• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. Apr 2011; 79(4): 1696–1705.
Published online Jan 18, 2011. doi:  10.1128/IAI.01099-10
PMCID: PMC3067541

Virulence Inhibition by Zinc in Shiga-Toxigenic Escherichia coli[down-pointing small open triangle]


Previously, our laboratories reported that zinc inhibited expression of several important virulence factors in enteropathogenic Escherichia coli (EPEC) and reduced EPEC-induced intestinal damage in vivo. Since EPEC is genetically related to Shiga-toxigenic E. coli (STEC), we wondered whether the beneficial effects of zinc extended to STEC as well. Treatment options for STEC infection are very limited, since antibiotics tend to exacerbate disease via enhanced toxin production, so a safe intervention for this infection would be welcome. In this study, we report that in STEC strains zinc inhibits adherence to cultured cells as well as expression of EHEC secreted protein A (EspA). In addition, zinc inhibits the expression of Shiga toxin (Stx) at both the protein and the RNA level. Zinc inhibits basal and antibiotic-induced Stx production and inhibits both Stx1 and Stx2 by ≥90% at a concentration of 0.4 mM zinc. Rabbit EPEC strains were selected for acquisition of Stx-encoding bacteriophages, and these rabbit STEC strains (designated RDEC-H19A and E22-stx2) were used to test the effects of zinc in vivo in ligated rabbit intestinal loops. In vivo, zinc reduced fluid secretion into loops, inhibited mucosal adherence, reduced the amount of toxin in the loops, and reduced STEC-induced histological damage (villus blunting). Zinc has beneficial inhibitory effects against STEC strains that parallel those observed in EPEC. In addition, zinc strongly inhibits Stx expression; since Stx is responsible for the extraintestinal effects of STEC infection, such as hemolytic-uremic syndrome (HUS), zinc might be capable of preventing severe sequelae of STEC infection.

Shiga-toxigenic Escherichia coli (STEC), also called enterohemorrhagic E. coli (EHEC) or verotoxigenic E. coli (VTEC), is a type of diarrhea-producing E. coli more common in developed countries than in developing countries. More-severe cases are characterized by bloody diarrhea, and some patients go on to develop severe systemic complications, such as hemolytic-uremic syndrome (HUS) and encephalopathy. STEC is genetically and evolutionarily related to enteropathogenic E. coli (EPEC) (7), an E. coli pathotype known for causing long-lasting watery diarrhea in children in developing countries.

Zinc has been shown to reduce the duration and severity of acute diarrhea in children in field trials on several continents (1, 24). In earlier decades, these beneficial effects of zinc were usually attributed to the correction of zinc deficiency (29), which is unfortunately still seen in poor, malnourished children. In more recent years, however, studies have shown that zinc can be beneficial against diarrhea even in patients without zinc deficiency (2, 21).

In a previous study, we showed that zinc reduced the expression of several important virulence factors in EPEC and reduced EPEC-induced intestinal damage and fluid secretion in ligated rabbit intestinal loops in vivo (5).

Since the locus of enterocyte effacement (LEE) is regulated in similar ways in STEC and EPEC, we investigated whether zinc had any similar protective properties against STEC virulence factors in the LEE. We also tested whether the inhibitory effects of zinc on the expression of EHEC secreted protein A (EspA) were dependent on Ler, the LEE-encoded regulator, as they were in EPEC. Next, we tested the ability of zinc to inhibit production of Shiga toxins (Stx), which are not present in EPEC but are the hallmark of STEC strains. Zinc inhibited basal as well as antibiotic-induced expression of both Stx1 and Stx2 in a variety of STEC strains, with ≥90% inhibition of Stx achievable with zinc concentrations of 0.4 to 0.5 mM. In vivo, zinc protected rabbit intestinal loops from STEC-induced histological damage, reduced STEC adherence to epithelium, and reduced Stx production in loop fluids. However, STEC bacterial numbers were similar in zinc-treated and control loops.

The antivirulence effects of zinc extend to STEC as well as EPEC and are independent of the zinc nutritional status of the host. Further work is needed to understand the mechanism of Stx inhibition by zinc and to better define the protective dose of zinc in humans. In addition, we wonder if zinc's inhibitory effects are limited to EPEC and STEC or if they extend to other bacterial enteropathogens as well.


Bacterial strains used.

Bacterial strains used are listed in Table Table1.1. Bacteria were grown overnight in LB broth at 37°C with shaking at 300 rpm and then subcultured into DMEM. In this report, when bacteria were subcultured in “DMEM,” this refers to Dulbecco's modified Eagle's medium-F-12 medium supplemented with 18 mM NaHCO3 and 25 mM HEPES, pH 7.4, but without serum or antibiotics. DMEM-F-12 medium contains 1.5 μM zinc. The Δler mutant of EDL933, CB49, was a gift from Alfredo Torres, University of Texas Medical Branch, Galveston, TX.

