OUP user menu

Survival characteristics of diarrheagenic Escherichia coli pathotypes and Helicobacter pylori during passage through the free-living ciliate, Tetrahymena sp.

Charlotte D. Smith, Sharon G. Berk, Maria T. Brandl, Lee W. Riley
DOI: http://dx.doi.org/10.1111/j.1574-6941.2012.01428.x 574-583 First published online: 1 December 2012


Free-living protozoa have been implicated in the survival and transport of pathogens in the environment, but the relationship between non-Shiga toxin-producing Escherichia coli or Helicobacter pylori and ciliates has not been characterized. Six diarrheagenic pathotypes of E. coli and an isolate of H. pylori were evaluated for their susceptibility to digestion by Tetrahymena, an aquatic ciliate. Tetrahymena strain MB125 was fed E. coli or H. pylori, and the ciliate's egested products examined for viable bacterial pathogens by the BacLight LIVE/DEAD assay, a cell elongation method, and by colony counts. All six diarrheagenic E. coli pathotypes survived digestion, whereas H. pylori was digested. Growth of E. coli on agar plates indicated that the bacteria were able to replicate after passage through the ciliate. Transmission electron micrographs of E. coli cells as intact rods vs. degraded H. pylori cells corroborated these results. Scanning electron microscopy revealed a net-like matrix around intact E. coli cells in fecal pellets. These results suggest a possible role for Tetrahymena and its egested fecal pellets in the dissemination of diarrheagenic E. coli in the environment. This bacterial–protozoan interaction may increase opportunities for transmission of diarrheagenic E. coli to mammalian hosts including humans.

  • protozoa
  • vesicle
  • fecal pellets
  • electron microscopy
  • confocal laser scanning microscopy
  • phagosome


Escherichia coli is an important cause of diarrhea and death among children, especially in developing countries (Hunter, 2003). Although the sources of some E. coli serotypes, such as O157:H7, are well established, the sources of other diarrheagenic E. coli pathotypes, such as diffusely adherent E. coli (DAEC), enteroaggregative E. coli (EAEC), enteroinvasive E. coli (EIEC), enteropathogenic E. coli (EPEC), and enterotoxigenic E. coli (ETEC), remain obscure. The high frequency of these infections (especially EAEC and ETEC) in children in developing countries and among travelers to these countries suggests environmental sources, including water and food. Helicobacter pylori DNA has been found in potable water systems (Watson et al., 2004), but the significance of drinking water as a route of exposure has not been defined.

Epidemiologic and environmental studies of the occurrence and control of enteric bacteria in water systems often overlook the fact that in the natural environment, these bacterial pathogens may exist in the same niches as free-living protozoa, including ciliates such as Tetrahymena. Several bacterial pathogens including E. coli O157:H7 (Nelson et al., 2003; Steinberg & Levin, 2007; Gourabathini et al., 2008), Legionella pneumophila (Berk et al., 2008; Faulkner et al., 2008), Salmonella enterica (Brandl et al., 2005), Francisella tularensis (Kormilitsyna et al., 1993; Thelaus et al., 2008), Campylobacter jejuni (Snelling et al., 2005), and several E. coli laboratory strains (Schlimme et al., 1997; Ghafari et al., 2008; Matsuo et al., 2010) have been shown to evade digestion in Tetrahymena phagosomes. However, not all bacterial pathogens can evade digestion even at the high bacteria/ciliate ratios used to induce ciliate feeding. For example, Listeria monocytogenes did not survive digestion in the Tetrahymena species used in our study (Brandl et al., 2005).

The aim of this study was to determine whether the six major pathotypes of diarrheagenic E. coli or an isolate of H. pylori is able to survive digestion by the ubiquitous ciliate, Tetrahymena. Knowing whether these bacteria are simply prey to Tetrahymena or whether the bacteria can survive the digestive processes and emerge sequestered in pellets can enhance the knowledge of the microbial ecology of these important human pathogens. Additionally, this work can provide a foundation for future studies on the control of these pathogens during water and wastewater treatment (e.g. filtration, chemical or ultraviolet light disinfection methods), thus improving the microbial quality of drinking water or water discharged to the environment.

