Testing the Efficacy and Toxicity of Adenylyl Cyclase Inhibitors against Enteric Pathogens Using In Vitro and In Vivo Models of Infection

Bacterial strains, reagents, and growth conditions.Bacterial strains used in this study include ETEC strains H10407 and DEC7A, enteropathogenic E. coli (EPEC) O127:H6 (isolate E2348/69), enterohemorrhagic E. coli (EHEC) O157:H7 (isolate 86-24), enteroaggregative E. coli (EAEC) O42, Salmonella enterica serovar Typhimurium 2157, Shigella flexneri, and V. cholerae 569B. Strains were routinely grown overnight in Luria-Bertani (LB) medium at 37°C, and prior to the animal infection, the bacterial culture concentration was estimated spectrophotometrically and samples were diluted to the required working concentration. DC5 for these studies was synthesized in-house from commercially available precursors. The compound was dissolved in a minimal volume of anhydrous dimethyl sulfoxide (DMSO, usually to 100 mM) and then diluted to 10 mM into phosphate-buffered saline (PBS) buffer plus two molar equivalents of NaOH. Samples were then diluted 10× (1 mM) or 100× (0.1 mM) into cell culture or other medium before use in cell culture assays or for animal studies. Appropriate controls indicated that the end concentration of DMSO (1% DMSO) had no effect on the assays. The stock of LT toxin (kindly provided by J. Peterson) was suspended in 1 ml of sterile PBS (1 mg/ml), and the LT toxin in solution was transferred to a freezer vial and stored at −80°C until use.

In vitro cell-based cAMP assay. (i) Cell propagation.Murine monocyte/macrophage cells (RAW 264.7) were propagated in T75 flasks containing Dulbecco's modified Eagle's medium (DMEM) (Mediatech, Inc., Herndon, VA) at 37°C with 5% CO₂. The culture media contained 10% heat-inactivated fetal bovine serum (FBS), l-glutamine, and 100 μg/ml penicillin/streptomycin.

(ii) cAMP assays.Cells were plated at a cell density of 5 × 10⁵ cells/ml in 48-well assay plates and incubated overnight. The next day, RAW 264.7 cells were incubated with various concentrations of the inhibitors (from 100 μg/ml to 0.1 μg/ml) in the presence of purified protective antigen (PA, 2.5 μg/ml), edema factor (EF, 0.625 μg/ml), and 50 μM 3-isobutyl-1-methylxanthine (IBMX) (BIOMOL, Plymouth Meeting, PA) in triplicate for 4 h. To determine the amount of cAMP released, we used the Correlate-EIA direct cyclic AMP kit (Assay Designs, Ann Arbor, MI) to quantitatively determine amounts of extracellular cAMP from supernatants in duplicate.

Determination of IC₅₀.Three experiments in triplicate were performed, the average values were used to construct dose-response curves, and the 50% inhibitory concentrations (IC₅₀s) were calculated (Tablecurve 2D; AISN Software, EUA, 1996). When possible, we used one algorithm (logistic dose response) as our equation of choice.

Small-animal model of infection with ETEC.All mouse intestinal assay procedures were approved by the University of Texas Medical Branch (UTMB) Institutional Animal Care and Use Committee. The diet-restricted, antibiotic-treated mouse model was used to evaluate the effect of the adenylyl cyclase inhibitors on various ETEC strains. We selected the mouse intestinal model to study enteropathogenicity of E. coli strains (6, 24) that our lab has modified from the work of Allen et al. (1). Briefly, we used female ICR mice of 20 to 25 g (Charles River Laboratories) that were housed in the pathogen-free animal facility at UTMB upon arrival for 72 h prior to experiments. The animals' food was restricted and altogether removed 12 h prior to inoculation. All animals received streptomycin (5 g/liter in drinking water supplemented with 7% fructose) for 48 h prior to oral inoculation with ETEC (32). Four h prior to inoculation, the fructose-treated water was replaced with unsweetened, sterile water. Cimetidine (50 mg/kg) was injected into all mice 1 to 3 h prior to oral inoculation with ETEC strains (27). Groups of 6 to 10 mice were orally inoculated with a suspension of ETEC bacteria in the presence or absence of the inhibitors (0.1 mM or 1 mM DC5 or PGE₂-imidazole) in a final volume of 0.4 ml delivered by gavage (20-gauge needle). In some cases, mice were dosed with various concentrations of the compounds (200 μl) via the intraperitoneal route at 12-h intervals (12, 24, 36, 48, and 60 h). The animals were maintained for 24 or 72 h depending on the experimental protocol, but 12 h prior to euthanasia, food was withdrawn. To collect the fluid accumulated in the small intestine, ligatures of 2-0 Vicryl suture were created in each mouse and placed at the junctions of the pylorus and duodenum, as well as the ileum and cecum. After euthanasia, the entire small intestine of each mouse was excised and weighed. To collect total bacteria, 200 μl of PBS was then injected into one end of the ligated intestine, the other end was cut, and the total content, containing bacteria, was then harvested from the small intestine and collected in microcentrifuge tubes. Serial dilutions of the intestinal content were plated in duplicate on MacConkey agar plates or LB-streptomycin agar plates and incubated at 37°C overnight. The quantification of the organisms is reported as CFU/intestinal segment. Fluid accumulation was analyzed by a two-tailed Student's t test for independent samples or by Dunnett's multiple group comparison test.

