|Home | About | Journals | Submit | Contact Us | Français|
Salmonella spp. and Shigella spp. are responsible for millions of cases of enteric disease each year worldwide. While these pathogens have evolved distinct strategies for interacting with the human intestinal epithelium, they both induce significant proinflammatory responses that result in massive transepithelial migration of neutrophils across the intestinal mucosa. It has previously been shown with Salmonella enterica serotype Typhimurium that the process of neutrophil transmigration is mediated in part by the secretion of hepoxilin A3 (HXA3; 8-hydroxy-11,12-epoxy-eicosatetraenoic acid), a potent neutrophil chemoattractant, from the apical surface of infected model intestinal epithelium. This study confirms that HXA3 is also secreted in response to infection by Shigella flexneri, that it is produced by a pathway involving 12/15-lipoxygenase (12/15-LOX), and that S. enterica serovar Typhimurium and S. flexneri share certain elements in the mechanism(s) that underlies the otherwise separate signal transduction pathways that are engaged to induce polymorphonuclear leukocyte (PMN) transepithelial migration (protein kinase C and extracellular signal-regulated kinases 1 and 2, respectively). PMN transepithelial migration in response to infection with S. flexneri was dependent on 12/15-LOX activity, the enzyme responsible for the initial metabolism of arachidonic acid to HXA3. Probing further into this pathway, we also found that S. enterica serovar Typhimurium and S. flexneri activate different subtypes of phospholipase A2, a critical enzyme involved in the liberation of arachidonic acid from cellular membranes. Thus, although S. enterica serovar Typhimurium and S. flexneri utilize different mechanisms for triggering the induction of PMN transepithelial migration, we found that their reliance on 12/15-LOX is conserved, suggesting that enteric pathogens may ultimately stimulate similar pathways for the synthesis and release of HXA3.
Salmonella spp. and Shigella spp. are frequently implicated in outbreaks of enteric disease due to the consumption of contaminated food. In the United States alone, the number of prevalent cases of salmonellosis is estimated to range from 80,000 to 3,700,000 annually (7). Moreover, Shigella sp. infections are responsible for ~163 million cases of dysentery and 1 million deaths each year worldwide (27). While the majority of infections occur in third-world countries, where contaminated food and drinking water are common (27), even developed nations cannot escape several outbreaks in a given year (65). Therefore, studying the molecular pathogenesis of these gram-negative enteric pathogens is of paramount importance, especially taking into consideration emerging antibiotic resistance and the lack of appropriate vaccines.
Salmonella enterica serotype Typhimurium and Shigella flexneri interact with the intestinal epithelium to induce intestinal inflammation characterized by the transepithelial migration of neutrophils (2, 12, 33, 41, 49). Symptoms of salmonellosis manifest within 48 h of the ingestion of contaminated food or water and include nausea, vomiting, and acute, mostly self-limiting diarrhea. Following its ingestion, this motile pathogen colonizes and penetrates the intestinal epithelium and gains access to systemic sites, such as the liver and spleen, through lymphatic and blood circulation (70). The ability of S. enterica serovar Typhimurium to cause disease in humans is related to the acquisition of virulence genes termed pathogenicity islands. Located on a 40-kb segment of the bacterial chromosome termed SPI-1 (Salmonella pathogenicity island 1) are 25 virulence genes that encode structural components and secreted substrates of a type III secretion system (TTSS) (11, 21). This TTSS is essential for the ability of S. enterica serovar Typhimurium to invade and induce proinflammatory responses (21).
Like S. enterica serovar Typhimurium, Shigella flexneri enters the human host by being ingested via contaminated food or water, passaging through the stomach, and infecting the large intestine. Infection of the colonic mucosa by Shigella results in bacillary dysentery (shigellosis), an acute inflammatory disease characterized by abdominal cramps, fever, and severe diarrhea often containing blood and mucus (66). In contrast to what is the case for Salmonella, humans (and certain higher primates) serve as the only natural hosts and reservoirs for Shigella (66). Thus, this organism has evolved to be highly host specific as well as efficient, with a particularly low infectious dose in comparison to other enteric pathogens (77). Also distinct from Salmonella, Shigella has a unique mode of entry requiring access to the basolateral surface. It has been established that Shigella alters the tight junctional complex, thereby allowing the pathogen to traverse the paracellular space and access the basolateral surface, an event that also decreases barrier function (64). Once at the basolateral surface, Shigella rapidly invades and disseminates through the epithelium, a consequence of the TTSS and additional proteins encoded in a 31-kb region (the mxi-spa locus) of the large 220-kb virulence plasmid (5, 69).
While these pathogens have evolved distinct strategies for interacting with the human intestinal epithelium, they both induce significant proinflammatory responses that result in the massive transepithelial migration of neutrophils across the intestinal mucosa (2, 12, 33, 41, 49). Both S. enterica serovar Typhimurium and S. flexneri induce epithelial cells to secrete a repertoire of chemokines that play an active role in recruiting polymorphonuclear leukocytes (PMNs) from the peripheral circulation and directing them across the epithelium to the intestinal lumen (15, 25, 43, 44, 61, 63, 68, 74). In the case of S. enterica serovar Typhimurium, the effector protein SipA is necessary to initiate the cellular events that lead to PMN transepithelial migration during acute states of active intestinal inflammation (10, 35, 73). Regarding the molecular mechanism underlying these cellular events, it has been revealed so far that SipA activates a novel ADP ribosylation factor 6- and phospholipase D-dependent lipid signaling cascade (10) that in turn activates protein kinase C α (PKC α) (73), events that ultimately lead to the apical secretion of a potent PMN chemoattractant, eicosanoid hepoxilin A3 (HXA3; 8-hydroxy-11,12-epoxy-eicosatetraenoic acid) (51). HXA3, an endogenous 12/15-lipoxygenase (12/15-LOX) product, forms a chemotactic gradient across the epithelial tight junctional complex that directs PMNs across the intestinal epithelium to the luminal surface (51).