Materials and methods

Microbial strains and growth conditions

Six diarrheagenic E. coli pathotypes from the Riley Laboratory Collection were evaluated in this study (Table 1). All strains were characterized for membership in each group by the Centers for Disease Control and Prevention and confirmed by PCR-based tests for the presence of genes that define their pathogenic groups (Tornieporth et al., 1995). All six pathotypes have been maintained in this collection for several decades including minimally passaged E. coli O157:H7 from the 1982 outbreak in Oregon, in which this strain was identified as a human pathogen (Riley et al., 1983). Nonpathogenic E. coli K12 (ATCC 10798) and 1 μm green fluorescent latex beads (Sigma, St. Louis, MO) were used for comparative purposes under the same conditions (i.e. concentrations, ratios, incubation temperature).

View this table:

Pathogenic Escherichia coli strains used in this study

Enteropathogenic E. coli (EPEC)O126:NM
Enteropathogenic E. coli (EPEC)O111:NM
Enterohemorrhagic E. coli (EHEC)O157:H7
Enterotoxigenic E. coli (ETEC)E2539
Enterotoxigenic E. coli (ETEC)1493
Enteroinvasive E. coli (EIEC)O144:H25
Enteroaggregative E. coli (EAEC)3–8
Diffusely adherent E. coli (DAEC)O8AD
Nonpathogenic E. coliK12

The E. coli strains were maintained on tryptic soy agar (TSA) slants [Becton-Dickson Difco (BD Difco), Sparks, MD] and grown overnight in Luria-Bertani broth (BD Difco) to an optical density of 600 nm (OD600) of 1, [c. 1 × 109 colony forming units (CFU) mL−1] for experimentation. Broth cultures were centrifuged at 5000 g for 5 min and then washed in municipal tap water that had been autoclaved for sterilization. Boiling in the autoclave also removed the chlorine residual of the tap water (Krasner & Wright, 2005). The absence of chlorine was confirmed by testing with N,N diethyl-p-phenylene diamine (Hach, Loveland, CO). Dechlorinated sterile tap water was selected because its source was a large pristine lake that would represent the natural environment of the protozoa better than synthetic media. Nevertheless, with regard to feeding, swimming, and forming fecal pellets, the ciliates in our study behaved as they did in other studies that used a buffered saline medium (Brandl et al., 2005; Power et al., 2006; Pinheiro et al., 2007; Faulkner et al., 2008).

Clinical isolates of H. pylori were grown to confluence on TSA with 5% sheep's blood (Hardy Diagnostics, Santa Maria, CA) for 48 h at 37 °C in airtight boxes with AnaeroPack sachets (Mitsubishi Gas Corp., Tokyo, Japan). Helicobacter pylori colonies were harvested from the agar plates for co-culture with Tetrahymena by gently washing the surface of the agar media with 2 mL of phosphate-buffered saline, pH 7.2. To check the condition of H. pylori, we applied a 2-μL aliquot of the remaining H. pylori solution to a copper grid and negatively stained it with 2 μL of 2% uranyl acetate. The negatively stained H. pylori were then examined by transmission electron microscopy (TEM; JEOL 100 CX electron microscope). We also tested viability by applying a 20-μL drop of the harvested H. pylori cells to a fresh TSA with 5% sheep's blood plate and incubated as described earlier.

Tetrahymena sp. strain MB125 (Brandl et al., 2005) was maintained axenically at 25 °C by transferring 500 μL of an existing culture into 5 mL of 2/3 strength plate count broth (0.66% tryptone, 0.33% yeast extract, 0.2% dextrose; BD Difco) in a 25-mL tissue culture flask (BD, Franklin Lakes, NJ), approximately once per week. Forty-eight hours before the Tetrahymena were to be used for co-culture with E. coli, a fresh Tetrahymena subculture was prepared in the same manner. The Tetrahymena cells were harvested by centrifugation (200 g) and resuspended in chlorine-free tap water. Fifty percent and then 75% of the plate count broth was replaced by tap water in two sequential steps to avoid osmotic shock to the ciliates. Therefore, the final volume was 0.6 mL of 2/3 strength plate count broth and 4.4 mL of tap water.