E. coli O157:H7 strains used in this research


The following reagents were obtained from Sigma-Aldrich Chemicals: zinc acetate, NiCl2, MnCl2, CuSO4, and Ga(NO3)3. FeSO4 was from MP Biomedicals, Irvine, CA.

Cell culture.

HeLa cells were grown in DMEM-F-12 medium supplemented with 7.5% fetal bovine serum (Gibco/Invitrogen, Grand Island, NY), 18 mM NaHCO3, 20 μg/ml vancomycin, and 15 μg/ml gentamicin as previously described (4).

Bacteriophage transduction of rabbit EPEC strains to generate Stx-producing rabbit-adapted E. coli strains.

The methods used to generate strain RDEC-H19A, in which bacteria carrying the antibiotic-resistance-tagged Stx1 phage from an O26 strain were cocultured with the recipient RDEC-1 strain, have been described previously (22). Similar methods were used to select strain E22-stx2. Briefly, O157:H7 strain 86-24 (Stx2 only) was used as the phage source. A 1.5-kb kanamycin resistance marker was introduced 27 kb downstream from the Stx2b coding sequence and between predicted open reading frames. This insertion did not affect Stx2 production (as determined by cytotoxicity analysis in vitro of cell lysates), and a lysogenic tagged strain was capable of being induced to produce entire phage particles. Following coculture with the recipient O103 E22 strain, kanamycin-resistant isolates which had acquired the Stx2 phage as confirmed by PCR and which were able to produce Stx without and with antibiotic induction were selected. One such isolate was named E22-Stx2. Rabbits inoculated with this Stx2-producing lysogenic rabbit EPEC (REPEC) strain developed more-severe histological lesions and inflammation than did those inoculated with the isogenic, non-toxin-producing wild-type strain E22.

Adherence assay.

Quantitative adherence assays were performed in duplicate or triplicate wells of HeLa cells grown in collagen-coated 24-well plates. HeLa cells were used at 90% confluence; cells were rinsed, changed into 0.25 ml per well of antibiotic-free DMEM, and then infected with STEC at a multiplicity of infection (MOI) of 100:1. STEC bacteria were grown overnight in LB broth and then subcultured for 2 h in DMEM. To enhance adherence, STEC bacteria were centrifuged onto the HeLa cells at 500 × g for 5 min at room temperature and then allowed to adhere for 2.5 h at 37°C in a 5% CO2 atmosphere. After 2.5 h of STEC adherence, the monolayers were washed twice with sterile phosphate-buffered saline (PBS) to remove unbound bacteria, and then the monolayers were solubilized in 0.1% Triton X-100 detergent in water. The Triton extracts were subjected to serial dilution and spread on LB agar plates in duplicate, and the number of colonies was counted. Adherence was calculated as parts per thousand (‰) of the input inoculum.

EIA for Shiga toxin protein.

Shiga toxin protein was measured by enzyme immunoassay (EIA) using a Premier EHEC EIA kit from Meridian Bioscience, Cincinnati, OH, which measures Stx1 and Stx2. Since different STEC strains produced different amounts of Shiga toxin, required different concentrations of antibiotics for induction of Stx, and so on, the optimal conditions for growth and EIA detection had to be determined by trial and error. The optimal antibiotic concentrations, culture durations, and dilutions of supernatant needed for successful Stx EIA detection for various STEC strains are summarized in Table Table2.2. Bacterial culture supernatants were filter sterilized prior to Stx measurement by EIA; therefore, any toxin trapped in the periplasmic space, often a substantial proportion for Stx1 (31), was not measured in our assay.

Conditions for detecting Stx production in STEC by enzyme immunoassay (EIA)a

Analysis of RNA expression by reverse transcription and quantitative real-time PCR.

Quantitative real-time PCR (qRT-PCR) was carried out as described previously (5, 6). Primers for stx1 were those described by Jinneman et al. (11), while those for stx2, STEC espA, and the normalizing gene gnd were those from the methods of Rashid et al. (18). Primers for measuring expression of espA and espB in rabbit EPEC strain E22 were reported previously (6).

Normalizing genes for analysis of expression were rrsB, using the primers reported by Leverton and Kaper (14), and gnd, as reported by Rashid et al. (18). Calculations of normalized expression using the two different normalizing genes gave quite similar results.

STEC infection in the ligated rabbit ileal loop model.

Details of the surgical procedures used to create the ileal loops were presented in the supplemental material of a previous article (5), including details of the methods used to provide analgesia and avoid pain. Experimental infection of rabbits was performed at the Boedeker laboratory in New Mexico. Animal use was approved by the Institutional Animal Care and Use Committee at the University of New Mexico. Briefly, 10-cm loops of ileum were created, and then the loops were injected with 108 to 109 CFU of RDEC-H19A (Stx1+) or E22-stx2 (Stx2+). Rabbits were reanesthetized 16 h later and killed humanely using an overdose of pentobarbital. The abdomen was reopened, the loop fluid was recovered, and the rabbit intestinal tissues were preserved in 10% buffered formalin for histological analysis. Bacterial CFU in the loop fluid were determined by dilutions and plate counts on LB medium plus the appropriate antibiotic. For counting of RDEC-H19A, the medium used was LB agar plus nalidixic acid (50 mg/liter) plus tetracycline (25 mg/liter), whereas for strain E22-stx2, it was LB plus nalidixic acid (50 mg/liter) plus kanamycin (50 mg/liter). The amount of Shiga toxin in loop fluid was measured by EIA as mentioned above. Quantitative histological analysis of infected intestinal tissues was done by an observer unaware of the treatment groups. Villus-to-crypt ratios and percent enteroadherence were measured as previously described (5).