The mechanisms underlying Shigella-induced PMN transepithelial migration are less defined. Studies addressing PMN movement across model intestinal epithelia indicate that PMN transmigration does occur as a result of Shigella infection and that this event further facilitates bacterial invasion (48, 62). Furthermore, the induction of PMN transepithelial migration is dependent not only on a functional TTSS and invasive ability (24, 67) but also on intercellular spread (16). Recently, lipopolysaccharide and the S. flexneri Osp proteins have been shown to be involved in the activation of mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) signaling pathway (26, 85, 86). In turn, ERK1/2 phosphorylation and its nuclear localization have been shown to play a role in PMN transepithelial migration (26).
The purpose of this study was to determine if S. enterica serovar Typhimurium and S. flexneri share certain elements in the mechanism(s) that underlies the otherwise separate signal transduction pathways that are engaged to induce PMN transepithelial migration (PKC and ERK1/2, respectively). We found that S. flexneri, similarly to S. enterica serovar Typhimurium, induced the apical secretion of the potent PMN chemoattractant HXA3 from model intestinal epithelia. In addition, PMN transepithelial migration in response to infection with S. flexneri was dependent on 12/15-LOX activity, the enzyme responsible for the initial metabolism of arachidonic acid to HXA3. Probing further into this pathway, we also found that S. enterica serovar Typhimurium and S. flexneri activate different subtypes of phospholipase A2 (PLA2), a critical enzyme involved in the liberation of arachidonic acid from cellular membranes. Thus, although S. enterica serovar Typhimurium and S. flexneri utilize different mechanisms for triggering the induction of PMN transepithelial migration, we found that their reliance on 12/15-LOX is conserved, suggesting that enteric pathogens may ultimately stimulate similar pathways for the synthesis and release of HXA3.
The human colon cancer-derived cell lines T-84 (passages 57 to 77) and HCT-8 (passages 30 to 40) were maintained in Dulbecco's modified Eagle's medium -F-12 (Invitrogen, Carlsbad, CA) supplemented with 15 mM HEPES, 14 mM NaHCO3, 40 μg ml−1 penicillin, 80 μg ml−1 ampicillin, 90 μg ml−1 streptomycin, and 6% fetal calf serum. Monolayers were grown on 0.33- and 4.5-cm2 suspended-collagen-coated permeable polycarbonated filters with pore sizes of 5 and 3 μm, respectively (Costar, Cambridge, MA). Inverted monolayers used for PMN transmigrations were constructed as previously described (39, 53, 59, 60). Monolayers were utilized after 7 to 14 or 14 to 21 days for 0.33- and 4.5-cm2 filters, respectively, once having reached a confluent, polarized, and differentiated state. A steady-state transepithelial cell resistance (TEER) of at least 500 Ω per cm2 was reached for all monolayers used, as measured using a Millicell ERS voltohmmeter (World Precision Instruments, New Haven, CT).
Prior to treatment or infection, monolayers were drained of medium, extensively washed, and allowed to equilibrate in Hanks balanced salt solution containing Mg2+, Ca2+, and 10 mM HEPES (HBSS+; pH 7.4; Sigma, St. Louis, MO) for 30 min at 37°C.
For 12-LOX inhibition, T-84 monolayers were incubated in the presence of 2 μM baicalein (stock concentration of 1 mM in dimethyl sulfoxide [DMSO]) in cell culture medium for 48 h at 37°C. For 5-LOX inhibition, the monolayers were incubated in the presence of 2 μM caffeic acid (stock concentration of 22 mM in DMSO) in cell culture medium for 24 h. Identical monolayers were incubated in the presence of DMSO in medium at the same concentration as during treatment to serve as vehicle controls. Following treatment, inhibitors/media were thoroughly washed away and monolayers were equilibrated in HBSS+ for 30 min at 37°C prior to infection. Both inhibitors were purchased from Biomol (Plymouth Meeting, PA).
For the general inhibition of all subfamilies of PLA2, T-84 monolayers were incubated with the paninhibitor ONO-RS-082 (stock concentration of 20 mM in DMSO; Biomol) diluted in HBSS+ at final concentrations of 5, 10, and 20 μM. For specific inhibition of secretory PLA2 (sPLA2), monolayers were incubated with 4-bromophenacyl bromide (stock concentration of 72 mM in methanol; Fluka, Switzerland) diluted in HBSS+ at concentrations of 0.07, 0.7, and 7 μM. For the inhibition of cytosolic PLA2 (cPLA2), cPLA2α inhibitor (stock concentration of 6 mM in DMSO; Calbiochem, Darmstadt, Germany) was used at final concentrations of 0.06, 0.6, and 6 μM in HBSS+. For the inhibition of calcium-independent PLA2 (iPLA2), monolayers were incubated in bromoenol lactone (BEL; stock concentration of 26 mM in methyl acetate; Cayman Chemical, Ann Arbor, MI). BEL was first diluted to 5 mM in HBSS+, and monolayers were incubated in final concentrations of 1, 5, and 25 μM in HBSS+. Monolayers were incubated with the PLA2 inhibitors for 2 h at 37°C, after which they were thoroughly washed with HBSS+ and infected as described below. Identical monolayers were exposed to the highest concentration of each vehicle in the absence of the inhibitors in order to control for any nonspecific effects. In addition, nonspecific effects of these inhibitors were also monitored by measuring TEER of the cell monolayers following treatment, and the concentrations used in this study were selected from those that had no effect on TEER.
