This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.
Agricola
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, November 2003, p. 6848-6855, Vol. 69, No. 11
0099-2240/03/$08.00+0     DOI: 10.1128/AEM.69.11.6848-6855.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.

Protistan Grazing Analysis by Flow Cytometry Using Prey Labeled by In Vivo Expression of Fluorescent Proteins

Yutao Fu,{dagger} Charles O'Kelly, Michael Sieracki, and Daniel L. Distel*

Bigelow Laboratory for Ocean Sciences, West Boothbay Harbor, Maine 04575

Received 23 June 2003/ Accepted 20 August 2003


arrow
ABSTRACT
 
Selective grazing by protists can profoundly influence bacterial community structure, and yet direct, quantitative observation of grazing selectivity has been difficult to achieve. In this investigation, flow cytometry was used to study grazing by the marine heterotrophic flagellate Paraphysomonas imperforata on live bacterial cells genetically modified to express the fluorescent protein markers green fluorescent protein (GFP) and red fluorescent protein (RFP). Broad-host-range plasmids were constructed that express fluorescent proteins in three bacterial prey species, Escherichia coli, Enterobacter aerogenes, and Pseudomonas putida. Micromonas pusilla, an alga with red autofluorescence, was also used as prey. Predator-prey interactions were quantified by using a FACScan flow cytometer and analyzed by using a Perl program described here. Grazing preference of P. imperforata was influenced by prey type, size, and condition. In competitive feeding trials, P. imperforata consumed algal prey at significantly lower rates than FP (fluorescent protein)-labeled bacteria of similar or different size. Within-species size selection was also observed, but only for P. putida, the largest prey species examined; smaller cells of P. putida were grazed preferentially. No significant difference in clearance rate was observed between GFP- and RFP-labeled strains of the same prey species or between wild-type and GFP-labeled strains. In contrast, the common chemical staining method, 5-(4,6-dichloro-triazin-2-yl)-amino fluorescein hydrochloride, depressed clearance rates for bacterial prey compared to unlabeled or RFP-labeled cells.


arrow
INTRODUCTION
 
Protists graze selectively on bacteria and so may strongly influence the taxonomic composition, morphology, physiology, and metabolic capacity of bacterial communities (11, 12, 14, 23, 28, 29). As a result, protistan bacterivory is an important determinant of natural bacterial diversity and may also strongly influence the survival of bacteria artificially introduced into environments (24). This may be of particular concern in predicting the efficacy of bacterial bioremediation efforts and the persistence of genetically modified bacteria in the wild.

It has been proposed that grazing pressure produces morphological and compositional changes in planktonic bacterial communities by removing species and morphotypes that are less grazing resistant and by inducing morphological and physiological changes in pleomorphic species (1, 10, 12). It has also been proposed that selective grazing may increase the abundance and physiological activity of certain bacterial species by removing competitors, releasing nutrients, and selecting for strains capable of increased reproductive rates (10, 21, 22, 31). However, bacterial grazing resistance is a concept that has remained poorly defined, largely because the factors that determine prey selection are numerous, diverse, and difficult to observe directly.

Factors that have been proposed to influence grazing selection are both predator and prey specific. These include taxonomic identity, physiological condition, and feeding capacity of the predator, as well as the identity, condition, nutrient content, surface characteristics, colony-forming ability, motility, shape, and size of the prey (10, 14, 27, 32). Of these factors, prey size and shape have been studied most extensively, in part because these parameters can be observed directly by microscopy.

Traditional microscopy-based techniques, however, have serious limitations. These methods generally require conditions that kill or otherwise alter predator and prey behavior, including fixation, staining, and mounting of cell preparations (3, 6, 25). Most frequently prey cells are stained prior to grazing by using either chemical or immunological labeling methods (3, 6, 17, 21). The most important and widely used labeling method has been the mortal stain, 5-(4,6-dichloro-triazin-2-yl)-amino fluorescein hydrochloride (DTAF) (21).

Traditional microscopy methods are also time-consuming and tedious. Although automated image analysis methods promise to reduce the time and effort involved in analyzing grazing results (13, 26), it remains difficult to obtain large numbers of observations over brief time intervals, thus limiting the effectiveness of statistical analyses. Finally, traditional microscopy methods can identify bacterial morphotypes but cannot identify specific taxa. Although the use of fluorescent in situ hybridization methods with taxon-specific oligonucleotide probes has dramatically improved this situation (13, 28), these methods are time-consuming and still require fixation and postmortem staining of prey cells. Improved methods are needed for real-time, in vivo observation of predator-prey interactions.

Flow cytometry provides a viable alternative to microscopy for studies of protistan bacterivory (2, 8, 33). Compared to conventional microscopy, the power of flow cytometry lies in its high operating speed for particle analysis and its capacity to measure multiple optical signals simultaneously. These optical signals can be used to identify and enumerate particle types according to their unique fluorescent and/or light-scattering properties. Applied to protistan bacterivory, optical signals may be used to rapidly detect, enumerate, and distinguish predator and prey cells in a mixed sample and to evaluate changes in these populations in real time. From these data, clearance rates may be calculated simultaneously for multiple prey types.

Using flow cytometry, predator cells typically can be distinguished from prey cells because their larger size is reflected in larger forward-scattered (FSC) light signals in correlation with distinctive side-scattered (SSC) light signals (20). However, since many prey types have similar optical properties, it is necessary to stain prey cells with fluorescent labels to allow individual types to be distinguished from other prey types and nonprey particles.