Measurement of zinc concentrations in gastrointestinal contents of the rabbit after orogastric zinc administration.

To determine what lumenal concentrations of zinc were achieved in response to oral zinc, we measured zinc in various segments of the rabbit gastrointestinal (GI) tract after orogastric administration of three different zinc salts: zinc acetate, zinc sulfate, and zinc oxide.

Zinc oxide (ZnO) is known to be less ionized and less bioavailable that the other 2 zinc salts. Since the 3 salts have different molecular weights, we adjusted the oral dose to achieve equal, 10-mg doses of elemental zinc and administered each zinc salt by orogastric tube to rabbits once daily for 3 consecutive days. Intestinal contents were collected from the small intestine, proximal cecum, midcecum, and large intestine. Rabbits used in this experiment had an average body weight of 2.4 kg, yielding a zinc dose of 4.1 mg/kg (mg elemental zinc per kg body weight). Rabbits were not surgically altered, nor were they infected with STEC. Zinc content of the samples was determined at Michigan State University, a national reference laboratory, by inductively coupled plasma mass spectrometry (ICPMS). In addition to samples collected at the time of euthanasia, fecal pellets were collected daily, and the rabbits were killed humanely on the evening of the third day. Fecal pellets collected on the portion of day 3 were insufficient for analysis. Fecal zinc was reported as μg of elemental zinc per gram (dry weight) of feces. For example, the zinc content of the fecal pellets from control rabbits was 96.7 μg of elemental zinc per gram (dry weight) of feces, compared to 251.3 μg/g from fecal pellets of rabbits that received zinc acetate for 2 days. In order to compare the zinc concentrations in fecal pellets with those measured in intestinal fluids, the concentration of zinc in feces was converted to millimoles per liter by using the measured water content of feces of 85% and assuming a density of 1 g per ml of feces. Zinc was well tolerated by the rabbits, as determined by good physical appearance and behavior and by normal consumption of food and water during the experiment. Zinc levels in the control rabbits were also in the measurable range (see Fig. 6A). The standard rabbit chow at the University of New Mexico contains a mineral supplement including 60 mg/kg of elemental zinc. Based on normal rabbit chow consumption of 50 to 70 g/day, this means that even the control rabbits were receiving 3 to 4 mg of elemental zinc daily via their food.

Data analysis.

Error bars shown in graphs are standard deviations. Statistical analysis was carried out by analysis of variance (ANOVA) using InStat 3.0 software (GraphPad Software, San Diego, CA), with the Tukey-Kramer posttest for multiple comparisons.


Initially, we tested whether zinc had any effect on the adherence of human STEC strains to cultured cells in vitro. Figure Figure11 shows that 0.2 to 0.4 mM zinc almost completely abolished STEC adherence to HeLa cells in a short-term adherence assay. The molecular basis for STEC adherence to host cells is less well understood than that for EPEC, with different laboratories reporting many putative adhesins, including intimin, EspA, H7 flagella, outer membrane protein A (OmpA), long polar fimbriae, and hemolysin, to name just a few (16, 18, 22, 23, 26, 27). Therefore, while the inhibition of STEC adherence by zinc is important, it does not immediately point to a molecular mechanism.

FIG. 1.
Effect of zinc on the adherence of two STEC strains to HeLa cells. HeLa cell monolayers were infected with STEC, and then the multiwell plates were centrifuged at 500 × g for 5 min to enhance adherence. Adherence was allowed to proceed for 2.5 ...

Zinc inhibited expression of the EPEC secreted proteins EspA and EspB in strain E2348/69, so we tested whether zinc also inhibited expression of the esp genes in STEC strain EDL933. Zinc did inhibit esp expression in EDL933 (see Fig. Fig.2A2A for espA). In EPEC, the inhibitory effect of zinc on the Esp proteins showed an unexpected dependence on the LEE-encoded regulator, Ler, and was abolished in the Δler mutant. We compared the effect of zinc on the Δler mutant of EDL933, designated CB49, with the effect on the wild-type parental strain. Figure Figure22 A and B show that the inhibitory effect of zinc was lost in the Δler mutant of EDL933, just as it was in the EPEC ler mutant. To try to extend these findings even more, we compared the effect of zinc on the wild-type rabbit EPEC strain E22 with the effect on the Δler mutant of E22, strain ECB159. As with human EPEC and STEC, the inhibitory effect of zinc on espA and espB was lost in the Δler mutant of E22 (Fig. 2C and D). The findings shown in Fig. 2A to D, together with our previous study of human EPEC, show that the presence of Ler is necessary for the inhibitory effects of zinc on the EPEC and EHEC secreted proteins (Esps).