Hematology.In order to perform a cursory examination of DC5's toxicity, various blood markers were examined. CD-1 mice were anesthetized with isoflurane, and a cardiac puncture was performed to remove roughly 1 ml of blood. One half of the blood volume was placed in a K2EDTA tube (BD Microtainer) to determine leukocyte counts via a Hemavet 950FS. The other half of the blood was placed in a Z tube (BD Microtainer) and allowed to coagulate at room temperature for 4 h, at which time the samples were centrifuged and serum was extracted. The serum was analyzed in a Vegasys blood chemistry analyzer for alanine aminotransferase (ALT), aspartate aminotransferase (AST), and creatine phosphokinase (CPK), common indicators of toxicity.

Preparation of samples for electron microscopy.Segments of mouse intestinal tissue infected with the wild-type ETEC strain were collected from ileal segments, washed gently with PBS, and fixed in a mixture of 2.5% formaldehyde, 0.1% glutaraldehyde, 0.03% trinitrophenol, and 0.03% CaCl₂ in 0.05 M cacodylate buffer (pH 7.2). After being washed in 0.1 M cacodylate buffer, the tissues were processed further by postfixing in 1% OsO₄ in the same buffer, stained en bloc in 1% uranyl acetate in 0.1 M maleate buffer (pH 5.2), dehydrated in ethanol, and embedded in Poly/Bed 812 (Polysciences, Warrington, PA). Specimens were examined in a Philips 201 electron microscope. To eliminate bias in our electron microscopy (EM) analysis, we took samples from representative mice and embedded 10 pieces of tissue from each sample, randomly selecting pieces for cutting and further processing. There was one observer examining the sections blindly and taking pictures randomly of each sample.

RNA extraction and preparation for RT-PCR.Animals infected with ETEC or ETEC plus DC5 were euthanized, and 5- to 10-cm sections of the small intestine were removed and immediately flushed with PBS to eliminate nonadherent bacteria. The intestinal segments were macerated and resuspended in 10-fold volumes of RNAprotect (Qiagen). Samples stabilized with RNAlater (Invitrogen) were snap-frozen with liquid nitrogen, stored at −80°C, and thawed on ice prior to bacterial extraction. Samples were treated with 1% Triton X-100 and vortexed to remove adherent bacteria. For purification, a 5-fold dilution of the RNAlater in distilled water was performed and connective tissue and cellular debris were removed by centrifugation at 1,000 × g for 10 min at 4°C. Supernatant was collected and bacteria were pelleted by centrifugation at 12,000 × g for 20 min at 4°C. Bacterial RNA was extracted using a Qiagen RNeasy Protect Bacteria mini kit. Quality of RNA was assessed by determining the ratio of optical density at 260 nm to the optical density at 280 nm (OD_260/280 ratio) and by visualization following agarose gel electrophoresis and ethidium bromide staining. The cDNA was synthesized using the SuperScript first-strand synthesis system for reverse transcriptase PCR (RT-PCR) (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The resulting cDNA was utilized for regular PCR with gene-specific primers to amplify elt (heat-labile enterotoxin of ETEC) (forward 5′-GGC GAC AGA TTA TAC CG TGC-3′ and reverse 5′-CGG TCT CTA TAT TCC CTG TT-3′) and using PCR conditions previously described by our group (5). A positive control with genomic DNA and a negative control with no reverse transcriptase added were also used in the assay.