Co-cultures were made by mixing 100 μL of harvested E. coli culture (c. 1 × 108 CFU mL−1) with Tetrahymena (c. 1 × 104 cells) in a final volume of 1 mL of tap water in triplicate wells of 24-well tissue culture plates. Similarly, 100 μL of harvested H. pylori cells were mixed with c. 1 × 104 Tetrahymena for co-culture. We performed serial dilutions of the harvested E. coli and counted CFU mL−1 on LB agar to check the concentrations used in the co-cultures. Helicobacter pylori cells were stained with BacLight LIVE/DEAD assay (Molecular Probes, Carlsbad, CA) and examined under a 63× objective on a Zeiss AxioObserver epifluorescent microscope fitted with a Zeiss 38HE filter cube (excitation: BP470/40; beamsplitter: FT495; emission: BP525/50) to ensure that the ratio of H. pylori to Tetrahymena was at least as high as was used for the E. coli co-cultures. For each co-culture, three identical replicates were made in individual wells of a 24-well tissue culture plate. Separate 24-well plates were used for each E. coli strain.

Cell morphology and ultrastructure

To examine E. coli and Tetrahymena cell morphology by scanning electron microscopy (SEM), we processed 25-μL aliquots from 24-h co-cultures (prepared as described earlier). To examine the ultrastructure of planktonic cells and ingested bacterial cells or bacterial cells in egested fecal pellets by TEM, we processed entire 1-mL E. coli–Tetrahymena or H. pylori–Tetrahymena co-cultures (prepared as described earlier) after 1 and 24 h to see intraciliate and egested bacteria, respectively. Processing for electron microscopy was carried out on duplicate co-cultures, and the entire process (from establishing a co-culture to imaging by SEM or TEM) was repeated on at least three occasions, except for TEM imaging of negatively stained H. pylori, which was carried out once (on 10 fields of view).

Samples to be examined by SEM were fixed in 2.0% glutaraldehyde in 0.1 M cacodylate buffer at pH 7.2. Fixed samples were dispersed onto poly-l-lysine-treated silicon wafers and rinsed twice with cacodylate buffer, postfixed in 1% osmium tetroxide in 0.1 M cacodylate buffer at pH 7.2, dehydrated with a graded series of ethanol (30–100%), then dehydrated in a critical point dryer, and sputter-coated with goldpalladium to a coating thickness of about 2 μm. Specimens were imaged with a Hitachi S-5000 electron microscope.

Samples to be examined by TEM were fixed in 2.0% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2), rinsed with cacodylate buffer, and postfixed in 1% osmium tetroxide containing 1.6% potassium ferricyanide. Following cacodylate and deionized distilled water washes, samples were stained with 0.5% uranyl acetate overnight at 4 °C, dehydrated with a graded series of acetone (30–100%) rinses, and embedded in Eponate resin. Thin sections were mounted on Formvar-coated copper grids and stained with methanolic uranyl acetate for 7 min, rinsed with 70% methanol and then with distilled water, stained with freshly prepared lead citrate for 5 min, and rinsed again with distilled water. Sections were imaged with a JEOL 100 CX electron microscope operated at an accelerating voltage of 80 kV.

Escherichia coli viability

Three methods were used to determine whether the E. coli pathotypes were able to survive digestion by the protozoa. Two of these used direct viable count, and one used culture on agar. For the first set of experiments, we examined E. coli both within Tetrahymena vacuoles and in egested fecal pellets with the BacLight LIVE/DEAD assay. This assay is based on the permeability of the DNA intercalating dye, Syto9, through cell membranes, and its displacement by propidium iodide, which penetrates disrupted cell membranes, typical of dead cells. We prepared a ‘BacLight stock solution’ by adding 1.5 μL of 3.34 mM Syto9 and 1.5 μL of 20 mM propidium iodide to 997 μL of nuclease-free water. Three microliters of the BacLight stock solution was added to 10 μL-aliquots of each of the E. coli–Tetrahymena co-cultures on microscope slides (from three separate wells of tissue culture plates). The stained co-culture remained in the dark for 15 min at room temperature to allow the stain to intercalate within the DNA of E. coli in Tetrahymena cells or in egested fecal pellets. Then 5 μL of 2.5% glutaraldehyde was added to fix the cells. Without the fixative, Tetrahymena cells would burst under laser microscopy. Coverslips were gently placed on the microscope slides to avoid disruption of Tetrahymena or fecal pellets. The coverslips were sealed with quick-dry clear nail polish to avoid dehydration and disruption of the ciliates.