A serious concern with regard to the use of labeled prey in bacterivory studies, however, is that the labeling method or the label itself may alter predator or prey properties or behavior, thereby altering clearance rates. This is of particular concern for DTAF labeling since this staining method kills most bacterial cells and chemically alters cell surfaces. An alternative approach to chemical staining in grazing studies is in vivo labeling by the introduction of genes expressing fluorescent proteins (FPs), e.g., green FP (GFP) or red FP (RFP), into bacterial prey species (7, 15, 18). These proteins are expressed in the cytoplasm and so are not expected to alter the surface properties of the prey cells. Furthermore, these labels can be detected in live cells without fixation or addition of reagents that may affect predator or prey behavior.

In this investigation, flow cytometry was used to observe the grazing behavior of the heterotrophic nanoflagellate Paraphysomonas imperforata when exposed to multiple microbial prey types labeled by in vivo expression of FPs. The resulting method, here called FCM-LIVE (flow cytometry with labeling by in vivo expression of FPs), (i) allows simultaneous observation of grazing parameters for multiple prey types, (ii) uses in vivo labeling that neither kills nor significantly alters prey properties, and (iii) employs data collection methods that are amenable to automated, real-time, high-throughput data generation and analysis. The results demonstrate that grazing preference of P. imperforata is influenced by prey type, size, and condition but is not affected by GFP and RFP expression. In contrast, DTAF staining resulted in significant decrease in clearance rates compared to unlabeled or FP-labeled cells.


arrow
MATERIALS AND METHODS
 
Plasmid construction.
Broad-host-range plasmids containing GFP or RFP genes were constructed as follows: the sequence encoding GFP (from positions 30 to 1082 numbered relative to the start of the multiple cloning site) from plasmid pQBIT7GFP (QBiogene, Carlsbad, Calif.) was excised by using HindIII/XbaI digestion. The resulting fragment (1,052 bp) was then inserted into the HindIII/XbaI sites of the broad-host-range expression plasmid pMMB67EH (ATCC 37622) at positions 8804 and 8828 to construct plasmid pMMBGFP. The lacIq fragment of pMMBGFP was then disrupted by excision of a PvuII fragment (pMMB67EH positions 7357 to 7450 bp) to give pMMBGFPß. The RFP coding sequence from pDsRed (Clontech, Palo Alto, Calif.) was excised by PvuII/StuI digestion (positions 55 to 1041), and the resulting 986-bp fragment was inserted into pBBR122 as described for GFP constructs above to produce the plasmid pBBRRFP. The location and orientation of all insertions were verified by restriction fragment length polymorphism analysis. Molecular recombination procedures were performed as described previously (19). Restriction digestions were done in 50-µl volumes with 0.5 U of enzyme (37°C for 2 h). Digested fragments were separated by agarose gel electrophoresis and recovered by using a QIAquick gel extraction kit (Qiagen, Valencia, Calif.) according to product protocols. Ligations were performed overnight with an insert/vector ratio of 3:1 in a 20-µl mixture containing 2 U of T4 DNA ligase (New England Biolabs, Beverly, Mass.) at 16°C for cohesive ended and room temperature for blunt-ended inserts.

Transformation and preparation of prey strains.
Escherichia coli BL21(DE3) (Novagen, Madison, Wis.), Enterobacter aerogenes NCDC 819-56, and Pseudomonas putida KT2442 cells were inoculated from frozen stocks (-80°C) into 2 ml of liquid Luria-Bertani (LB) medium (19) (Fisher Scientific) and were grown with shaking (250 rpm) for 12 h at 37°C (E. coli and E. aerogenes) or at 30°C (P. putida) prior to transformation.

E. coli was transformed by calcium heat shock as described earlier (19). Competent cells were prepared by washing early-log-phase LB culture (optical density at 600 nm [OD600], ca. 0.4 to 0.6) twice with 1/5 volume of ice-cold 0.1 M CaCl2 and resuspending cells in 0.1 M CaCl2 to 1/10 the original volume. After overnight incubation on ice, 0.1-ml portions of competent cells were mixed with 0.5 µg of plasmid DNA, heated shocked at 42°C for 90 s, and then mixed with 0.9 ml of SOC medium (19). After a 1-h incubation at 37°C with shaking (250 rpm), the transformed cells were spread onto LB plates amended with the appropriate antibiotic.

E. aerogenes was transformed by inoculating wild-type cells into 5 ml of LB medium with 1 mM MgCl2, followed by incubation at 30°C with shaking (250 rpm) until mid-log phase (OD = ~0.6). Cells were then sedimented by centrifugation (8,000 x g, 10 min, 4°C). The cell pellet was washed twice with 1 ml of ice-cold 0.01 M CaCl2, resuspended in 0.5 ml of cold CaCl2, and allowed to incubate for 2 h on ice. Cells were precipitated again (5,000 x g, 3 min, 4°C), washed twice with 1 ml of ice-cold 10% glycerol, and resuspended in the same solution. An aliquot of 0.1 ml of the resultant competent cells were mixed with 1 µg of plasmid. After a 10-min incubation on ice, the transformation mixture was frozen at -80°C for 5 min and then thawed on ice. The cells were then electroporated (15 kV, 200 {Omega}, 25 µF, {tau} {approx} 4.6) by using a GenePulser electroporation device (Bio-Rad, Hercules, Calif.). Electroporated cells were gently transferred to a culture tube and allowed to stand on ice for 1 min, followed by a 3-min incubation at 37°C. SOC medium (0.9 ml) was then added, and the cells were incubated at 30°C for 1 h. Cells were then spread onto LB agar plates amended with the appropriate antibiotic.