Adhesion assays with different strains in the presence or absence of DC5.Adherence of bacterial strains to cultured HeLa cells was evaluated as previously described (5). DMEM with 10% (vol/vol) heat-inactivated fetal bovine serum, 2 mM l-glutamine, penicillin (100,000 IU/liter), and streptomycin (100 mg/liter) was used to grow HeLa cells prior to bacterial infection. Bacteria were grown statically in LB broth overnight at 37°C and inoculated at a multiplicity of infection of approximately 10:1 onto semiconfluent cultured epithelial cell monolayers grown in 24-well microtiter plates. Before use, the cells were washed with sterile phosphate-buffered saline (PBS, pH 7.4) and replenished with DMEM. For in vitro experiments assaying the effect of exogenous LT, HeLa cells received 100 ng/ml LT in the presence or absence of DC5 (0.1 mM) compound prior to ETEC infection, and they were incubated for 3 h at 37°C and 5% CO₂. To quantify bacterial adherence, the cells were washed two times with PBS and bacteria were recovered with 0.1% Triton X-100 in PBS buffer and plated on Luria agar plates.

DC5 prevents accumulation of cAMP by cultured cells.Recently, a structure-based procedure, involving compound library screening using a fragment-based 3D-pharmacophore, was used to identify compounds that should bind to the active site of edema factor, an adenylyl cyclase toxin (3). A small number of diverse compounds were identified that reversed cAMP secretion induced by these toxins and had no apparent toxicity on cultured cells (3, 4). One of the leading compounds for blocking cAMP accumulation by cultured cells, named DC5 {3-([9-oxo-9H-fluorene-1-carbonyl]-amino)-benzoic acid}, was selected for our analysis. The inhibitory effect of the DC5 compound on preventing accumulation of cAMP was compared to that of a previously known inhibitor, PGE₂-imidazole (21). Data illustrated in Fig. 1A demonstrated that DC5 had more activity than PGE₂-imidazole inhibiting the accumulation of cAMP caused by edema toxin (EDTx) treatment, as shown by the lower IC₅₀ (IC₅₀ = 8.99 μM for DC5 [R² = 0.986] compared to IC₅₀ = 14.10 μM for PGE₂-imidazole [R² = 0.992]). Similarly, DC5 was effective in reducing the accumulation of cAMP due to treatment of RAW 264.7 cells with cholera toxin (Fig. 1B; CT and LT share an identical mode of action), as indicated by the lower IC₅₀ (IC₅₀ = 2.04 μM [R² = 0.987]). These data suggest that DC5 is an effective inhibitor of toxin activity in vitro.

• Open in new tab
• Download powerpoint

FIG. 1.

In vitro cell-based cAMP assay. (A) Bioassay with different concentrations of PGE₂-imidazole and DC5 demonstrating that these compounds inhibited cAMP production induced by edema toxin (EdTx) on RAW 264.7 cells, with IC₅₀s of 14.1 and 8.99, respectively. Controls were medium alone or with the addition of 100 μM each individual compound tested (protective antigen [PA] or edema factor [EF]). (B) DC5 effect on cAMP production by cells treated with CT (equivalent to LT, IC₅₀ of 2.04).

Bacterial growth kinetics in the presence or absence of DC5.Next, we determined whether DC5 had any effect on growth of bacterial strains producing the LT toxin or other enteric organisms (Fig. 2 and data described below). First, we performed a series of in vitro growth studies to examine whether DC5 had any antibacterial effect on cultures of our prototype ETEC H10407 grown in Luria-Bertani (LB) broth in the presence or absence of 0.1 mM DC5. Growth of ETEC was monitored by measuring the OD₆₀₀ at 1-h intervals for 6 h, as well as quantification by serial dilution and plating bacteria at 0 h and 6 h. Our data showed that DC5 had no effect on the growth of ETEC (the numbers of CFU recovered on the plates matched the OD₆₀₀ data), which suggested that the effects observed during intestinal colonization of our animal model (see sections below) are not linked to inhibition of the growth or killing of the bacteria by the DC5 treatment (Fig. 2A). Based on these results and previous studies demonstrating that DC5 inhibits the activity of CT in the murine ileal loop model (data not shown), we tested whether this compound had activity against LT-producing ETEC.