The second test of bacterial cell viability combined 30 μL of 24-h E. coli–Tetrahymena co-culture with 30 μL of LB broth and 30 μL of 64 μM nalidixic acid, a DNA gyrase inhibitor that allows all metabolic processes to proceed except cell division, producing elongated cells (Kogure et al., 1979). This mixture was created once for each of three replicate E. coli–Tetrahymena co-cultures from 24-well plates as described earlier and placed in three wells of a 96-well plate. The plate was then incubated at 37 °C for approximately 2 h, but observed about every 20 min for cell elongation under a 40× objective of an inverted Olympus IX80 bright field microscope. When elongated bacteria were seen jetting out of pellets, 10-μL aliquots from each of the three replicate wells of the 96-well plate were placed on microscope slides, stained with BacLight, fixed as mentioned earlier, and then examined with confocal laser scanning microscopy (CLSM) as follows:

Slides from E. coli viability (LIVE/DEAD staining) and elongation experiments were imaged at the UC Berkeley Molecular Imaging Center using the inverted laser scanning confocal microscopes (Zeiss LSM 510 NLO Axiovert or Zeiss LSM 710 AxioObserver) and Zeiss AIM 3.5 or Zeiss Zen 2010 software, respectively. Syto9 and propidium iodide were imaged sequentially, using a 488-nm laser and a 543- or 561-nm laser, with either a Zeiss Plan-Apochromat 63×/1.4 NA oil objective or Zeiss Plan-Apochromat 100×/1.4 NA oil objective, with a pinhole of 1 airy unit. The bright field image was captured simultaneous to the Syto9 acquisition using a transmitted light photomultiplier tube. Z-stacks were acquired using Nyquist sampling, with a 0.5-nm step size. Postprocessing to alter brightness and contrast was performed in the Zeiss LSM Image Browser software to allow for better visual inspection. No gamma adjustment was performed. Manual counts were performed by looking through Z-stacks to identify live (Syto9, green) and dead (propidium iodide, red) E. coli.

Quantitation of E. coli after passage through Tetrahymena

To verify that E. coli in egested fecal pellets were capable of replication, Tetrahymena were allowed to feed for 2 weeks, allowing ample time for planktonic bacteria to be packaged into fecal pellets that subsequently aggregated into flocs. A decrease in planktonic bacteria in the co-cultures over time was periodically assessed by observation of the tissue culture plates' wells under the 40× objective on the inverted Olympus IX80 microscope. After 2 weeks, we extracted six, 2-μL aliquots from a single floc-containing well as follows: One set of three aliquots was extracted by placing a pipette tip directly into the floc (that was visible by eye), and a second set of three aliquots was extracted from the same well at a position away from the visible floc. Four- to sevenfold serial dilutions from each aliquot were plated on LB agar and incubated at 37 °C overnight for enumeration of CFU mL−1 for bacteria in the floc vs. bacteria not associated with floc.

Tetrahymena enumeration

Approximately 1 × 103 Tetrahymena cells were fed c. 1 × 108 CFU mL−1 of six strains of E. coli (the nonpathogenic strain K12; two EPEC strains: O111:NM and 126:NM; two ETEC strains: E2539 and 1493; and the DAEC strain: O8AD) in a final volume of 1-mL tap water. For comparative purposes, a monoculture of washed Tetrahymena was prepared without the addition of E. coli. Each co-culture and the Tetrahymena monoculture were prepared in duplicate and incubated at 25 °C. Immediately and after 24 and 48 h, 10-μL aliquots were withdrawn from the replicate wells, placed on microscope slides, and fixed with 5 μL of 2.5% glutaraldehyde. After coverslips were gently placed on the aliquots and sealed with clear nail polish, the Tetrahymena cells were enumerated under the 10× objective of the inverted Olympus IX80 microscope. Duplicate counts were averaged and multiplied by 100 to obtain the number of Tetrahymena cells mL−1 at the three time periods.