FIG. 2.
Role of Ler in the effect of zinc on expression of the EHEC secreted proteins (Esps). Expression of RNA transcribed from esp genes of STEC and rabbit EPEC was measured by quantitative real-time PCR (qRT-PCR) as described in Materials and Methods. (A and ...

Figure 2E and F show experiments testing whether the presence or absence of Ler had any effect on Shiga toxin (Stx) production. Figure Figure2E2E shows that the amounts of Stx produced under basal and antibiotic-stimulated conditions were identical in the wild-type strain EDL933 and the Δler mutant CB49. Similarly, zinc inhibited Stx production in the Δler mutant with efficacy and potency equal to those in the wild type (Fig. (Fig.2F).2F). To summarize, Ler is required for zinc's inhibitory effects on EspA and EspB (Fig. 2A to D) but is not required for its inhibitory effects on Stx (Fig. 2E and F).

Stx is an important virulence factor in STEC because Stx1 and Stx2 are responsible for the bloody diarrhea and extraintestinal complications observed in STEC infection. Therefore, we further tested zinc's ability to inhibit Stx expression in various human-derived STEC strains, including strains that produce Stx1 alone, Stx2 alone, or both toxins. In addition, since several antibiotics are known to increase the expression and release of the Shiga toxins, we tested whether zinc inhibited antibiotic-induced Stx expression. We measured Stx protein expression by enzyme immunoassay (EIA) and stx RNA abundance by quantitative real-time PCR (qRT-PCR).

Figure Figure33 shows the effect of zinc on Stx expression as assessed by EIA performed on culture supernatants; the EIA used detects both Stx1 and Stx2. Figure Figure3A3A shows that zinc inhibited basal toxin expression in strain EDL933 by 90%. In contrast, NiCl2 and MnCl2 stimulated Stx production to about 50% above control at lower concentrations (0.05 to 0.2 mM) before Stx fell off again at higher concentrations. We and others have noted the biphasic stimulatory effect of nickel on other virulence factors, including the bundle-forming pilus of EPEC (5, 17). Pilot experiments showed that ciprofloxacin induced Stx production in all strains tested, so ciprofloxacin was used in tests of antibiotic-induced toxin production. Figure Figure33 shows that zinc was able to inhibit ciprofloxacin-induced toxin production in strain Popeye-1 (Stx2 only) (Fig. (Fig.3B)3B) and in strain TSA14 (Stx1 only) (Fig. (Fig.3C)3C) as well as in EDL933 (Fig. (Fig.3E3E and data not shown).

FIG. 3.
Zinc inhibition of Shiga toxin production in various STEC strains and under various conditions, as measured by EIA with supernatant medium. Optimal conditions, including antibiotic concentrations and supernatant dilutions used, are summarized in Table ...

In order to test the effect of zinc in a rabbit intestinal model of STEC infection, two rabbit STEC strains were also tested for susceptibility to inhibition by zinc. Figure Figure3D3D shows that zinc inhibited Stx2 production by rabbit strain E22-stx2. Toxin production from rabbit strain RDEC-H19A was also inhibited by zinc (data not shown).

In addition to its ability to inhibit ciprofloxacin-induced Stx, zinc inhibited trimethoprim-induced Stx production and release (Fig. (Fig.3E).3E). Zinc also inhibited H2O2-induced Stx production (Fig. (Fig.3F).3F). H2O2 has been implicated as an inducer of Stx production in vivo in the absence of antibiotics (28). Figure Figure33 shows that zinc was able to inhibit Stx production from a variety of STEC strains, including strains that produce Stx1, Stx2, and both toxins; furthermore, zinc inhibited basal as well as antibiotic-induced toxin production.

Because of the promising results found using Stx protein measurements, we also measured stx RNA by qRT-PCR. Figure Figure44 A shows that zinc inhibited the abundance of RNA transcripts from stx2 in STEC strain Popeye-1. Basal stx2 expression was inhibited by 50%, while ciprofloxacin-stimulated stx2 expression was inhibited by 90%.

FIG. 4.
Zinc inhibition of stx2 RNA in strain Popeye-1, as determined by qRT-PCR. (A) Zinc inhibition of stx2 RNA under basal and ciprofloxacin-stimulated conditions at 4 h. (B) Comparison of the effects of various metals and the semimetal gallium on ciprofloxacin-stimulated ...