P. putida was transformed by electroporation as described earlier (19). Briefly, cultured P. putida cells were washed twice and resuspended in ice-cold 10% glycerol. An aliquot of 0.1 ml of competent cells was mixed with 1 µg of plasmid, transferred to a prechilled electroporation cuvette, incubated on ice for 10 min, and electroporated (15 kV, 200 {Omega}, 25 µF, {tau} {approx} 4.6). Electroporated cells were gently mixed with 0.9 ml of SOC medium and incubated at 30°C for 1 h before plating them on LB-ampicillin agar plates.

Transformed strains are denoted by codes for the strains (Ec for Escherichia coli, Ea for Enterobacter aerogenes, or Pp for Pseudomonas putida) with an appended suffix -GFP or -RFP, to indicate the plasmid type used for transformation. Transformed bacteria were cultivated in the presence of appropriate antibiotics (50 µg of ampicillin/ml for GFP strains, except Pp-GFP [see below]) or 10 µg of kanamycin/ml for RFP strains. Optimum fluorescence was achieved for strain Pp-GFP after cultivation in liquid medium containing 250 µg of ampicillin/ml plus IPTG (isopropyl-ß-D-thiogalactopyranoside; Sigma) at 1 µg/ml, followed by overnight growth on an LB ampicillin agar plate at 37°C and inoculation from a single plate colony to new LB liquid medium. Similarly, Ec-RFP achieved optimum fluorescence after extended incubation (48 h) at 37°C with shaking, followed by 24 h at 4°C.

Micromonas pusilla IB4 strain CCMP490 was obtained from the Provasoli-Guillard Center for Culture of Marine Phytoplankton, Bigelow Laboratory, West Boothbay Harbor, Maine, and was maintained in f/2 mineral medium (9) at 21°C with 12-h light-dark cycling (9).

Chemical staining of prey cells.
DTAF (Sigma) staining was performed according to Vazquez-Dominguez et al. (8). Bacterial cells in mid-logarithmic phase of growth (OD ~0.6) were harvested by centrifugation at 5,000 x g for 3 min and washed twice by resuspension with alkaline phosphate-buffered saline (0.05 M Na2HPO4-0.85% NaCl adjusted to pH 9.0) of equal volume. Cells were then resuspended in phosphate-buffered saline at 109 cells/ml. DTAF was added to a final concentration of 200 µg/ml, and cells were allowed to incubate at 60°C for 2 h. Stained cells were washed six times by centrifugation (3,600 x g for 3 min at room temperature) and resuspension in carbonate-bicarbonate buffer (0.1 M, pH 9.5).

PicoGreen staining was used to enumerate nonfluorescent wild-type prey cells. Cells in mid-logarithmic growth phase (OD ~0.6) were harvested and washed as described above and were diluted 1:1,000 with 0.2-µm (pore size)-filtered seawater (FSW; obtained from the Center for Culture of Marine Phytoplankton). One milliliter of diluted bacterial cells was fixed by addition of 50 µl of 37% paraformaldehyde (30 min, 4°C, in the dark), and then stained by the addition of 10 µl of Pico Green stock solution (Molecular Probes, Inc., Eugene, Oreg.) for 15 min at room temperature.

Microscopy.
Cells were observed by using a Labophot-2 epifluorescence microscope (Nikon) by both fluorescence and transmitted light. Transformed and wild-type bacterial cells were measured with a stage micrometer (Spencer Lens Co., Buffalo, N.Y.), and protist and bacterial cells were enumerated visually with a hemocytometer (Fisher Scientific). To observe fluorescence in food vacuoles of P. imperforata, cells were immobilized by using 0.5% NiSO4. Fluorescence images were taken with a SPOT-2 charge-coupled digital camera (Diagnostic Instruments, Inc., Sterling Heights, Mich.). The following filter sets and settings were used for epifluoresence microscopy: (i) red filter set—exciter filter (540 to 580 nm), dichroic mirror (595 nm), and barrier filter (600 to 660 nm); (ii) green filter set—exciter filter (450 to 490 nm), dichroic mirror (480 nm), and barrier filter (510 to 550 nm); and (iii) red-green dual-wavelength filter—exciter filter (450 to 490 nm), dichroic mirror (505 nm), and barrier filter (520 nm).

Bacterivory assays.
Stock cultures of P. imperforata originally obtained from D. Caron (University of Southern California) were maintained in 100 ml of L1 mineral medium (9) or organic medium (30) in 250-ml flasks without shaking at 21°C in the dark. One milliliter of culture was transferred to fresh medium at 30-day intervals. Cultures were fed with E. aerogenes aseptically washed from LB plates into L1 and pipetted into culture flasks to a final concentration of ca. 109 cells/ml.