A two-tailed Student's t-test was used for comparison of CFUs arising from bacteria in visibly aggregated pellets with those from a portion of the co-culture without visible floc. Tetrahymena counts from cultures with and without E. coli were compared using a one-way analysis of variance with repeated measures (for the replicates). These analyses were conducted with stata version 11 (StataCorp, College Station, TX).


Bacterial ingestion

Morphologically distinct (green) Syto9-stained E. coli rods were apparent within Tetrahymena sp. MB125 vacuoles within an hour of feeding indicating live ETEC serotype E2539 (Fig. 1a). Identical images were produced from CLSM of the other diarrheagenic E. coli serotypes and E. coli K12. In contrast, vacuoles of Tetrahymena sp. MB125 that had been fed H. pylori contained some (green) Syto9 staining, but bacterial cells were not discernable (Fig. 1b) indicating that the H. pylori had been digested, and residual DNA was present, but that propidium iodide did not penetrate the vacuoles or was in insufficient concentration to be visualized.

BacLight-stained bacterial cells in Tetrahymena sp. MB125. (a) Syto9 (green)-stained Escherichia coli are alive and have maintained their rod shape, whereas distinct Helicobacter pylori (b) rods were not detectable and appeared to be undergoing digestion. Bars = 10 μm.

Cell morphology and ultrastructure

Egestion (defecation) of the vacuoles (vesicles) under bright field microscopy started at about 2 h and continued through the next day. Twenty-four-hour co-cultures examined by SEM revealed Tetrahymena cells of various stages of the bacterial ingestion and egestion processes. Figure 2a shows E. coli associated with the oral apparatus (cytopharynx) of Tetrahymena cells. After passage through the cytoplasm to the posterior end of the ciliate, E. coli were egested from the cytoproct in individual fecal pellets (Fig. 2b). Additionally, SEM revealed the presence of a net-like matrix around E. coli cells egested in fecal pellets (Fig. 3a). This was unlikely a bacterial product because a similar matrix was observed around latex beads fed to, and egested by, Tetrahymena (Fig. 3b). Observed differences in the texture of the matrix may be an artifact of the critical point drying step of the SEM protocol.

(a) Escherichia coliO157:H7 (EHEC) cells adjacent to the cytopharynx (mouth) of Tetrahymena sp. (MB125). (b) A fecal pellet being egested from the cytoproct. Bars = 1 μm (a), 2 μm (b).

SEM images of Escherichia coli-containing fecal pellets from Tetrahymena. (a) A net-like sac containing E. coliO157:H7 (EHEC). The bacteria were packaged into a fecal pellet (inset). (b) Latex beads were egested in a similar pellet (b,inset) with a casing like that of the pellets containing bacteria. Bars = 0.5 μm.

The rod-shaped morphology of H. pylori cells (Fig. 4) fed to Tetrahymena is an accepted characteristic of culturable H. pylori cells (Adams et al., 2003). We also noted that the 20-μL drop of harvested H. pylori cells applied to TSA with 5% sheep's blood plates exhibited growth, which further indicated that the H. pylori cells were viable when added to the co-culture.

TEM image of negatively stained Helicobacter pylori. The rod shape indicates that bacteria were viable when fed to Tetrahymena. Bar = 2 μm.

The cell morphology of E. coli in Tetrahymena phagosomes (Fig. 5a) and in fecal pellets (Fig. 5b) indicated intact rod-shaped cells. However, ‘whorls’ characteristic of digested cells (Berk et al., 2008) were observed in the phagosomes (Fig. 5c) and in fecal pellets (Fig. 5d) when Tetrahymena grazed on H. pylori. All six diarrheagenic pathotypes exhibited similar morphologies by TEM (ETEC strain E2539 is provided as the example in Fig. 5).