For inhibition of diacylglycerol (DAG) lipase, T-84 monolayers were incubated in the presence of 50 μM RHC-80267 (in DMSO; Biomol) in HBSS+ for 2 h at 37°C or the equivalent dilution of DMSO in HBSS+ to serve as a vehicle control. After 2 h, the monolayers were washed free of the inhibitor and infected.
S. enterica serovar Typhimurium SL1344 is an invasive, virulent strain known to induce PMN transepithelial migration (45, 81). The noninvasive mutant vv341 (hilA::Kan-339) is unable to induce PMN migration (22, 46). S. flexneri 2457T, a wild-type strain of S. flexneri serotype 2a, is invasive in HeLa and T-84 cells (48). BS103, a mutant of 2457T, is cured of the large virulence plasmid and lacks the ability to invade or stimulate neutrophil transepithelial migration (42, 48).
S. enterica serovar Typhimurium SL1344 and vv341 cultures were grown in Luria-Bertani broth (LB; Becton Dickinson, Sparks, MD), according to the work of McCormick et al. (44). S. flexneri 2457T and BS103 cultures were grown aerobically at 37°C in Bacto tryptic soy broth (Becton Dickinson, Sparks, MD), according to the work of McCormick et al. (48). Bacteria were pelleted by centrifugation, washed, and suspended to the appropriate concentration in HBSS+ to yield multiplicities of infection (MOI) of 1,000 and 200 for S. enterica serovar Typhimurium and S. flexneri, respectively, for PMN transepithelial migration assays (44, 48).
The PMN transepithelial migration assay was previously described in detail (39, 59, 60) and modified for Salmonella enterica serovar Typhimurium and Shigella flexneri as outlined by McCormick et al. (44, 48). Human PMNs were purified from whole blood (anticoagulated with 13.2 g of citrate and 11.2 g of dextrose in 500 ml of water; pH 6.5) collected by venipuncture from healthy human volunteers of both sexes in accordance with the Massachusetts General Hospital Institutional Review Board approval (protocol number P-007782/7) as previously described (59, 60). PMNs were suspended at a concentration of 5 × 107 ml−1 in modified HBSS (without Ca2+ and Mg2+) and were 95% pure with 98% viability (47).
Inverted polarized T-84 or HCT-8 monolayers on 0.33-cm2 filters were infected with 25 μl S. enterica serovar Typhimurium (apically) or S. flexneri (basolaterally), and the transepithelial migration of isolated PMNs was performed as described previously (44, 48).
Plasmids used to generate small interfering RNAs (siRNAs) were constructed using the pSUPER vector (Oligoengine, Seattle, WA) by the method described by Brummelkamp et al. (4). Oligonucleotides were designed incorporating a 19-nucleotide sequence (in italics in sequences below) from the targeted human iPLA2 (PLA2G6; GenBank accession number, NM_003560) or arachidonate 15-LOX (ALOX15; GenBank accession number, NM_001140) transcript and its reverse complement (also in italics) separated by a short spacer region, along with BglII and HindIII restriction sites. For iPLA2, the oligonucleotide sequences were 5′-GATCCCCTGACCGACCTCATCCGTAATTCAAGAGATTACGGATGAGGTCGGTCATTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAATGACCGACCTCATCCGTAATCTCTTGAATTACGGATGAGGTCGGTCAGGG-3′ or the following random control sequences: 5′-GATCCCCAGATGGATGTCACCGACTATTCAAGAGATAGTCGGTGACATCCATCTTTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAAAGATGGATGTCACCGACTATCTCTTGAATAGTCGGTGACATCCATCTGGG-3′. For 15-LOX, the oligonucleotide sequences were 5′-GATCCCCTCGTGAGTCTCCACTATAATTCAAGAGATTATAGTGGAGACTCACGATTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAATCGTGAGTCTCCACTATAATCTCTTGAATTATAGTGGAGACTCACGAGGG-3′ or the following random control sequences: 5′-GATCCCCCATCCTATCTTCAAGCTTATTCAAGAGATAAGCTTGAAGATAGGATGTTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAACATCCTATCTTCAAGCTTATCTCTTGAATAAGCTTGAAGATAGGATGGGG-3′. Oligonucleotides were annealed, yielding double-stranded DNAs with overhanging restriction sites, and ligated into digested pSUPER. Constructs were transformed into competent Escherichia coli DH5α by standard methods and plated on LB with ampicillin (50 μg/ml). Plasmids were extracted (QIAprep spin miniprep kit; Qiagen, Valencia, CA) and sequenced for confirmation prior to transfection.
The HCT-8 cell line (kind gift of Cheleste Thorpe, Tufts University School of Medicine) is a transformed polarizing intestinal human cell line and was used for this protocol because of its high transfection efficiency. HCT-8 cells were transfected with modified pSUPER (4 μg) by use of Lipofectamine 2000 (Invitrogen, Carlsbad, CA) per the manufacturer's instructions. Cells were passaged into fresh media with a selection agent (neomycin-G418, 1 mg/ml; Sigma-Aldrich, St. Louis, MO) added the next day. Cells underwent two additional cycles of growth/passage in G418 prior to use.