Grazing trials were performed in 100-ml plastic tissue culture flasks each containing 25 ml of FSW to which predator cells were added. Final predator cell density in experimental flasks was adjusted to ~105 cells/ml. A FACScan flow cytometer (Becton Dickinson, Inc., Franklin Lakes, N.J.) equipped with a 15-mW argon laser at 488 nm and three-color fluorescence (emissions: fluorescence channel 1 [FL1], 510 to 525 nm; FL2, 560 to 590 nm; and FL3, >650 nm) was used to estimate predator cell numbers by analysis of forward-scatter (FSC) and side-scatter (SSC) optical signals. Prey cells were washed with FSW five times by centrifugation (5,000 x g, 3 min at room temperature), followed by resuspension and dilution with FSW to an final estimated density of 106 prey cells/ml as determined by FL1/FSC optical signals.

Sampling for each grazing trial was done with at least three time points and at least three replicates for each time point. Each grazing trial included an experimental treatment flask as well as no predator and no prey control flasks, which were identical to the experimental flask except for the omission of predator or prey cells. Grazing trials were started by the addition of prey cells and were conducted in the dark at room temperature without shaking for the time intervals specified in individual experiments. At each sampling time point the flask was mixed thoroughly, and aliquots of 0.5 ml were transferred to 10-ml glass FACScan sample tubes. Tubes were weighed before and after sampling to determine the flowthrough volume. Each tube was sampled for 3 min at a sample flow rate of 10 µl/s. For all experiments, the following settings were used: an SSC voltage of 300 and an amplifier gain 1.0; an FSC voltage of E00 and an amplifier gain of 1.0; and an FL1 of 550 V, an FL2 of 520 V, and an FL3 of 550 V. Counting was triggered by SSC, with a threshold voltage of 150. Sheath and sample fluids were FSW.

Before grazing analysis, prey density was determined and detector settings were optimized for each prey strain in pure culture by using the FACScan. Cultured bacteria were washed twice with phosphate-buffered saline and twice with FSW by centrifugation at 5,000 x g for 3 min and then were resuspended in an equal volume of FSW prior to sampling.

The dominant hypothesis currently used to describe protistan bacterivory proposes that clearance rate is limited primarily by the frequency of prey encounters. Therefore, prey density time curves may be simulated to the kinetics of first-order reactions (8, 14, 21). Using proper gating criteria, densities of free-living prey cells can be extracted from FCM data. Decrease of prey densities along a time series reflects the intensity of grazing. For prey species that display rapid production or high mortality, it is necessary to subtract the change in density due to prey production and nongrazing death from the apparent prey loss to obtain accurate estimates of clearance due to bacterivory. Clearance rate (R) can be calculated according to the formula

where M is the density of predator cells, T is the time elapsed, N is the density of prey cells in experimental flasks, and C is the prey density in no-predator control flasks.

Data analysis.
Flow cytometry data were saved as FCS2.0 list mode files by using CellQuest (Becton Dickinson) and were analyzed by using the WinMDI flow cytometry interface (Joseph Trotter [The Scripps Research Institute], unpublished data). Optical parameter ranges for recognition of prey and predator cells were determined empirically and were used to gate scattergrams (see Fig. 2) and histograms to produce plain-text readouts for predator and prey statistics. WinMDI output was analyzed by GR, a Perl program written to calculate clearance rates and signal kinetics (available online [http://phoenix.umecit.maine.edu/~yutao/gr]). The functions of GR include management and annotation of FCM data, clearance rates calculation, and statistical analysis of experimental data or grazing results. Analysis of variance (ANOVA) was used to calculate the probability (P) for statistically significant difference in clearance rates in dilution experiments, and Student t test (paired) was used for the remaining experiments in which more than one prey species was examined.



View larger version (20K):
[in this window]
[in a new window]
  		FIG. 2. Optical signals generated by flow cytometry. (A and B) P. imperforata (A) and E. aerogenes (B) expressing GFP (strain Ea-GFP); (C) E. coli expressing RFP (strain Ec-RFP) in a mixed population. Solid outlines indicate empirically determined gate values assigned to delineate predator and prey populations.


arrow
RESULTS AND DISCUSSION
 
Three bacterial species, E. coli, E. aerogenes, and P. putida were transformed with broad-host-range GFP or RFP expression plasmids pMMBGFPß or pBBRRFP to produce six fluorescent strains: Ec-GFP, Ec-RFP, Ea-GFP, Ea-RFP, Pp-GFP, and Pp-RFP. When cultivated in LB medium, transformed E. aerogenes, E. coli, and P. putida cells are rods of 0.6 by 1.2 µm, 0.9 by 2.4 µm, and 1.3 by 4.2 µm, respectively (average of 100 cells as determined by using a microscope fitted with an stage micrometer) and do not differ significantly in size and shape from wild-type cells. Fluorescent signals produced by all transformed bacteria were sufficiently intense to allow direct observation by epifluorescence microscopy (Fig. 1).



View larger version (61K):
[in this window]
[in a new window]
  		FIG. 1. In vivo expression of RFP and GFP by genetically modified bacteria. (A) E. coli (Ec-RFP; red) and E. aerogenes (Ea-GFP; green); (B) E. coli (Ec-RFP; red) and P. putida (Pp-GFP; green). (C to F) Fluorescence of Ec-RFP (red) and Ea-GFP (green) after ingestion by a single cell of P. imperforata. (C) GFP epifluorescence image; (D) RFP epifluorescence image; (E) phase-contrast image; (F) combined images (i.e., panels C to E). Fluorescence is observed both from distinct bacterial cells and from coalesced food vacuoles, suggesting the ongoing digestion of prey. Images were taken 3 min after the introduction of prey. Scale bars: A and B, 10 µm; C to F, 1.0 µm.