Transmission electron micrographs showing ETEC E2539 cells within food vacuoles of Tetrahymena sp. MB125 (a) and in egested fecal pellets (b), compared with whorls of digested Helicobacter pylori cells in food vacuoles (c) and in egested fecal pellets (d). Red circles highlight described ultrastructures. Bars = 5 μm (a, b, d), 0.5 μm (c).

Bacterial viability

Less than 1% of E. coli cells in egested fecal pellets were dead (propidium iodide-stained, red), as determined by CLSM observation of 10 fields of view per E. coli pathotype. For example, the flocs in the field of view represented by Fig. 6 had approximately 370 pellets averaging 5 μm in size and contained at least 25 live E. coli 0157:H7 cells per pellet for a total of almost 1000 pelleted bacterial cells. Fourteen dead E. coli cells were observed in pellets by examining the Z-stacks of this field of view. There were 142 planktonic E. coli identified in optical slices of the Z-stacks (to reveal bacteria not associated with a fecal pellet). These proportions were typical for 24-h co-cultures of all six diarrheagenic E. coli and the nonpathogenic E. coli K12. Tetrahymena fed with H. pylori produced detritus that did not contain distinct bacterial cells.

Syto9- (green) and propidium iodide- (red-) stained co-culture of DAEC serotype O8AD showing Tetrahymena sp. MB125 (T), Escherichia coli cells in an egested fecal pellet (P), and fecal pellets aggregated into floc (F) after 2 weeks of incubation. Syto9-stained bacterial cells are alive. All six pathotypes of diarrheagenic E. coli produced similar features. Bar = 5 μm

Nalidixic acid added as a DNA gyrase inhibitor prevented division of live cells, as evidenced by elongated bacterial cells in the pellets. BacLight staining improved visualization of the elongated cells and served as a confirmation of cell viability (Fig. 7). All six diarrheagenic pathotypes and K12 exhibited the same sensitivity to nalidixic acid as depicted in Fig. 7 where the EAEC strain 3–15 serves as the example.

Confocal micrograph of Escherichia coli in pellets exposed to the DNA gyrase inhibitor nalidixic acid, which allows replication but not separation of daughter cells. The cells were then stained with Syto9 and propidium iodide. Elongated green cells are alive. The EAEC strain 3–15 serves as the example. Bar = 5 μm.

Bacterial replication

Colonies arising from the floc indicated that E. coli remained viable in the pellets during a 2-week period and had the capacity for replication. For each of the six pathogenic E. coli strains and K12 fed to Tetrahymena for 2 weeks, aliquots taken from a floc of aggregated fecal pellets produced significantly more colonies on LB agar than aliquots taken from a part of the tissue culture plate well that had some nonaggregated pellets and planktonic bacteria (Fig. 8, P < 0.05). We used the term ‘pelleted’ to describe aliquots from visible floc and the term ‘planktonic’ to describe aliquots from the solution (that contained some loose pellets and planktonic bacteria). Evaluation of CLSM optical slices of Z-stacks indicated very few viable bacteria that were not associated with an intact or bursting pellet in the floc. That is, the floc was mostly comprised of pellets, and only a few planktonic bacteria were observed imbedded in the floc. Planktonic bacteria may have moved away from the floc facilitated by their own motility or agitation caused by the swimming activity of the ciliates. We also observed Tetrahymena picking at the floc, like fish on a coral reef, presumably to ingest available planktonic bacteria.

Bacterial concentration arising from 2-μL aliquots of aggregated pellets (floc) (dark bars) and of bacteria not associated with floc (light bars) from co-cultures of Tetrahymena sp. MB125 with various pathotypes of diarrheagenic Escherichia coli or nonpathogenic E. coliK12 incubated for 2 weeks. Bacteria from aggregated pellets produced significantly more colonies on LB agar plates than planktonic cell suspensions from the same co-cultures. Each bar represents the mean of three replicate co-cultures. Error bars are standard deviations. Two-tailed Student's t-test: *P< 0.05, **P< 0.01, ***P< 0.001.

Tetrahymena replication

Although the power of this analysis was low, there was no evidence to reject the hypothesis that counts of Tetrahymena in a control monoculture (without E. coli) were similar to counts when fed various E. coli pathotypes, or different strains within the same pathotype (P> 0.05). The ciliates did not appear to obtain adequate nutrients from E. coli to increase their concentrations significantly above the control (Fig. 9). Bars represent averages of duplicate aliquots from the co-cultures or monoculture, and lines represent standard deviations.