Transfected HCT-8 cell monolayers were grown on 4.5-cm2 Transwell filters and infected with 1 ml of bacterial suspension at MOI of 500 and 200 for S. enterica serovar Typhimurium and S. flexneri, respectively; uninfected monolayers served as the negative control. Following infection, the cell monolayers were harvested on ice in 300 μl of lysis buffer and the cytosolic proteins were collected as previously described (64). The cytosolic proteins (amounts equivalent to 30 μg for each sample) were electrophoresed on a gradient Tris-glycine polyacrylamide gel electrophoresis gel (8 to 16%) and transferred to nitrocellulose. For signal generation, the membrane was incubated at room temperature first for 1 h with 5% nonfat dried milk in Tris-buffered saline with 0.1% Tween 20 (TBST; 0.1 M, pH 7.4) and then for 16 h at 4°C with an antibody against 15-LOX (H-235) or group VI iPLA2 (both from Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or with glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Fitzgerald Industries International Inc, Concord, MA) diluted 1:1,500 in 5% nonfat dried milk in TBST. The membranes were washed four times in TBST and incubated with the following horseradish peroxidase-conjugated secondary antibodies: goat anti-rabbit for 15-LOX (MP Biomedicals, Solon, OH), donkey anti-goat (Santa Cruz) for iPLA2, and goat anti-mouse for GAPDH (Santa Cruz). Secondary antibodies were diluted 1:2,000 in 5% nonfat dried milk in TBST for 1 h at room temperature. The membranes were washed four times in TBST and the bands were visualized by enhanced chemiluminescence with the ECL system (Pierce) and analyzed using ImageJ software (W. S. Rasband, 1997 to 2007; http://rsb.info.nih.gov/ij/).
Infection of T-84 monolayers was performed as described previously (44) with slight modification (48). Transwells were treated with either the cPLA2α inhibitor (0.06 μM) or the iPLA2 inhibitor BEL (5 μM) for 2 h as described above. After treatment, the inhibitors were removed with HBSS+ washes and S. enterica serovar Typhimurium or S. flexneri was administered to the apical or basolateral surface, respectively, of T-84 cell monolayers (MOI of 25) at 37°C for 1 h for S. enterica serovar Typhimurium and 1.5 h for S. flexneri. Monolayers were then washed, treated with gentamicin, and lysed (44, 48). The cell lysates were diluted and plated on MacConkey agar plates. Data are expressed as percentages of initial inoculum added to the monolayer.
The cytotoxic effects of the inhibitors used in this study were determined by the release of lactate dehydrogenase (LDH) from the monolayers. Monolayers were exposed to specific inhibitors and infected as described above. Following treatment and infection, monolayers were incubated in HBSS+ for up to 3 h. The buffer in the apical and basolateral compartments was collected and pooled for each Transwell to represent the amount of LDH released. The remaining monolayers were lysed with 1% Triton X-100, the amounts of LDH in both fractions were evaluated using an LDH kit (Sigma; St. Louis, MO) according to the manufacturer's instructions, and the percentage of LDH released by the monolayers following treatment and infection was determined.
T-84 monolayers on 4.5-cm2 Transwells were drained of medium and washed with warm HBSS+ prior to basolateral infection with S. flexneri (MOI of 200) for 1.5 h at 37°C. The bacteria were washed away with warm HBSS+ and 750 μl of HBSS+ was added to the apical surface, with 2 ml HBSS+ in the basolateral chamber. The Transwells were incubated at 37°C and the buffer in the apical compartment was collected every 2 h and replenished with the same volume of fresh HBSS+. Collected buffer was pooled and stored on ice in the dark. After 6 h (three collections), all pooled samples received 0.5 ng of prostaglandin B2 (PGB2) (Cayman Chemical) as an internal standard and were brought to a pH of 3 to 4 with HCl prior to extraction. Samples were extracted by solid-phase extraction as previously described (19). Samples were eluted in 5 ml of methanol, concentrated under nitrogen, and resuspended in 100 μl of methanol. Samples were stored at −80°C until analyzed.
HXA3 was identified by liquid chromatography-tandem mass spectrometry (LC/MS-MS)-based lipidomic analyses (52). Extracted samples suspended in methanol were analyzed by a triple quadrupole linear ion trap LC/MS-MS system (MDS Sciex 3200 QTRAP; Applied Biosystems, Foster City, CA) equipped with a LUNA C18(2) minibore column (Phenomenex, Torrance, CA) using a mobile phase (methanol:water:acetate, 65:35:0.03 [vol/vol/vol]) with a 0.35-ml/flow rate. MS-MS analyses were carried out in negative-ion mode and fatty acids were identified and quantified by multiple-reaction-monitoring mode using specific transitions for HXA3 (335→127 m/z), and PGB2 (333→175 m/z). The specific retention time for HXA3 was established with a synthetic standard (Biomol) and transition ions established by MS-MS analyses using enhanced-product-ion mode with appropriate selection of the parent ion in quadrupole 1. Calibration curves (1 to 1,000 pg) for HXA3 and PGB2 were established with synthetic standards.
PMN transmigration results are shown as the percentage of the PMNs that had completely traversed untreated monolayers within a given condition (i.e., 100% of wild-type or formylmethionyl-Leu-Phe [fMLP] response), as derived from a standard daily PMN dilution curve. PMN isolation was limited to repetitive donations by 10 different donors over the course of the experiments. Due to variations in both PMN and transepithelial resistance between monolayers (baseline resistance, 500 to 1,000 Ω per cm2), data were analyzed within an individual experiment and not between experiments. However, the overall trends associated within an experiment were reproducible between experiments. The data are expressed as the mean ± standard deviation for triplicate samples and represent one of at least three independent experiments performed. Statistical analysis was performed by Student's t test.
We previously reported that upon infection with S. enterica serovar Typhimurium, model intestinal epithelia apically release a potent neutrophil chemoattractant, a molecule that was originally termed pathogen-elicited epithelial chemoattractant (47). We have since identified this molecule as the eicosanoid HXA3 and have determined that it is also released from epithelial cells in response to other pathogens, as in the case of lung epithelial cells exposed to Pseudomonas aeruginosa (23). Given that a hallmark of shigellosis is the massive influx of PMNs into the intestinal lumen, we postulated that HXA3 is released in response to infection with Shigella flexneri and plays a role in guiding PMNs across the intestinal epithelium.