• Open in new tab
• Download powerpoint

FIG. 2.

Bacterial growth in LB medium in the presence or absence of DC5. The different bacterial cultures were monitored spectrophotometrically for a period of 6 h. (A) ETEC H10407, EPEC E2348/69, EHEC 86-24. (B) ETEC DEC7A, EPEC E2348/69, S. Typhimurium, EAEC O42.

Establishing the optimal DC5 dose.The initial studies were used to establish an optimal DC5 concentration in our murine model. CD-1 mice (n = 8/group) were divided into four groups on the basis of dose received: (i) DMSO (at the concentration used to solubilize DC5) plus PBS; (ii) ETEC (1 × 10⁸) bacteria; (iii) ETEC bacteria plus 1 mM DC5; (iv) ETEC bacteria plus 0.1 mM DC5. The DC5 intraperitoneal (i.p.) boosters were administered at 12, 24, 36, 48, and 60 h prior to euthanasia at 72 h. At that time point, intestine weights and CFU counts were established; as shown in Fig. 3, DC5 effectively reduced the intestinal fluid accumulation after experimental ETEC infection, and although the difference is not dramatic, it was statistically significant (Fig. 3A; P = 0.02). Statistically significant differences were also observed between results for mice infected with ETEC and those receiving the vehicle (PBS plus DMSO) control (P = 0.03) and between results for animals treated with PBS plus DMSO control and ETEC-plus-DC5-treated animals (P = 0.003). To our surprise, the number of bacteria recovered from the intestine after infection was significantly reduced in the infected animals treated with DC5 compared with results for controls (Fig. 3B). Because the differences between 0.1 and 1 mM DC5 were not significant, we decided to use the lower DC5 concentration (0.1 mM in the 0.4-ml dose) for subsequent experiments.

• Open in new tab
• Download powerpoint

FIG. 3.

Establishing the optimal DC5 dose. (A) Increase in fluid (estimated by weight of the intestine) was reduced in the presence of DC5. P values (t test): *, <0.05 (comparing DMSO plus PBS control with ETEC treatment); **, <0.05 (ETEC treated with ETEC plus 1 mM DC5). (B) The DC5 treatment had an impact on the amount of bacteria (ETEC) recovered from the intestinal lumen at 72 h, particularly at 0.1 mM DC5. P values (t test): *, <0.01 (comparing ETEC treatment with ETEC plus 0.1 mM DC5 or ETEC plus 1 mM DC5). Each group included eight mice.

Comparing the effects of DC5 and PGE₂-imidazole in the murine model of ETEC infection.Because our data in vitro indicated that DC5 has more activity than PGE₂-imidazole, we then determined whether dosing mice with DC5 was more effective than, or as effective as, treating the animals with PGE₂-imidazole at the same concentration. Similar to Fig. 3, animals were divided into groups and received ETEC (1 × 10⁸) with or without DC5 (0.1 mM) or PGE₂-imidazole (0.1 mM). DC5 or PGE₂-imidazole i.p. boosters were administered at 12, 24, 36, 48, and 60 h prior to sacrifice of the animals at 72 h. As shown in Fig. 4, DC5 was slightly more effective in reducing the fluid accumulation in the intestine than PGE₂-imidazole (Fig. 4A). Most importantly, both compounds had a direct effect on the number of bacteria recovered from the intestinal lumen (Fig. 4B). These results strongly suggested that the ability of the ETEC bacteria to colonize and persist in the intestine was associated with the presence of the LT toxin, an effect that was diminished upon the addition of PGE₂-imidazole or DC5.

• Open in new tab
• Download powerpoint

FIG. 4.

Comparing the effects of DC5 and PGE₂-imidazole. (A) Reduction in the weight of the intestine (indicative of fluid accumulation) was observed in DC5- and PGE₂-imidazole-treated mice. P values (t test): *, <0.05 (comparing ETEC treatment with ETEC plus 0.1 mM DC5). (B) Both compounds had a significant effect on bacterial recovery from the intestinal lumen at 72 h postinfection. P values (t test): *, <0.01 (comparing ETEC treatment with ETEC plus 0.1 mM DC5 or ETEC plus 0.1 mM PGE₂-imidazole). Each group included eight mice.