Using the same strain, we also compared the inhibitory ability of zinc with that of other divalent metals and gallium nitrate at a 0.3 mM concentration (Fig. (Fig.4B).4B). Gallium was tested because of its position adjacent to zinc on the periodic table and because Kaneko et al. reported that gallium has strong antivirulence and antibiofilm activity against Pseudomonas aeruginosa (12). Zinc and nickel chloride were the most inhibitory, followed by copper and then manganese and gallium. In contrast, addition of FeSO4 actually stimulated stx2 RNA even more than ciprofloxacin alone. Although nickel strongly inhibited stx2 expression, nickel is also much more toxic to humans than is zinc (Table (Table3).3). In EPEC, nickel also had a paradoxical stimulatory effect on expression of the bundle-forming pilus and on the EPEC secreted proteins (5), and the adherence-enhancing ability of nickel was confirmed in another E. coli strain (17). Because of these two undesirable features of nickel (increased toxicity and enhancement of pathogen adherence), we believe that zinc is the divalent metal with the best combination of desirable properties: inhibition of virulence expression in the pathogen and low toxicity toward the host.

Comparative toxicities of selected metals in humans

Figure Figure4C4C shows the effect on stx2 of delaying the addition of zinc for various amounts of time after addition of an inducing antibiotic (ciprofloxacin, in this case). Delay in zinc addition gradually reduced the inhibitory effect of zinc compared to that seen with immediate zinc addition. However, even after a 3-h delay, zinc still inhibited ciprofloxacin-induced stx2 expression by ~50%. The fact that zinc can still have an inhibitory effect on stx2 RNA levels when added 3 h after the inducing antibiotic has implications for the molecular mechanism by which zinc inhibits Stx production. However, the loss in inhibitory efficacy when zinc is added hours after the inducing antibiotic means that zinc might not be very effective as a “rescue” treatment for patients who have already received antibiotics for STEC. Instead, it appears that zinc would have to be given before or, at the latest, simultaneously with an antibiotic if the Stx induction phenomenon is to be prevented.

Figure Figure55 shows the effect of zinc in vivo against two different rabbit STEC strains in the ligated rabbit intestinal loop model. Rabbit ileal loops were infected with either the Stx1-producing strain RDEC-H19A or the Stx2-producing strain E22-stx2, with or without zinc. Figure Figure5A5A shows that zinc inhibited fluid secretion into the loop, measured as the volume-to-length ratio. For strain E22-stx2, zinc's inhibitory effect on secretion reached statistical significance (P < 0.05), while for RDEC-H19A, this reduction did not reach significance (P = 0.1).

FIG. 5.
Effect of zinc in vivo on STEC infection in ligated rabbit ileal loops. Loops of ileum (10 cm) were created as described in Materials and Methods and infected with 109 CFU of RDEC-H19A (Stx1+) or 108 CFU of E22-stx2 (Stx2+). Infection ...

Another measure of EPEC- and STEC-mediated damage to the intestinal mucosa is the villus-to-crypt ratio. EPEC and STEC cause cell damage and death to villus enterocytes, resulting in villus blunting, while cells in the crypts compensate by proliferation. Figure Figure5B5B shows that zinc strongly protected the rabbit intestines from the villus blunting induced by infection with both of the rabbit STEC strains, and Fig. 5H and I show zinc's protective effect against E22-stx2-induced blunting.

The percentage of the mucosal surface with adherent mats of pathogenic E. coli is yet a third quantitative measure of infection. Zinc significantly reduced the percent enteroadherence for strain E22-stx2, which adheres avidly (Fig. 5C, F, and G). For strain RDEC-H19A, which adheres less avidly, zinc's inhibitory effect on adherence was not statistically significant (Fig. (Fig.5C5C).

In addition to the parameters mentioned, zinc inhibited the amount of Stx that accumulated in the loops infected with rabbit STEC (Fig. (Fig.5D).5D). In contrast to the inhibitory effects on fluid secretion, adherence, villus blunting, and Stx production, zinc did not reduce the number of STEC bacteria recovered from the loops. For example, the number of E22-stx2 bacteria recovered from rabbit intestinal loops treated with 1 mM zinc acetate was 1.2 × 1012 CFU/ml, compared to 1.0 × 1012 CFU/ml recovered from the no-zinc control loops. Likewise, there was no difference in the numbers of RDEC-H19A bacteria recovered from the zinc-treated and control loops (P = 0.29). As with EPEC, the beneficial effects of zinc in vivo appear to be mediated via an inhibition of virulence rather than via an inhibition of bacterial growth. Lack of inhibition of bacterial growth has been cited as a possible advantage for other virulence inhibitors, such as quorum-sensing inhibitors (19), because of a theoretical lower chance of emergence of resistance.

Figure Figure5E5E shows massive submucosal edema in loops infected with E22-stx2, an effect also observed with strain RDEC-H19A (data not shown). Since submucosal edema is not observed with the parental REPEC strains (E22 and RDEC-1), the edema appears to be an effect of Stx. Figure Figure5F5F shows that infection with E22-stx2 causes villus blunting. Figure Figure5G,5G, at higher power (×600), shows E22-stx2 adhering in thick mats or biofilms (arrow) (E. coli bacteria stain blue with hematoxylin and eosin [H&E] stain). In contrast, loops infected with E22-stx2 in the presence of 1 mM zinc acetate (Fig. 5H and I) show histology that is difficult to distinguish from that of normal, uninfected ileum (data not shown), with normal slender villi and almost no adherent pathogenic E. coli.