To determine if HXA3 is apically secreted from model intestinal epithelia infected with S. flexneri, infected T-84 monolayers were incubated in HBSS+ and the buffer within the apical chamber was collected every 2 h and replaced over the course of 6 h. The apical secretions were purified and analyzed by LC/MS-MS, as described in Materials and Methods. HXA3 was identified based on diagnostic fragment ions (Fig. (Fig.1A)1A) and a high-performance LC retention time that are consistent with an HXA3 structure. Significant levels of released HXA3 were detected in the apical compartment of S. flexneri-infected monolayers, and these levels were substantially greater than those of HXA3 from uninfected monolayers (Fig. (Fig.1B).1B). A synthetic standard of HXA3 was used to identify the specific transition and retention time.
Given that HXA3 is a hydroxy epoxide derivative formed from 12S-hydroperoxyeicosa-5Z,8Z,10E,14Z-tetraenoic acid (12S-HpETE), the main product of 12S-LOX activity on arachidonic acid (13, 56), we investigated the effect of LOX activity inhibitors on the ability of S. flexneri to induce PMN transepithelial migration. We hypothesized that an inhibitor of 12-LOX activity would reduce the secretion of HXA3 and that this effect would culminate in an overall reduction in pathogen-elicited PMN migration, as we have previously reported for epithelial monolayers infected with S. enterica serovar Typhimurium and P. aeruginosa (23, 51). Monolayer exposure to the 12-LOX inhibitor baicalein prior to infection nearly abolished S. flexneri-induced PMN transmigration, while exposure to the 5-LOX inhibitor caffeic acid had no effect (Fig. (Fig.2A).2A). Neither inhibitor displayed nonspecific effects, as observed by the lack of change in PMN migration across inhibitor-treated monolayers in response to a noninvasive mutant of S. flexneri, BS103 (Fig. (Fig.2A),2A), or across uninfected monolayers exposed to the potent neutrophil chemoattractant fMLP or buffer controls (Fig. (Fig.2B).2B). Vehicle control monolayers exposed to DMSO alone during inhibitor incubation resulted in a level of PMN migration equivalent to that seen for buffer-treated, infected monolayers (data not shown). Furthermore, the reduction in migration in baicalein-treated monolayers was not due to cytotoxic effects, as LDH assays revealed no significant cellular damage following treatment and infection (data not shown).
As an alternative approach to drug studies, we utilized a more specific means of reducing 12-LOX activity. There are four enzymes known to possess 12-LOX activity in humans: platelet type, 12R, epidermal, and 12/15-LOX (82-84). Since previous studies using mouse leukocyte-type 12-LOX showed that intestinal epithelial cells selectively express epithelial 12-LOX during active states of inflammation (72), we chose to generate siRNA constructs by use of the mRNA sequence for ALOX15, which encodes 12/15-LOX, the human homologue of murine leukocyte-type 12-LOX (8). This vector, or a control vector producing nonspecific siRNA, was transfected into HCT-8 cells. HCT-8 cells were more suitable for this particular procedure, since, similar to T-84 cells, they are a transformed polarizing human intestinal cell line and PMNs are able to traverse HCT-8 monolayers in response to infection with low background levels. However, unlike T-84 cells, the HCT-8 cell line is easily and highly transfectable by standard methods.
PMN transepithelial migration in response to both S. enterica serovar Typhimurium and S. flexneri was significantly reduced across HCT-8 monolayers expressing siRNA against ALOX15 mRNA in comparison to what was seen for monolayers expressing nonspecific siRNA (Fig. 3A and B). Similar to the drug inhibitor study, positive control fMLP-induced migration remained unaffected by the reduction in 12/15-LOX activity (Fig. (Fig.3C).3C). As observed with buffer-treated controls, there may also be a reduction in background levels of PMN migration across the monolayers, suggesting that 12/15-LOX activity may play a role in the processes that stimulate nonspecific or basal levels of PMN movement as well (Fig. 3A and B). Western blotting was performed to confirm and quantify the loss of 12/15-LOX by siRNA with densitometric analysis by ImageJ (Rasband, 1997 to 2000; http://rsb.info.nih.gov/ij/), indicating that there was an ~70% reduction in the amount of 12/15-LOX in uninfected and infected HCT-8 monolayers expressing ALOX15 siRNA from the level for the control monolayers expressing nonspecific siRNA (Fig. (Fig.3D).3D). In support of the 12-LOX inhibitor study, these results demonstrate that a reduction in 12/15-LOX activity significantly reduces PMN transepithelial migration in response to the enteric pathogens S. enterica serovar Typhimurium and S. flexneri.
HXA3 is generated from 12S-HpETE, a primary product of 12-LOX activity on arachidonic acid (13, 56). In humans, 12/15-LOX is capable of synthesizing 12-HpETE as a minor product and 12/15-LOX is also expressed in numerous epithelial cell types, including the intestinal epithelium (31, 32). The liberation of arachidonic acid is considered to be the rate-limiting step in the generation of eicosanoids. Presuming that arachidonic acid must first be made available to 12/15-LOX in order to generate the precursor to HXA3, we next investigated the potential mechanisms by which arachidonic acid may be released for its subsequent involvement in PMN migration. Two enzymes are capable of generating free arachidonic acid and thus are central to the production of various eicosanoids. The first is DAG lipase, an enzyme that releases arachidonic acid from DAG, and the second is PLA2, a family of proteins that release arachidonic acid from cell and organelle membranes (6, 14, 30).
To address the roles of DAG lipase and PLA2 in S. enterica serovar Typhimurium- and S. flexneri-induced PMN migration, we again employed a number of inhibitors to block the specific enzyme activity. The DAG lipase inhibitor RHC-80267 was previously shown to decrease bacterially induced arachidonic acid release by 25% (24). However, it was unable to significantly reduce PMN migration in response to either S. enterica serovar Typhimurium or S. flexneri. Untreated T-84 monolayers infected with S. enterica serovar Typhimurium resulted in 11.5 × 104 ± 2 × 104 PMNs having traversed the monolayer, while 10.3 × 104 ± 1.5 × 104 PMNs crossed infected monolayers treated with RHC-80267. Similarly, 32 × 104 ± 7 × 104 PMNs migrated across untreated monolayers infected with S. flexneri, while 38 × 104 ± 5 × 104 PMNs migrated across RHC-80267-treated, S. flexneri-infected monolayers.
