori effectively inhibited the RAW264.7 cell proliferation, regardless of CagPAI or VacA activities.

H. pylori infection attenuates proliferation of primary macrophage cells.
Because the above assays were performed using RAW264.7 cell line, we would like to use primary macrophage cells, to confirm the infectivity and anti-proliferative effect of H. pylori in the macrophages. Bone marrow cells were isolated from C57BL/6 mice and stimulated with 20 ng/ml M-CSF for 7 days to obtain bone marrow-derived macrophage (BMDM) cells. H. pylori infection for 24 h resulted in decreased proliferative cells within BMDM population (Fig. 7), supporting the ability of H. pylori to effectively block the macrophage cell proliferation.

Discussion
H. pylori is a Gram-negative bacterium colonizing nearly half of the human population and is a well-established etiological agent of gastritis, peptic ulcer and gastric cancer. In this study, we reported disrupted gene transcriptional program in H. pylori-infected RAW264.7 monocytic macrophage cells. Consequently, cell cycle progression and proliferation of these infected macrophages were greatly suppressed. H. pylori infection shifted cells into Ki-67 low non-proliferative stage, arrested cells at G0/G1 phases, and impeded entry of cells into S or G2/M phases. To date, there is no evidence showing that H. pylori causes cell cycle arrest in immune cells, although H. pylori has been reported to retain epithelial cells at G1/S and G2/M phases through increasing p27Kip1 and decreasing cyclin E/Cdk2 complex activities 61 . Besides, the H. pylori L-asparaginase can function as a cell cycle inhibitor by preventing entry into S phase in gastric cells 62 .
The cell cycle progression through G1/S checkpoint is controlled by sequential activation of cyclin/Cdk complexes 63 . At G1 phase, D-type cyclins (D1, D2 and D3) associate with Cdk4/Cdk6, while Cyclin E forms a complex with Cdk2. These cyclin-Cdk complexes sequentially activate retinoblastoma (Rb) protein and E2F that control the expression of a cluster of S phase-associated genes including Cyclin A and Cdk1 64 . During the G1-to-S transition, AurkB kinase plays an essential role in augmenting phosphorylation and activation of Rb, Cdk1 and Cdk2 38 . Our study showed that H. pylori-mediated transcriptional inhibition of genes encoding for Cyclin D1, Cdk1 and Cdk2 coupled with suppression of AurkB and Cdk1 phosphorylation may halt G1-to-S transition. Doubling of DNA content via DNA replication occurs during the S phase prior to M phase. During this stage, members of Mcm family form pre-replication complex that unravels the double helix and initiates replication at early S phase 65,66 whereas the Pol 67,68 and Rfc 69 complexes are important for DNA elongation. Strikingly, we observed that expression of numerous genes within the Mcm, Pol and Rfc families were strongly down-regulated in the H. pylori-infected RAW264.7 cells. Furthermore, KEGG pathway analysis showing depletion of DNA replication pathway upon H. pylori infection supports the hypothesis that H. pylori blocks DNA replication and inhibits cell cycle at S phase.
It is important to note that the H. pylori-mediated cell cycle inhibition is not limited to G1-to-S transition, but it occurs simultaneously at G2/M phase. This is supported by cell cycle assay and microarray data which showed significant suppression by H. pylori infection of multiple genes encoding for molecules associated with G2/M progress. AurkB may also play a key role in this process because in addition to regulating G1-to-S transition 38 , it participates in mitosis by modulating spindle function 70 in which AurkB controls centromere protein complex that includes histone Cenp-a protein which is responsible for assembly of kinetochore proteins 71,72 . Moreover, expression of Mad1l1, a protein vital for  mitosis progression and checkpoint control, is reduced by H. pylori 73,74 . In our study, H. pylori-infected RAW264.7 cells exhibited reduced mRNA and protein levels of AurkB and its substrate Cenp-a. In addition, genes encoding for Zwilch kinetochore proteins and for Separase (encoded by Espl1) indispensable for anaphase spindle elongation 75 were significantly suppressed by H. pylori (Supplementary Figure S2). Therefore, disruption of these processes can result in mitotic checkpoint failure, resulting in premature mitotic exit and chromosome mis-segregation that underlie tumorigenesis.
Macrophage is a key player in H. pylori pathogenesis in which its depletion causes reduced pathology in the gastric 3 . One of the major roles of macrophage is to trigger adaptive immune response. H. pylori-infected macrophage is able to produce BAFF which promotes T H 17 cell expansion by creating a pro-T H 17 milieu or by direct control of naïve T cell differentiation 76 . The ability of H. pylori to form chronic colonization in the host relies on their effective immune evasion strategies 77 . Previous study showed the H. pylori infection in macrophage can cause cell death by induction of macrophage arginase II 17 . In this study, we suggest a different strategy by which the H. pylori is able to block various cell proliferation-associated genes thus inhibits the macrophage cell growth. In addition to inhibiting macrophage, H. pylori VacA exotoxin interferes with the T cell activation through inhibiting calcium influx thus preventing NFAT nuclear translocation and the subsequent cytokine transactivation 78 . Besides, VacA is also able to interfere with antigen presentation by major histocompatibility complex in B cells 79 .
In summary, we observed that H. pylori infection impaired mitotic proliferation of RAW264.7 monocytic macrophage cells. G1-to-S cell cycle transition was inhibited in the infected cells subsequent to depleted expression of AurkB-and cyclins/cdks-encoding genes. We anticipate that H. pylori-mediated interference of macrophage proliferation is possibly one of the strategies employed by H. pylori to limit the quantity of macrophages at the infection site and to evade efficient clearance by host immune system.