In our previous work with zinc and EPEC and in the present study of zinc and STEC, we observed that fairly high concentrations of zinc, in the 0.1 to 0.5 mM range, were required to inhibit virulence. We were unsure if such high concentrations of zinc were achievable in the lumen of the gastrointestinal tract. To determine what concentrations of zinc were produced in response to oral zinc, we measured zinc in various segments of the rabbit gastrointestinal (GI) tract after orogastric administration of three different zinc salts: zinc acetate, zinc sulfate, and zinc oxide.

As shown in Fig. Fig.6A,6A, the 10-mg dose of zinc resulted in levels of zinc in the intestinal tract that were significantly higher than those in the control, non-zinc-treated rabbits. Zinc levels that we measured in intestinal and cecal lumen were in the 0.3 to 0.4 mM range for both of the more bioavailable zinc salts (acetate and sulfate), ranges that we have shown are inhibitory for both EPEC and STEC virulence. In the fecal pellets, zinc levels were even higher, reaching 0.5 to 0.6 mM (Fig. (Fig.6B);6B); higher levels are expected due to reabsorption of water from intestinal contents during passage through the colon. Furthermore, increased levels of zinc were observed in the fecal pellets after just 1 day of administration, showing that zinc supplements would not have to be given for a prolonged period to achieve protective levels in the distal GI tract. In contrast to the large, ~3-fold increase in zinc levels in the intestinal and fecal contents, zinc levels in rabbit plasma increased only 28% in the zinc acetate- and zinc sulfate-supplemented groups, compared to the level in control rabbits. This finding is in accord with the nutrition literature (15) in that the ability of the upper GI tract to absorb zinc is limited, like its ability to absorb iron. High doses of oral zinc can apparently exceed the absorptive ability of the upper GI tract and thereby reach therapeutic concentrations in the lower GI tract, where EPEC and STEC adhere and cause disease. The doses of zinc required for the latter purpose well exceed the amounts of zinc needed to maintain normal zinc nutritional status, so that high-dose zinc should be viewed as acting more like a drug and less as a nutrient in this context.

FIG. 6.
Concentrations of zinc achieved in intestinal lumenal contents and rabbit fecal pellets after orogastric administration of zinc at 10 mg elemental zinc daily. (A) Concentrations of zinc in intestinal lumenal contents after 3 days of daily zinc administration, ...


Our previous work showed that zinc blocked the production of virulence factors in EPEC, with strong inhibition by zinc of the production of the bundle-forming pilus, several EPEC secreted proteins (Esps), intimin, and Tir. Since EPEC is considered the evolutionary progenitor of most STEC strains (7, 30), we wondered whether zinc's inhibitory effects extended to STEC virulence factors as well.

As in EPEC, zinc inhibited expression of the Esps in STEC (Fig. (Fig.2),2), which is not surprising considering the similarity in the operons of the locus for enterocyte effacement (LEE) between EPEC and STEC, including the regulation of the LEE4 operon, which encodes the Esps. In our previous work, we noticed that the inhibitory effects of zinc were abolished in an isogenic strain with a deletion mutation in ler, the LEE-encoded regulator. Here we show again that in the Δler mutant of STEC strain EDL933 and in the Δler mutant of rabbit EPEC E22 zinc's inhibitory effects on espA and espB were lost (Fig. 2A to D). Therefore, in human EPEC, rabbit EPEC, and STEC O157:H7, Ler is needed to observe the inhibitory effects of zinc on the EPEC and EHEC secreted proteins (Esps). Further work is needed to determine the molecular basis of this interesting observation.

Production of the Shiga toxins is the main characteristic which distinguishes STEC from EPEC. The Shiga toxins are encoded on lysogenic bacteriophages within the STEC chromosome, and regulation of Stx production is quite different from that of other STEC virulence factors. For example, Stx production is independent of Ler (9), quorum sensing (10), bicarbonate, and other environmental signals and transcription factors which regulate virulence gene expression in EPEC and STEC. Although Stx production is regulated differently than that of the Esps, zinc nevertheless inhibited Stx production (Fig. (Fig.33 and and4)4) and even inhibited Stx in the Δler mutant (Fig. (Fig.2F).2F). Therefore, although the principle of parsimony makes it desirable to postulate a single, unifying target for zinc action, our current data force us to hypothesize that there are at least two independent targets for zinc in STEC: one which involves Ler, and a separate Ler-independent target that mediates inhibition of Stx.

Zinc's inhibitory effects on Stx were observed in strains expressing Stx1 only, Stx2 only, and both toxins. Zinc inhibited basal as well as antibiotic-induced Stx expression, and its effects were measurable at the level of toxin protein as well as RNA expression. Zinc-mediated inhibition of Stx generally achieved or exceeded 90% inhibition of Stx at 0.3 to 0.5 mM zinc. In addition to inhibition of ciprofloxacin-induced Stx production, zinc inhibited Stx induced by trimethoprim and by H2O2.