The other mediator of arachidonic acid release is PLA2, which functions by liberating arachidonic acid from membranes (14). PLA2 represents a family of at least 19 distinct proteins categorized into multiple subfamilies, three of which have been associated with eicosanoid generation: sPLA2, cPLA2, and iPLA2 (14). The general PLA2 inhibitor ONO-RS-082 inhibits each of these three subfamilies of interest and significantly reduced PMN migration in response to S. enterica serovar Typhimurium and S. flexneri at each dose tested (Fig. (Fig.4A),4A), while migration levels in positive and buffer control monolayers were unaffected by the inhibitor (Fig. (Fig.4B).4B). Vehicle control monolayers exposed to DMSO alone followed by infection resulted in a level of PMN migration equivalent to that seen for buffer-treated, infected monolayers (data not shown). In addition, LDH assays revealed no significant cellular damage following treatment and infection, indicating that reductions in migration across ONO-RS-082-treated monolayers were not due to cytotoxic effects (data not shown).
Given that the general PLA2 inhibitor reduced PMN migration by at least 40% compared to what was seen for untreated controls (Fig. (Fig.4A),4A), we next sought to determine which PLA2 subfamily may be responsible. To determine the involvement of each PLA2 subfamily, we separately employed inhibitors specific for sPLA2, cPLA2, and iPLA2. As shown in Fig. Fig.5,5, sPLA2 activity blocked with 4-bromophenacyl bromide showed no reduction in PMN transepithelial migration observed at any dose for either pathogen.
No reduction in PMN transepithelial migration was observed for S. enterica serovar Typhimurium-infected monolayers pretreated with the cPLA2α inhibitor. However, PMN migration in response to S. flexneri was significantly reduced by 70% at the lowest dose tested (0.06 μM) and increased with higher doses, with the highest dose resulting in an ~30% decrease compared to the level for untreated controls (Fig. 6A and B). The rise in migration with higher doses may be due to stress or damage caused by the combined effects of the inhibitor and infection, although no differences were seen for positive control or uninfected monolayers (Fig. (Fig.6C)6C) or vehicle (DMSO) control monolayers (data not shown). Furthermore, LDH assays did not reveal any significant cellular damage following treatment and infection, indicating that reductions in migration across cPLA2α inhibitor-treated monolayers were not due to significant cytotoxic effects (data not shown). Therefore, while stress may have resulted in increased migration with higher doses, this did not correspond to observable or significant damage to the monolayers. Nonetheless, it should be noted that despite the rise in PMN migration, all doses resulted in significant reductions in migration compared to what was seen for untreated, S. flexneri-infected monolayers.
It is also important to note that the reduction in neutrophil movement across the monolayer was not an effect of impaired invasion by S. flexneri. An invasion assay revealed equivalent levels of invasion by the pathogen in both untreated and treated monolayers, with 0.05% ± 0.01% and 0.04% ± 0.01% of the initial inocula having invaded untreated and cPLA2α inhibitor-treated monolayers, respectively.
While effects were seen with the cPLA2α inhibitor with regard to S. flexneri, S. enterica serovar Typhimurium-induced PMN migration remained unaltered by blocking cPLA2 activity. Interestingly, the opposite was true following incubation with the iPLA2 inhibitor BEL. Monolayers treated with 5 μM BEL prior to infection with S. enterica serovar Typhimurium displayed a 90% reduction in the number of PMNs that traversed the monolayer compared to the level for untreated controls (Fig. (Fig.7A).7A). Similar to what was seen for the cPLA2α inhibitor and S. flexneri, the highest dose (25 μM) of BEL resulted in an increase in migration over what was seen for the 5 μM dose, although a significant reduction compared to what was seen for untreated, infected monolayers was still observed. Again, no significant effects were observed with positive control, uninfected (Fig. (Fig.7C)7C) or vehicle (methyl acetate) control monolayers, and LDH assays did not reveal significant cytotoxic effects following treatment and infection (data not shown). Further, the reduction in neutrophil migration was not due to effects on invasion by S. enterica serovar Typhimurium, as an invasion assay revealed equivalent invasion levels by S. enterica serovar Typhimurium in both untreated and treated monolayers, with 0.42% ± 0.03% and 0.41% ± 0.03% of the initial inocula having invaded, respectively. However, S. flexneri-induced PMN migration was unaffected by targeting iPLA2 inhibition (Fig. (Fig.7B);7B); similarly, S. enterica serovar Typhimurium-induced migration was unaffected by targeting cPLA2 inhibition (Fig. (Fig.6A6A).
To confirm our findings from the drug studies by a more selective, non-inhibitor-based approach, we attempted to generate HCT-8 cells with reduced expression of cPLA2α and iPLA2 via RNA interference. We constructed siRNA-generating plasmids aimed at reducing mRNA transcripts for the human genes encoding cPLA2α and iPLA2. Unfortunately, efforts to manipulate the level of cPLA2α within HCT-8 cells were unsuccessful, as the reduction of this enzyme proved lethal for HCT-8 cells (and other transfectable human intestinal cell lines). However, we were able to successfully reduce the levels of iPLA2 by siRNA. As seen in Fig. Fig.8,8, PMN transepithelial migration in response to S. enterica serovar Typhimurium was significantly reduced (~50% reduction) across HCT-8 monolayers expressing siRNA against PLA2G6 mRNA in comparison to that across monolayers expressing nonspecific siRNA (Fig. (Fig.8A),8A), while there were no differences in PMN migration in response to S. flexneri with the reduction in iPLA2 (data not shown). Positive control fMLP-induced migration remained unaffected by the reduction in iPLA2, confirming that there were no direct effects on the ability of the neutrophils to migrate (Fig. (Fig.8B).8B). Western blotting was performed to confirm and quantify the loss of iPLA2 due to siRNA with densitometric analysis by ImageJ (Rasband, 1997 to 2000; http://rsb.info.nih.gov/ij/), indicating that there was an ~50% reduction in iPLA2 in uninfected and infected HCT-8 monolayers expressing siRNA against PLA2G6 mRNA in comparison to what was seen for the control monolayers expressing nonspecific siRNA (Fig. (Fig.8C8C).