Transformed bacteria could also be detected by the FACScanflow cytometer. The excitation and emission maxima (474-nm excitation and 509-nm emission) of the red-shifted GFP variant used in the present study matched sufficiently well with the 488-nm argon laser and the FL1 detector used by the FACScan device to allow efficient detection of all GFP transformants. However, the RFP variant (558-nm excitation and 585-nm emission) was only weakly excited by the argon laser and could be detected above background by the FL2 detector channel only for Ec-RFP, the brightest of the RFP-labeled strains. The FL2 channel also efficiently detected cells of the green alga Micromonas pusilla due to its strong red autofluorescence. Therefore, Ec-RFP or M. pusilla cells were used for all experiments requiring red-labeled prey.

Predators and fluorescent prey cells could be detected and distinguished readily by using various combinations of FSC, SSC, and fluorescent (FL1 and FL2) optical signals generated by the FACScan flow cytometer. For example, Fig. 2 demonstrates resolution of a mixed population of P. imperforata, Ec-RFP, and Ea-GFP. The P. imperforata cells could be distinguished from the bacterial prey populations by their distinct combination of FSC and SSC signals (Fig. 2A). This reflects the significantly larger size of P. imperforata and differences in its surface properties compared to prey cells. Due to overlap in their size ranges and surface properties, unlabeled prey cell types could not be reliably distinguished from one another by differences in their FSC and SSC signals. However, red and green fluorescent strains could easily be distinguished from predators, other prey types, and nonprey particles by using FSC and fluorescent signals: FL1 for GFP-labeled strains (Fig. 2B) and FL2 for Ec-RFP (Fig. 2C) and M. pusilla.

Using these combined optical signals it was possible to observe grazing by P. imperforata and to calculate individual grazing rates for red- and green-labeled prey simultaneously and in real time. Significant reduction in prey density was observed for all prey types in the presence of P. imperforata. For example, Fig. 3 shows changes in density for Ea-GFP and Ec-RFP with time after the introduction of P. imperforata. Prey densities in control flasks, which received no predator cells, remained stable (<5% variation from initial values) during the total time course of the experiment (135 min), whereas prey densities declined exponentially under grazing pressure from P. imperforata. Clearance rates calculated for this experiment were 0.203 ± 0.119 and 0.271 ± 0.044 nl predator-1·h-1 for Ea-GFP and Ec-RFP, respectively. In all grazing trials, clearance rates for individual prey types did not vary significantly over successive time intervals examined, indicating that grazing capacity of predators was undiminished throughout the experimental periods (0 to 180 min).



View larger version (12K):
[in this window]
[in a new window]
  		FIG. 3. Grazing by P. imperforata. Change in prey cell density for FP-labeled prey over time in the presence (solid symbols) or absence (open symbols) of grazing by P. imperforata (Ea-GFP, squares; Ec-RFP, diamonds).

Clearance rates were strongly dependent on predator culture conditions and varied widely from batch to batch of predator culture; therefore, grazing comparisons were made only within individual batches of predator culture. Clearance rates determined by FCM in this investigation (0.04 to 1.55 nl predator-1·h-1) were consistent with values reported in the literature (0.1 to 10 nl predator-1·h-1) (8, 14, 21).

Grazing could also be observed and monitored by the accumulation of FP-labeled bacteria in the food vacuoles of P. imperforata, as determined either by epifluorescence microscopy (Fig. 1) or by FACScan signals (Fig. 4). When P. imperforata was exposed to bacteria expressing GFP or RFP, individually or in combination, fluorescence could be detected in food vacuoles within 1 min after a mixing of predators and prey (Fig. 1). Well-defined individual prey cells, as well as coalesced fluorescent zones, were detected in food vacuoles, indicating various stages of digestion of labeled prey cells.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 4. Normalized geometric mean of FL signals from protist food vacuoles. The fluorescent signals in food vacuoles of P. imperforata during grazing on FP-labeled bacterial strains Ec-RFP ({blacktriangleup}) and Ea-GFP ({diamondsuit}) and chemically stained bacteria Ea-DTAF ({blacksquare}) were detemined. Values are expressed as the mean fluorescence per P. imperforata cell normalized by using the 135-min readings as 100%.

Rapid increase in fluorescence intensity per predator cell was observed in the first 45 min after initiation of feeding with Pp-GFP and Ec-RFP, as well as with DTAF-labeled Ea-DTAF prey cells. Accumulation of fluorescence in predator cells from both FP-labeled prey types reached an asymptote within 90 min, whereas that for the DTAF-labeled cells continued to increase throughout the 135-min time course (Fig. 4). The observed decreases in the rates of fluorescence accumulation, however, were not accompanied by significant changes in clearance rates for all prey types. These results suggest that fluorescence was lost during the course of the experiment due to digestion of labeled prey rather than egestion or decreased ingestion of these prey cells. Given that the clearance rate for DTAF-stained cells was lower than that for FP-labeled cells, these data suggest that DTAF is not as readily degraded by P. imperforata or that Ea-DTAF cells are digested at a lower rate than those of Ec-RFP or Ea-GFP cells.