Growth of Tetrahymena in a tap water monoculture (no Escherichia coli) did not significantly differ from growth when fed diarrheagenic E. coli pathotypes or different of strains within the same pathotype. Bars are means of two replicates. Lines are standard deviations.


Our study demonstrated that representative strains of six diarrheagenic E. coli pathotypes were able to evade digestion from an environmental isolate of Tetrahymena (MB125). Unlike prior studies that focused on survival of E. coli inside Tetrahymena cells (Nilsson, 1987; Schlimme et al., 1995; Steinberg & Levin, 2007; Matsuo et al., 2010; Oguri et al., 2011), we reported on the bacterial survival in fecal pellets, which were expelled starting approximately 2 h after the bacterial prey was ingested. Fecal pellets adhered together to form floc, the components of which were not characterized in this study, but may be similar to ciliate-associated floc comprised primarily of polysaccharides described by others (Arregui et al., 2007).

The short time of bacterial residence within the ciliate was consistent with previous reports of E. coli residing in Tetrahymena for 2–4 h (Nilsson, 1977). However, as this residence time is short, the survival of E. coli in fecal pellets over long periods of time may have a greater impact on its ecology. While survival of E. coli O157:H7 ingested by Tetrahymena had been previously reported (Nelson et al., 2003; Steinberg & Levin, 2007; Gourabathini et al., 2008), it was unknown whether this trait is shared among all diarrheagenic pathotypes of E. coli. Our results from microscopic observations and viability studies indicate that resistance to digestion by Tetrahymena is a common trait among E. coli strains, as all six pathotypes tested were egested as viable cells in fecal pellets that gave rise to colonies on LB agar. Growth on agar was interpreted as the ability to replicate under appropriate conditions. The number of colonies arising from plated aliquots of floc under-represents the actual number of E. coli bacteria in a sample of floc, because competitive inhibition of colony growth would occur without disaggregation of a pellet to individual bacterial cells before plating. Attempts to disaggregate the fecal pellets by filtration or physical disruption through hypodermic needles did not produce reproducible results. Therefore, the colonies arising from plated floc should be interpreted relative to colonies arising from solution as presented earlier for comparison rather than actual values.

Unlike E. coli, the enteric pathogen H. pylori, reported to survive digestion by the free-living protozoan Acanthamoeba castellanii (Winiecka-Krusnell et al., 2002), was digested by Tetrahymena sp. MB125. Our results provide the first evidence that H. pylori cannot evade digestion by this ciliate. These results also support previous findings that certain enteric pathogens interact with Tetrahymena differently than E. coli and S. enterica and that bacterial survival in protozoa may depend on the protozoan species (Brandl et al., 2005).

Recent evidence suggests the enhancement of bacterial virulence factors, acid resistance (Rehfuss et al., 2011), and enhanced horizontal gene transfer (Schlimme et al., 1997; Matsuo et al., 2010; Oguri et al., 2011), after passage through Tetrahymena. Another potentially important aspect of the interaction of enteric pathogens with this ciliate is the stickiness of fecal pellets, as evidenced by aggregation of the pellets to form visible floc. This property may play a role in the epidemiology of human pathogens, because entrapment in the adhesive floc may contribute to their survival in environments where ciliates encounter bacterial concentrations sufficiently high to stimulate feeding and egestion of bacterial cells in fecal pellets. Therefore, it is reasonable to expect survival and replication of diarrheagenic E. coli when associated with Tetrahymena in our laboratory microcosm to extend to the natural environment. This association may offer an explanation for environmental exposure of animal hosts, including humans, to diarrheagenic E. coli pathotypes.


C.D.S. was supported by USEPA STAR Fellowship F09D40688. We thank Holly Arron, Oishee Bose, and Joanne Dai for excellent laboratory assistance. Helicobacter pylori was a gift of Richard Peek, Vanderbilt University. We thank the reviewers for constructive comments on the manuscript.


  • Editor: Patricia Sobecky


View Abstract