The current paradigm, based largely on studies using S. enterica serovar Typhimurium, predicates that enteric pathogens provoke PMN movement across the intestinal epithelium by stimulating the polarized release of distinct PMN chemoattractants from the epithelium. For example, both S. enterica serovar Typhimurium and S. flexneri activate the transcription factor NF-κB, resulting in the epithelial synthesis and basolateral release of the potent PMN chemokine interleukin-8 (15, 25, 44, 46, 48, 66, 68). Such basolateral secretion of interleukin-8 recruits PMNs through the matrix (lamina propria) to the subepithelial space but is not involved in PMN migration across epithelial tight junctions, the final step of crypt abscess formation (18, 29, 45). More recently, it was determined that the final step of S. enterica serovar Typhimurium-induced PMN movement across the epithelium is directed by the NF-κB-independent release of HXA3, a potent PMN chemoattractant and the primary product of arachidonic acid metabolism formed from the 12/15-LOX pathway (51).
Hepoxilins possess a wide range of biological activities, with the A3 form known to potentiate glucose-dependent insulin secretion (8, 56), open S-type K+ channels in Aplysia californica (55), modulate synaptic neurotransmission in rat hippocampus (58), increase vascular permeability in rat skin (34), and induce the chemotaxis of neutrophils (79). Our recent findings are the first to demonstrate that HXA3 is secreted from epithelial cells in a manner that is regulated by conditions which contribute to inflammation (51). These findings substantiate those made by Shannon et al., who found that in patients with inflammatory bowel disease the leukocyte-type 12-LOX is upregulated at sites of intestinal inflammation (72). Given that the human homologue of mouse leukocyte-type 12-LOX is 12/15-LOX (8), this reflects our usage of the 12/15-LOX nomenclature in reference to this part of the pathway.
Previously, we have shown that HXA3 is secreted apically by intestinal epithelial cells and subsequently forms a paracellular gradient through the tight junctional complex that prompts the guidance of PMNs across the epithelium (23, 51). Thus, the action of HXA3 represents the final gate to PMN transmigration across polarized epithelial barriers. In the present study, we demonstrate for the first time that S. flexneri is able to coordinate the signaling cascades that eventuate in the apical release of HXA3. We also found that disruption of the 12/15-LOX pathway (required for HXA3 synthesis) dramatically reduced S. flexneri-induced PMN transepithelial migration, supporting the notion that HXA3 is a key regulator of mucosal inflammation. Although previous studies have demonstrated the 5-LOX pathway to be a prominent proinflammatory cascade (17), we determined that an inhibitor of the 5-LOX pathway (caffeic acid) failed to inhibit S. flexneri-induced PMN migration (Fig. (Fig.2),2), suggesting that 5-LOX and its products participate in other aspects of inflammation. These results are consistent with prior reports that established the involvement of the 12-LOX pathway and the synthesis of HXA3 as critical steps underlying the mechanism by which S. enterica serovar Typhimurium elicits PMN transepithelial migration (51). These observations also suggest that the rate-limiting step in the production of HXA3 and the ensuing PMN movement may be at the level of availability of arachidonic acid (to be used as a substrate for the 12/15-LOX pathway). There are four enzymes that possess 12-LOX activity in humans [platelet type, 12(R), epidermal-type 12-LOX, and 12/15-LOX] (82-84). In humans, unlike what is seen for mice, 12/15-LOX is capable of synthesizing both 15-HpETE and 12-HpETE at a 4:1 ratio, where 12-HpETE is the minor product (31, 32). Since 12/15-LOX is highly expressed in epithelial cells, including intestinal epithelial cells (31, 32), the synthesis of 12-HpETE through this pathway could be significant. However, at present it is unclear which 12-LOX genes specifically contribute to HXA3 production, and thus the potential contribution of elox-3 and 12-platelet-LOX cannot be excluded. Future experiments will help delineate the involvement of the respective 12-LOX genes in HXA3 production.
The eicosanoids include two main families of arachidonic acid metabolites derived through the cyclooxygenase pathways (COX-1 and COX-2), producing prostaglandins and thromboxanes, or the LOX pathways, producing leukotrienes, lipoxins, and hepoxilins (57, 71, 80). Because of the potent pro- and anti-inflammatory effects exerted by these bioactive lipids, the enzymatic release of arachidonic acid from membrane phospholipids must be tightly controlled and its intracellular concentrations must be maintained at low levels in resting cells (1). A predominant pathway triggering the release of arachidonic acid from the membrane is that by the hydrolysis of membrane glycerophospholipids at the sn-2 position by the action of PLA2 (1, 3). There are at least three broad classes of phospholipases involved in generating eicosanoids; these are grouped based on cellular deposition and calcium dependence and include sPLA2, iPLA2, and cPLA2 (14). In our model of pathogen-induced PMN transepithelial migration, we envisage that arachidonic acid once released from the membrane becomes an available substrate for 12/15-LOX, which is metabolized to produce HXA3. Remarkably, we found that S. enterica serovar Typhimurium and S. flexneri induce PMN transepithelial migration through the utilization of distinct PLA2 enzymes; S. enterica serovar Typhimurium is dependent on iPLA2, whereas S. flexneri is dependent on cPLA2 activity.