An important assumption in experimental grazing systems is that labeled and unlabeled bacterial cells are equally suitable to serve as prey for the experimental predators. One type of evidence that has been presented to support the suitability of prey has been to determine whether labeled and unlabeled prey cells produce similar growth of the predator (21). In long-term predation experiments, the growth of P. imperforata populations was statistically indistinguishable (P > 0.5 [ANOVA]) when P. imperforata was fed with wild-type E. aerogenes cells compared to Ea-GFP and Ea-RFP (Fig. 5). These results indicate that FP-labeled E. aerogenes strains are not significantly different from wild-type E. aerogenes with respect to suitability as prey items under the conditions examined. However, such experiments do not address prey preference since the predator is exposed to only a single prey type at a single density in each experiment.



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 5. Growth of P. imperforata is supported by grazing on FP-labeled bacterial strains. Change in cell densities of P. imperforata over time in the presence of Ea-GFP ({diamondsuit}), Ea-RFP (•), or unlabeled E. aerogenes ({blacktriangleup}) or without added bacterial prey ({blacksquare}). No significant differences in growth rate or yield of were observed for P. imperforata grown on FP-labeled and unlabeled prey of the same species. Error bars are omitted for clarity.

To observe prey preference directly, clearance rates by P. imperforata were determined for two prey types presented simultaneously. Table 1 presents the results of grazing trials in which P. imperforata was presented with a single bacterial prey type (Ea-GFP or Pp-GFP cells) either alone or in combination with an equal density of algal prey (M. pusilla). Clearance rates measured simultaneously for both prey types show that in each case the bacterial prey was consumed at a significantly greater rate (P < 0.05) than the algal prey. Furthermore, the presence of algal cells had no significant effect on clearance rates for the Ea-GFP or Pp-GFP strains, indicating that the observed differences in clearance rates were due to preferential grazing for the bacterial prey types over the algal prey rather than to changes in grazing behavior in the presence of mixed prey types.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Clearance rates for GFP-labeled bacteria and an autofluorescent algal prey strain

Although prey type had a significant influence on prey selection, selection for prey size could also be observed in the same experiments. Ea-GFP cells are similar to M. pusilla in mean lengths (~1.2 and 1.8 µm, respectively) and maximum lengths (~1.5 and 2.0 µm, respectively), while the mean and maximum lengths of Pp-GFP cells are significantly larger (4.3 and 6.0 µm, respectively). Because the FSC signal is proportional to the particle size, it is possible to observe changes in the mean cell size of the bacterial population at each time point during a grazing trial. Figure 6A shows that, for Pp-GFP, the larger of the two bacterial prey types, the mean cell size, as reflected by the FSC signal, increased significantly after each time interval in the presence of the predator P. imperforata. No significant change in mean cell size was observed for Pp-GFP cells in the absence of the predator. These results suggest that smaller cells within the Pp-GFP population were consumed preferentially over larger cells. From the change in the mean cell size of the Pp-GFP population remaining after grazing, it is possible to infer the mean size of the Pp-GFP cells that were consumed during each grazing interval (Fig. 6B). These values demonstrate that the prey cells consumed in the first and second grazing intervals were significantly smaller on average than those consumed in the third interval.



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 6. Time kinetics for change in mean cell size as reflected by FSC light signals for the bacterial prey strain Pp-GFP. (A) Mean FSC signals for the prey population remaining after indicated time intervals with (solid bars) and without (open bars) grazing by P. imperforata; (B) inferred mean FSC signals for prey ingested during successive time intervals. During the initial time intervals, the mean cell size for ingested cells was significantly smaller than the mean size for the total prey population, resulting in a gradual increase in the mean size of the residual prey population. These results indicate that smaller Pp-GFP cells were consumed preferentially by P. imperforata.

We speculate that smaller cells were consumed preferentially and that the average size of consumed cells increased only after grazing depleted the smaller size classes. In contrast, no significant change in the mean cell size was observed for the smaller prey type, Ea-GFP, suggesting that most cells of Ea-GFP fell within the prey size range preferred by P. imperforata, whereas the largest cells of Pp-GFP exceeded that range.

Another significant concern in grazing experiments that require labeling of prey cells is the possibility that labeling methods may influence clearance rates. In the competitive grazing trials described here, no significant difference in clearance rates were observed between GFP- and RFP-labeled strains of the same species or between similar-sized GFP- and RFP-labeled bacteria of different species (Table 2). Therefore, GFP and RFP-labeled cells appear to be equivalent with respect to grazing by P. imperforata.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Clearance rates for GFP- and RFP-labeled bacterial strains

It is not possible to directly determine whether clearance rates for labeled cells differ from those for unlabeled cells in this system since the latter cannot be unambiguously identified by flow cytometry. However, by comparing clearance rates for labeled cells in the presence or absence of unlabeled cells of the same species it is possible to determine whether consumption of unlabeled cells either depresses or stimulates grazing for labeled cells. Ea-GFP cells were consumed at statistically indistinguishable clearance rates (P = 0.99 [ANOVA]) whether presented to prey undiluted (0.594 ± 0.15 nl predator-1 h-1) or diluted 10-fold (0.592 ± 0.261 nl predator-1 h-1) or 100-fold (0.609 ± 0.285 nl predator-1 h-1) with unlabeled cells. Thus, predator populations feeding primarily on unlabeled cells consume labeled prey at the same rate as cells feeding entirely on labeled prey, indicating that consumption of GFP-labeled cells has no inhibitory or stimulatory effects on clearance rates for GFP-labeled cells.