Zinc and other divalent metals have been noted to be able to inhibit protein and RNA synthesis at concentrations similar to those used in our study (0.1 to 0.3 mM) (3, 8). There appears to be some general ability of divalent metals (zinc, nickel, copper, cobalt, etc.) to inhibit E. coli growth and protein synthesis, but among this group of metals, zinc appears to stand out as the most potent inhibitor of EPEC and STEC virulence factor expression (Fig. (Fig.3A3A and and4B)4B) (5).

In non-surgically-altered rabbits, zinc in GI tract contents achieved levels in the range (0.3 to 0.5 mM) that our in vitro experiments showed were sufficient to inhibit adherence, Esp expression, and Stx production. The data presented in Fig. Fig.6,6, however, were obtained in rabbits which were not infected with EPEC or STEC. The zinc levels achieved in the presence of active infection could be different from those we measured in uninfected rabbits. EPEC is known to induce a malabsorptive state in humans and animals (20, 25), so it is possible that zinc absorption would be reduced during infection and that larger amounts of zinc would therefore reach the ileum, cecum, and colon. On the other hand, zinc concentrations could be reduced by the dilutional effect of diarrheal fluid secretion. Determining the concentrations of zinc achievable in the intestinal lumen in the presence of active infection is therefore a goal for future research.

For many years, there has been a strong presumption in the nutrition literature that the beneficial effects of zinc supplementation in infectious diseases, including infectious diarrhea, were due to correction of a preexisting zinc deficiency. This view has unfortunately persisted despite the fact that field studies of zinc have shown benefits even in subsets of children with normal plasma zinc levels (21). That “zinc deficiency” view has led to the assumption that zinc supplements would offer no benefit in diarrheal illness in previously well-nourished subjects. The present study, our previous work (5), and the work of others (13) strengthen our belief that zinc, by acting directly against pathogen virulence, could be therapeutic even in humans and animals with completely normal zinc nutritional status. The doses of zinc required for this protective effect are supraphysiologic, meaning that zinc acts in a drug-like fashion. Furthermore, zinc can act within the lumen of the gastrointestinal tract, meaning that its antimicrobial effects do not necessarily require systemic absorption. To gain acceptance and to be translated into practical benefit, these new concepts require a very small but important paradigm shift in understanding zinc's actions in acute infectious diarrheal disease.


We acknowledge the support of the National Institutes of Health via grants R21 AI 066055 and RO1 AI081528.


Editor: S. M. Payne


[down-pointing small open triangle]Published ahead of print on 18 January 2011.