These results raise a number of interesting points. First, it appears that S. enterica serovar Typhimurium and S. flexneri trigger distinct signal transduction cascades that lead to the differential utilization of PLA2 enzymes. In particular, we have previously shown that the S. enterica serovar Typhimurium effector protein, SipA, activates a novel signal transduction pathway involving an ADP ribosylation factor 6- and phospholipase D-dependent lipid signal transduction cascade that, in turn, activates PKC α, leading to PMN transepithelial migration (10, 73). Our finding that S. enterica serovar Typhimurium engages a PKC-dependent signal transduction cascade and requires iPLA2 for the induction of PMN transepithelial migration is consistent with studies that have demonstrated that an increase in iPLA2 activity is mediated by PKC (9, 50, 76). Comparatively, the underlying mechanisms which direct PMN transmigration induced by S. flexneri are less clear. Recently, however, we identified two S. flexneri effector proteins, OspF and OspC1, that activate the mitogen-activated protein kinase/ERK pathway (85). We also determined that the phosphorylation of ERK1/2 results in the production of signaling molecules required for PMN migration (26, 85). Again, our evidence demonstrating that S. flexneri induces an ERK1/2-dependent signaling cascade and depends on cPLA2 for the promotion of PMN transepithelial migration is substantiated by many studies (36, 38, 40, 75, 78) and suggests that an increase in cPLA2 activity is mediated by ERK1/2 (Fig. (Fig.9).9). We have also found that an ERK inhibitor which attenuates S. flexneri-induced PMN migration also caused a marked decrease in the phosphorylation of cPLA2 during S. flexneri infection (K. L. Mumy and B. A. McCormick, unpublished observations). However, further investigation is required to validate the linkage of these signal transduction pathways (i.e., PKC and ERK1/2) induced by S. enterica serovar Typhimurium and S. flexneri to the activation of iPLA2 and cPLA2, respectively.
Second, our results demonstrate that the mechanisms underlying PMN transepithelial migration induced by these pathogens converge at the point of arachidonic acid release, indicating that this distal part of the signal transduction pathway is conserved (Fig. (Fig.9).9). In line with this evidence, we have previously shown that PLA2 activity is pivotal for mediating Pseudomonas aeruginosa-induced PMN migration across airway epithelial cells through its ability to generate the precursor arachidonic acid as a substrate for 12/15-LOX and its subsequent conversion to HXA3 (24). On the balance of these observations, we hypothesize that increased PLA2 activity, which mediates arachidonic acid release, is a conserved innate immune mechanism for the detection and eradication of pathogens interfacing with the host mucosal surface.
Only a limited number of studies have explored the role of PLA2 in the pathogenesis of enteric bacteria. Pace et al. showed that an sPLA2 inhibitor reduced the ability of wild-type S. enterica serovar Typhimurium to invade epithelial cells (54). However, in our study using the same inhibitor, we failed to observe attenuation in the PMN transepithelial migration response induced by either pathogen. It should be noted that our maximum dose of the sPLA2 inhibitor was 15-fold less than that used by Pace et al. (54). We were unable to perform our experiments using higher doses of the inhibitor, since we observed cytotoxic effects at higher doses (M. Pazos and B. A. McCormick, unpublished observations). Furthermore, we did not observe differences in the invasion capacities of S. enterica serovar Typhimurium and S. flexneri when the intestinal cell monolayers were exposed to the iPLA2- and cPLA2-specific inhibitors, respectively. These results indicate that in our model system invasion was not adversely influenced by PLA2 enzyme inhibitors at the doses used and thus is not the basis for the noted reduction in PMN transepithelial migration under these conditions.
To our knowledge, our current study is the first to have directly assessed the effects of PLA2 or arachidonic acid metabolites in the pathogenesis of Shigella flexneri. PLA2 activity has been reported to be involved in the pathophysiology of intestinal inflammation. Most studies, however, have focused on the expression and activity of PLA2 in Crohn's ileitis. Such reports indicate that Crohn's disease is associated with increased gene expression of PLA2-II (an sPLA2 family member) at the site of the active inflammation in the ileal and colonic mucosa (20, 37). However, it remains unclear whether the upregulation of PLA2-II in the intestinal mucosa of Crohn's disease patients is a beneficial host defense mechanism or a detrimental reaction of the disease process that may contribute to tissue destruction. Thus, while the role of PLA2 in the intestinal mucosa during active states of intestinal inflammation needs to be better defined, recent studies have demonstrated that in a model of 2,4,6-trinitrobenzene sulfonic acid-induced colitis, treatment with an sPLA2 inhibitor profoundly attenuated the development of colonic inflammation (28).
In conclusion, we found for the first time that S. flexneri induces the apical release of the potent PMN chemoattractant HXA3. We also provide evidence to suggest that S. enterica serovar Typhimurium and S. flexneri mediate PMN transepithelial migration through the use of different PLA2 isoforms, revealing a fundamental step underlying the molecular basis of this induced inflammatory response (Fig. (Fig.9).9). Future studies into the molecular mechanisms leading to the activation of specific PLA2 enzymes may not only improve our understanding of intestinal inflammation provoked by infectious or idiopathic processes but also lead to the development novel therapeutics aimed at ameliorating inflammation at mucosal surfaces (i.e., gut and lung).
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases grants DK-56754 and DK-33506 to B. A. McCormick, National Eye Institute grant EY01613604 to K. Gronert, and a T32 training grant sponsored by Harvard Medical School and the Department of Surgery at Massachusetts General Hospital to K. L. Mumy.
Editor: J. B. Bliska
Published ahead of print on 27 May 2008.