In contrast, Ec-RFP cells were grazed by P. imperforata at significantly (P = 0.007 [Student t test]) greater rates (0.554 ± 0.113 nl predator-1 h-1) than the identical Ec-RFP strain stained with DTAF (Ec-RFP/DTAF) (0.389 ± 0.098 nl predator-1 h-1) in competitive grazing trials. Furthermore, clearance rates for Ea-DTAF cells were significantly lower when presented undiluted (0.841 ± 0.290 nl predator-1 h-1) than when diluted 10-fold (1.33 ± 0.40 nl predator-1 h-1; P = 0.045 [Student t test]) or 100-fold (1.55 ± 0.54 nl predator-1 h-1; P = 0.019 [Student t test]) with unlabeled cells of the same strain. Thus, DTAF staining has a negative effect on clearance rate (P = 0.035 [ANOVA]) that is positively related to the ratio of labeled to unlabeled cells. The exponential model predicts no change in clearance rate with dilution of prey. Therefore, these results suggest that the observed clearance rate depression was due to a physiological response to consumption of DTAF-labeled prey. This effect is at least partially overcome by reducing the proportion of consumed cells that are DTAF labeled.

Thus, in these experiments, DTAF labeling of prey cells affected grazing in two ways: first, by altering grazing preference and, second, by reducing overall clearance rates. The cause of these effects is unclear. DTAF staining methods include a heat treatment step that kills most bacteria. It is conceivable that living cells are grazed preferentially by P. imperforata over dead cells of the same strain. Alternatively, DTAF binding to cell surfaces could negatively affect grazing selection, possibly by altering prey recognition or by some type of toxic effect. Unfortunately, it is not possible to distinguish between these alternatives with this experimental system because GFP- and RFP-labeled cells lose fluorescence after cell death (16). Therefore, it is not possible to DTAF label live cells or to detect GFP-labeled dead cells. These results are in agreement with apparent reductions in grazing rates observed previously for DTAF-stained prey compared to live prey stained by using vital stains (6) or prey stained after consumption by the predator (3).

In conclusion, the methods described here provide the means to monitor protistan bacterivory that is, in some respects, superior to previously described methods. The results demonstrate quantifiable differences in grazing behavior of a model protist in response to differences in prey type (algal versus bacterial cells), condition (chemically stained dead cells versus live unstained or FP-labeled cells), and prey size (large versus small cells within a population composed of a single bacterial species). Moreover, multiple preference parameters, e.g., size and taxonomic type, can be observed and quantified within a single experiment. The methods are well suited for exploring, modeling, and predicting the parameters that modify the grazing behavior of protists in the wild and for predicting the influence of protistan grazing on indigenous and introduced bacterial species. Although the method as currently described is designed for microcosm studies, we suspect that similar approaches may be feasible for observation of field samples, e.g., monitoring of grazing effects on genetically modified bioremediative bacteria after release into contaminated environments (4, 5).


arrow
ACKNOWLEDGMENTS
 
We thank Roy Turner (University of Maine) and Martin Polz (Massachusetts Institute of Technology) for thoughtful discussions of the manuscript and the GR program and Ed Thier, Terry Cucci, and Wendy Bellows for technical assistance with cell culture and FCM.

This work was supported by research grants from the Office of Naval Research (N00014-99-1-0514) and the National Science Foundation (IBN-998298).


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Department of Biochemistry, Microbiology, and Molecular Biology, University of Maine, 5735 Hitchner Hall, Orono, ME 04469-5735. Phone: (207) 581-2824. Fax: (207) 581-2801. E-mail: distel{at}maine.edu Back

{dagger} Present address: Bioinformatics Program, Boston University, Boston, MA 02115. Back