We then sought to perform an ultrastructural analysis of the infected intestinal tissues by electron microscopy to determine whether the inhibitors had a detrimental impact on the epithelia while diminishing the effect of the LT toxin. As shown in Fig. 5, ETEC-only infected tissues showed destruction of the microvilli of the cells, with areas of cell death detected in the sections analyzed (Fig. 5A to C). The integrity of the tissue was consistently different from that observed in control experiments where the intestine was infected with a nonpathogenic E. coli strain (data not shown). When the ETEC-infected animals were treated with PGE₂-imidazole, we noticed changes in the structural integrity of the intestinal epithelial barrier, however, and upon close examination, we noticed that the microvilli of the cells looked slightly damaged (small areas of microvillus destruction), but no evidence of cell death was found (Fig. 5D to F). In contrast, the integrity of the intestinal epithelial barrier was fully maintained in those CD1 mice infected with ETEC but receiving DC5 (Fig. 5G to I). We then evaluated whether treatment with DC5 caused damage to other organs (spleen, liver, or the rest of the intestine) or affected blood cell counts or liver enzymes. Our analysis indicated that the number of leukocytes in circulation was not affected in the animals treated with DC5. Further, no differences in the weight or the ultrastructure of liver, spleen, and intestine were observed compared to those of control mice (data not shown). Our results suggest that the microvilli on the epithelial cells remained intact upon treatment with the two compounds, with no evidence of cellular death, and that DC5 treatment was more effective than PGE₂-imidazole in preventing intestinal damage due to ETEC infection.

• Open in new tab
• Download powerpoint

FIG. 5.

EM images of infected intestine. Ultrastructural studies of the intestine infected with ETEC only (A to C), ETEC treated with PGE₂-imidazole (D to F), and ETEC with DC5 (G to I). ETEC-infected intestines demonstrated destruction of the microvilli and cell death, while integrity of the microvilli in the compound-treated animals remained intact. Control tissue (noninfected) was used for comparison (not shown). Magnification: panels A, D, and G, ×16,000; panels B, E, and H, ×32,000; panels C, F, and I, ×64,000.

To confirm that the effect in the microvilli was due to the interaction of the bacteria with the intestinal mucosa, resulting in production of toxins, and not just to a lack of interaction with the intestinal cells, we performed an RT-PCR analysis of bacteria attached to the intestinal mucosa (n = 6 per group). Our result indicated that a positive PCR was obtained from all mice infected with ETEC but only in half of mice receiving ETEC plus DC5. This result suggested that the presence of DC5 inhibits LT toxin activity, affecting the ability of ETEC to colonize the intestine (data not shown). Subsequent experiments were designed to determine whether the reduction of fluid accumulation observed at 72 h postinfection in animals treated with DC5 was achieved early during infection. In this case, animals infected with ETEC (1 × 10⁸) and treated with DC5 (0.1 mM) at time zero received DC5 boosters administered i.p. at 12, 24, 36, 48, and 60 h. Groups of animals were euthanized at 24, 48, and 72 h. We observed a transient increase in fluid accumulation at 24 h in animals treated with DC5 compared to accumulation in ETEC-only infected animals, but this fluid accumulation progressively decreased over a 24- to 72-h time period (data not shown). Interestingly, we still observed 3-log differences in the number of bacteria recovered from animals euthanized at 72 h (P < 0.01 in animals treated with ETEC plus DC5 versus those treated with ETEC alone).

Determining whether DC5 route of administration had an effect on bacterial recovery.To evaluate whether the route of administration of DC5 had an impact on bacterial recovery, ETEC-infected animals received DC5 by the oral or the i.p. routes at time zero and DC5 boosters were administered via i.p. or orally at 12, 24, 36, 48, and 60 h postinfection. Animals were euthanized at different time points. Physical signs of damage (inflammation, fluid accumulation, bleeding, etc.) were evaluated, along with determination of the bacterial counts. No signs of bleeding or inflammation were observed in any of the animals receiving DC5, but fluid accumulation was observed at 24 h (ETEC plus DC5 oral dosing). Bacteria were recovered in animals infected with ETEC alone, and a significant reduction (P < 0.01) in the ETEC bacterial counts was observed in mice treated with DC5, especially at 72 h postinfection, regardless of the route of administration (Fig. 6). No bacteria were recovered in animals receiving only DC5 (data not shown).