1. Bhandari, N., et al. 2002. Substantial reduction in severe diarrheal morbidity by daily zinc supplementation in young north Indian children. Pediatrics 109:e86. [PubMed]
2. Bhatnagar, S. 2007. Effects of zinc supplementation on child mortality. Lancet 369:885-886. [PubMed]
3. Blundell, M., and D. Wild. 1969. Inhibition of bacterial growth by metal salts: a survey of the effects on the synthesis of ribonucleic acid and protein. Biochem. J. 115:207-212. [PMC free article] [PubMed]
4. Crane, J., S. Choudhari, T. Naeher, and M. Duffey. 2006. Mutual enhancement of virulence by enterotoxigenic and enteropathogenic Escherichia coli. Infect. Immun. 74:1505-1515. [PMC free article] [PubMed]
5. Crane, J., T. Naeher, I. Shulgina, C. Zhu, and E. Boedeker. 2007. Effect of zinc in enteropathogenic Escherichia coli infection. Infect. Immun. 75:5974-5984. [PMC free article] [PubMed]
6. Crane, J. K., and I. Shulgina. 2009. Feedback effects of host-derived adenosine on enteropathogenic Escherichia coli. FEMS Immunol. Med. Microbiol. 57:214-228. [PMC free article] [PubMed]
7. Donnenberg, M. S., and T. S. Whittam. 2001. Pathogenesis and evolution of virulence in enteropathogenic and enterohemorrhagic Escherichia coli. J. Clin. Invest. 107:539. [PMC free article] [PubMed]
8. Easton, J., P. Thompson, and M. Crowder. 2006. Time-dependent translational response of E. coli to excess Zn(II). J. Biomol. Tech. 17:303-307. [PMC free article] [PubMed]
9. Elliott, S., et al. 2000. The locus of enterocyte effacement (LEE)-encoded regulator controls expression of both LEE- and non-LEE-encoded virulence factors in enteropathogenic and enterohemorrhagic Escherichia coli. Infect. Immun. 68:6115-6126. [PMC free article] [PubMed]
10. Hughes, D. T., and V. Sperandio. 2008. Inter-kingdom signalling: communication between bacteria and their hosts. Nat. Rev. Microbiol. 6:111. [PMC free article] [PubMed]
11. Jinneman, K. C., K. J. Yoshitomi, and S. D. Weagant. 2003. Multiplex real-time PCR method to identify Shiga toxin genes stx1 and stx2 and Escherichia coli O157:H7/H serotype. Appl. Environ. Microbiol. 69:6327-6333. [PMC free article] [PubMed]
12. Kaneko, Y., M. Thoendel, O. Olakanmi, B. Britigan, and P. Singh. 2007. The transition metal gallium disrupts Pseudomonas aeruginosa iron metabolism and has antimicrobial and antibiofilm activity. J. Clin. Invest. 117:877-888. [PMC free article] [PubMed]
13. Kelleher, S., I. Casa, N. Carbajal, and B. Lonnerdal. 2002. Supplementation of infant formula with the probiotic Lactobacillus reuteri and zinc: impact on enteric infection and nutrition in infant rhesus monkeys. J. Pediatr. Gastroenterol. Nutr. 35:162-168. [PubMed]
14. Leverton, L. Q., and J. B. Kaper. 2005. Temporal expression of enteropathogenic Escherichia coli virulence genes in an in vitro model of infection. Infect. Immun. 73:1034-1043. [PMC free article] [PubMed]
15. Lichten, L. A., and R. J. Cousins. 2009. Mammalian zinc transporters: nutritional and physiologic regulation. Annu. Rev. Nutr. 29:153-176. [PubMed]
16. Mahajan, A., et al. 2009. An investigation of the expression and adhesin function of H7 flagella in the interaction of Escherichia coli O157:H7 with bovine intestinal epithelium. Cell. Microbiol. 11:121-137. [PubMed]
17. Perrin, C., et al. 2009. Nickel promotes biofilm formation of curli-proficient K-12 Escherichia coli. Appl. Environ. Microbiol. 75:1723-1733. [PMC free article] [PubMed]
18. Rashid, R., et al. 2006. Expression of putative virulence factors of Escherichia coli O157:H7 differs in bovine and human infections. Infect. Immun. 74:4142-4148. [PMC free article] [PubMed]
19. Rasko, D. A., et al. 2008. Targeting QseC signaling and virulence for antibiotic development. Science 321:1078-1080. [PMC free article] [PubMed]
20. Rothbaum, R., A. McAdams, R. Giannella, and J. Partin. 1982. A clinicopathologic study of enterocyte-adherent Escherichia coli: a cause of protracted diarrhea in infants. Gastroenterology 83:441-454. [PubMed]
21. Sazawal, S., et al. 1995. Zinc supplementation in young children with acute diarrhea in India. N. Engl. J. Med. 333:839-844. [PubMed]
22. Shaw, R. K., et al. 2008. Enterohemorrhagic Escherichia coli exploit EspA filaments for attachment to salad leaves. Appl. Environ. Microbiol. 74:2908-2914. [PMC free article] [PubMed]
23. Sinclair, J., and A. O'Brien. 2004. Intimin types alpha, beta, and gamma bind to nucleolin with equivalent affinity but lower avidity than to the translocated intimin receptor. J. Biol. Chem. 279:33751-33758. [PubMed]
24. Strand, T. A., et al. 2002. Effectiveness and efficacy of zinc for the treatment of acute diarrhea in young children. Pediatrics 109:898-903. [PubMed]
25. Tai, Y.-H., T. Gage, C. McQueen, S. Formal, and E. Boedeker. 1989. Electrolyte transport in rabbit cecum. I. Effect of RDEC-1 infection. Am. J. Physiol. 256:G721-G726. [PubMed]
26. Torres, A. G., and J. B. Kaper. 2003. Multiple elements controlling adherence of enterohemorrhagic Escherichia coli O157:H7 to HeLa cells. Infect. Immun. 71:4985-4995. [PMC free article] [PubMed]
27. Torres, A. G., et al. 2007. Ler and H-NS, regulators controlling expression of the long polar fimbriae of Escherichia coli O157:H7. J. Bacteriol. 189:5916-5928. [PMC free article] [PubMed]
28. Wagner, P. L., D. W. K. Acheson, and M. K. Waldor. 2001. Human neutrophils and their products induce Shiga toxin production by enterohemorrhagic Escherichia coli. Infect. Immun. 69:1934-1937. [PMC free article] [PubMed]
29. Wapnir, R. A. 2000. Zinc deficiency, malnutrition and the gastrointestinal tract. J. Nutr. 130:1388S-1392S. [PubMed]
30. Wick, L. M., W. Qi, D. W. Lacher, and T. S. Whittam. 2005. Evolution of genomic content in the stepwise emergence of Escherichia coli O157:H7. J. Bacteriol. 187:1783-1791. [PMC free article] [PubMed]
31. Yoh, M., E. Frimpong, and T. Honda. 1997. Effect of antimicrobial agents, especially fosfomycin, on the production and release of Vero toxin by enterohaemorrhagic Escherichia coli O157:H7. FEMS Immunol. Med. Microbiol. 19:57-64. [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...