Bacteria.
A mouse-adapted strain of H. pylori SS1 strain 57 , was provided by the H. pylori Research Laboratory, University of Western Australia. J99 strain was from Amerian Type Culture Collection (ATCC, Rockwille, MA) 59 while 298 strain was derived from a local clinical isolate, UM032, as previously described 60 . Bacteria was grown on chocolate agar plate supplemented with 7% laked horse blood (Oxoid, Basingstoke, UK) under microaerophilic conditions at 10% CO 2 , 37 °C in a humidified incubator and were subcultured every 3 days. For infection, H. pylori was harvested in brain heart infusion (BHI) broth and quantified by a spectrophotometer (OD 650 nm of 1 = 1 × 10 8 cells/ml). Viable cell count was predetermined by calculating colony forming units after serially diluted bacteria were drop plated onto chocolate agar plate.
Tissue culture. RAW264.7 cells were purchased from America Type Culture Control (ATCC TIB-71). RAW264.7 cells were cultured in Dulbecco's Modified Essential Medium supplemented with 10% heat inactivated fetal bovine serum and incubated at 37 o C, 5% CO 2 . One day prior to inoculation, cells were seeded in a T25 flask at 5 × 10 5 cells/ml. Cells were then infected with H. pylori SS1 at MOIs of 1, 5 or 10 for 24 h. Primary macrophage cell preparation. C57BL/6 mice were purchased (Jackson Laboratory, Bar Harbor, ME). Preparation of primary macrophages was adapted from a previous report 80 . Two male mice at 8-12 weeks old were euthanized and bone marrow cells were isolated from the femurs. Cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 100 μ g/ml streptomycin and 100 U/ml penicillin, 1× non-essential amino acids, 1 mM HEPES and stimulated with 20 ng/ml M-CSF (Biolegend, San Diego, CA). After 3 days, non-adherent cells were collected and cultured for another 3 days to obtain BMDM 81 . At day 7, adherent cells were infected with H. pylori SS1 at MOI 10 for 24 h. RNA extraction and qRT-PCR. RNA was isolated from cells using TRIzol reagent (Invitrogen, Carlsbard, CA) as described 82 . RNA integrity number (RIN) was > 9.5 as determined using Bioanalyzer 2100. cDNA was prepared using M-MLV reverse transcriptase (Invitrogen). qRT-PCR was carried out with SsoAdvanced SYBR Green Supermix (Biorad, Hercules, CA) in a Real-Time PCR 7500 (Applied Biosystems, Foster City, CA) using designed primers (Supplementary Table S2). Relative fold change was calculated using comparative 2 −ΔΔCT method. All experiments were run in triplicates and were presented as mean ± SD. Microarray analysis. Microarray analysis was performed with Agilent Technologies microarray platform using Agilent SurePrint G3 Human GE 8 60k containing 55,821 probes (Design ID: G4851A, Lot: 0006097429). Total RNA (100 ng) was primed with an oligo-dT containing the recognition site for RNA polymerase. RNA was labelled using Low Input Quick Amp Labeling Kit, One-Color (Agilent p/n 5190-2305) to produce cyanine 3-CTP labeled cRNA. cRNA (600 ng) was hybridized onto 8-array slide at10 rpm for 17 h at 65 °C. The slide was washed and scanned on Agilent High Resolution Microarray Scanner (C-model). Raw signal data were extracted from the TIFF image with Agilent Feature Extraction Software (V107.1.1). Pathway analysis was performed using the KEGG database 83,84 . Heat maps were generated with multiexperimental viewer (MeV) software 85 .
Statistical analysis. Data were analyzed with unpaired two-tailed Student's t-test or Benjamini-Horchberg False Discovery Rate (FDR) multiple testing correction. Samples were considered significant if P < 0.05.