• Open in new tab
• Download powerpoint

FIG. 6.

Comparing effects of DC5 routes of administration on bacterial survival. ETEC-infected animals received DC5 treatment using the oral or the intraperitoneal routes at 24, 48, and 72 h postinfection, and the CFU in the intestine were calculated. P values (t test): *, <0.01 (comparing ETEC treatment with ETEC plus 0.1 mM DC5 or ETEC plus 0.1 mM PGE₂-imidazole). P values (t test): *, <0.01 (comparing treatments at 72 h, ETEC treatment versus ETEC plus 0.1 mM DC5 IP or ETEC plus 0.1 mM DC5 orally administered). Each group included eight mice.

Bacterial adhesion to cultured epithelial cells.First, we determined whether DC5 had any antibacterial effect (growth defect) on cultures of other enteric pathogens (including isolates of pathogenic E. coli [ETEC, EHEC, EPEC, and EAEC] and S. Typhimurium) grown in Luria-Bertani (LB) broth in the presence or absence of 0.1 mM DC5 (Fig. 2). Similar to our results with the prototype ETEC H10407, we showed that DC5 had no effect on the growth of any of the strains tested, which further supports that the effects observed during ETEC intestinal colonization are not the result of inhibition of the growth or killing of the bacteria by the DC5 treatment (Fig. 2). Because we did not observe any difference in growth rate of the bacteria treated with DC5, we next determined whether DC5 affected bacterial adhesion to epithelial cells. Semiconfluent cultured HeLa cell monolayers (20) were infected (mutiplicity of infection [MOI] of approximately 10:1) with the pathogenic E. coli (ETEC [two strains], EPEC, EAEC, and EHEC) as well as S. flexneri, S. Typhimurium, and V. cholerae, in the presence or absence of DC5. ETEC strains produced LT; V. cholerae produced CT, and the other pathogens produced a variety of secreted/translocated toxins. As shown in Fig. 7, we found a reduction in adherence with the two ETEC strains treated with DC5 (P < 0.05), compared to that with the cells infected with only the ETEC strains. Interestingly, we also observed reduction in adherence in EAEC, EPEC, and EHEC, strains that do not produce LT toxin. In contrast, a slight increase in adherence was observed in V. cholerae, S. enterica, and S. flexneri (not statistically significant). It was surprising to observe an increase in adherence of V. cholerae in the presence of DC5, which suggests that adherence of this pathogen might be independent of the production of the CT toxin.

• Open in new tab
• Download powerpoint

FIG. 7.

Effect of DC5 and LT on bacterial adhesion to HeLa cells. Bacteria were inoculated at a multiplicity of infection of 10:1 onto semiconfluent cultured epithelial cell monolayers in the presence or absence of DC5 or exogenous LT toxin. Bacteria and cells were incubated for 3 h, and the adherent bacteria were quantified by serial dilutions. (A) ETEC H10407, EAEC O42, V. cholerae 596B; (B) EPEC E2348/69, EHEC O157:H7, ETEC DEC7A; (C) S. enterica serovar Typhimurium, S. flexneri. Each panel represents duplicate independent experiments performed in triplicate. *, P = 0.05; **, P = 0.01 (t test; comparing treated samples to their corresponding wild-type strains).

To determine whether the LT toxin was playing a selective role in adherence for the ETEC strains, we performed additional experiments where exogenous toxin was added to the cultured media in the presence or absence of DC5 and prior to the pathogens' infections (Fig. 7). We observed that the adherence of the majority of the strains was not increased in the presence of the LT toxin and in contrast, several strains displayed a reduction in adherence compared to that of their respective reference wild-type strains. Due to the presence of the exogenous LT toxin, DC5 effect on adherence was eliminated, perhaps due to binding of the compound to the toxin found in solution. Overall, our results suggested that DC5 is a promising compound that can be used to treat infections caused by LT-positive bacteria and, potentially, other toxin-producing isolates, and they supported the importance of adenylyl cyclase LT toxin as a mediator of binding of ETEC strains to epithelial cells.