arrow
REFERENCES
 
    1
  1. Boenigk, J., and H. Arndt. 2002. Bacterivory by heterotrophic flagellates: community structure and feeding strategies. Antonie Leeuwenhoek 81:465-480.
    2. 2
  2. Christaki, U., C. Courties, H. Karayanni, A. Giannakourou, C. Maravelias, K. A. Kormas, and P. Lebaron. 2002. Dynamic characteristics of Prochlorococcus and Synechococcus consumption by bacterivorous nanoflagellates. Microb. Ecol. 43:341-352.[CrossRef][Medline]
    3. 3
  3. Christoffersen, K., O. Nybroe, K. Jurgens, and M. Hansen. 1997. Measurement of bacterivory by heterotrophic nanoflagellates using immunofluorescence labeling of ingested cells. Aquat. Microb. Ecol. 13:127-134.
    4. 4
  4. Daubaras, D. A. M. C. 1992. The environment, microbes, and bioremediation: microbial activities modulated by the environment. Biodegradation 3:125-135.
    5. 5
  5. Ellis, L. B., C. D. Hershberger, and L. P. Wackett. 2000. The University of Minnesota Biocatalysis/Biodegradation database: microorganisms, genomics, and prediction. Nucleic Acids Res. 28:377-379.[Abstract/Free Full Text]
    6. 6
  6. Epstein, S. S., and J. Rossel. 1995. Methodology of in situ grazing experiments: evaluation of a new vital dye for preparation of fluorescently labeled bacteria. Mar Ecol. Prog. Ser. 128:143-150.
    7. 7
  7. Errampalli, D., K. Leung, M. B. Cassidy, M. Kostrzynska, M. Blears, H. Lee, and J. T. Trevors. 1999. Applications of the green fluorescent protein as a molecular marker in environmental microorganisms. J. Microbiol. Methods 35:187-199.[CrossRef][Medline]
    8. 8
  8. Evaristo Vazquez-Dominguez, F. P., J. M. Gasol, and D. Vaqué. 1999. Measuring the grazing losses of picoplankton: methodological improvements in the use of fluorescently labeled tracers combined with flow cytometry. Aquatic Microb. Ecol. 20:119-128.
    9. 9
  9. Guillard, R. R. L., and P. E. Hargraves. 1993. Stichochrysis immobilis is a diatom, not a chyrsophyte. Phycologia 32:234-236.
    10. 10
  10. Hahn, M. W., and M. G. Hofle. 1999. Flagellate predation on a bacterial model community: interplay of size-selective grazing, specific bacterial cell size, and bacterial community composition. Appl. Environ. Microbiol. 65:4863-4872.[Abstract/Free Full Text]
    11. 11
  11. Hahn, M. W., and M. G. Hofle. 1998. Grazing pressure by a bacterivorous flagellate reverses the relative abundance of Comamonas acidovorans PX54 and Vibrio strain CB5 in chemostat cocultures. Appl. Environ. Microbiol. 64:1910-1918.[Abstract/Free Full Text]
    12. 12
  12. Jurgens, K., and C. Matz. 2002. Predation as a shaping force for the phenotypic and genotypic composition of planktonic bacteria. Antonie Leeuwenhoek 81:413-434.
    13. 13
  13. Jurgens, K., J. Pernthaler, S. Schalla, and R. Amann. 1999. Morphological and compositional changes in a planktonic bacterial community in response to enhanced protozoan grazing. Appl. Environ. Microbiol. 65:1241-1250.[Abstract/Free Full Text]
    14. 14
  14. Kinner, N. E., R. W. Harvey, K. Blakeslee, G. Novarino, and L. D. Meeker. 1998. Size-selective predation on groundwater bacteria by nanoflagellates in an organic-contaminated aquifer. Appl. Environ. Microbiol. 64:618-625.[Abstract/Free Full Text]
    15. 15
  15. Leung, K. T., J. S. So, M. Kostrzynska, H. Lee, and J. T. Trevors. 2001. Using a green fluorescent protein gene-labeled p-nitrophenol-degrading Moraxella strain to examine the protective effect of alginate encapsulation against protozoan grazing. J. Microbiol. Methods 39:205-211.
    16. 16
  16. Lowder, M., A. Unge, N. Maraha, J. K. Jansson, J. Swiggett, and J. D. Oliver. 2000. Effect of starvation and the viable-but-nonculturable state on green fluorescent protein (GFP) fluorescence in GFP-tagged Pseudomonas fluorescens A506. Appl. Environ. Microbiol. 66:3160-3165.[Abstract/Free Full Text]
    17. 17
  17. Nygaard, K., and D. O. Hessen. 1990. Use of C-14 protein-labeled bacteria for estimating clearance rates by heterotrophic and mixotrophic flagellates. Mar Ecol. Prog. Ser. 68:7-14.
    18. 18
  18. Parry, J. D., K. Heaton, J. Drinkall, and H. L. Jones. 2001. Feasibility of using GFP-expressing Escherichia coli, coupled with fluorimetry, to determine protozoan ingestion rates. FEMS Microbiol. Ecol. 35:11-17.[CrossRef][Medline]
    19. 19
  19. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
    20. 20
  20. Shapiro, H. M. 1995. Practical flow cytometry, 3rd ed. Wiley-Liss, New York, N.Y.
    21. 21
  21. Sherr, B. F., E. B. Sherr, and R. D. Fallon. 1987. Use of monodispersed fluorescently labeled bacteria to estimate in situ protozoan bacterivory. Appl. Environ. Microbiol. 53:958-965.[Abstract/Free Full Text]
    22. 22
  22. Sherr, B. F., E. B. Sheer, and T. Berman. 1982. Decomposition of organic detritus: a selective role for microflagellate protozoa. Limnol. Oceanogr. 27:765-769.
    23. 23
  23. Sherr, E. B., and B. F. Sherr. 2002. Significance of predation by protists in aquatic microbial food webs. Antonie Leeuwenhoek 81:293-308.
    24. 24
  24. Sibille, I., T. Sime-Ngando, L. Mathieu, and J. C. Block. 1998. Protozoan bacterivory and Escherichia coli survival in drinking water distribution systems. Appl. Environ. Microbiol. 64:197-202.[Abstract/Free Full Text]
    25. 25
  25. Sieracki., M. E., L. W. Haas, D. A. Caron, and E. J. Lessard. 1987. Effect of fixation on particle retention by microflagellates: underestimation of grazing rates. Mar. Ecol. Prog. Ser 38:251-258.
    26. 26
  26. Simek, K., J. Armengol, M. Comerma, J. C. Garcia, P. Kojecka, J. Nedoma, and J. Hejzlar. 2001. Changes in the epilimnetic bacterial community composition, production, and protist-induced mortality along the longitudinal axis of a highly eutrophic reservoir. Microb. Ecol. 42:359-371.[CrossRef][Medline]
    27. 27
  27. Simek, K., and T. H. Chrzanowski. 1992. Direct and indirect evidence of size-selective grazing on pelagic bacteria by fresh-water nanoflagellates. Appl. Environ. Microbiol. 58:3715-3720.[Abstract/Free Full Text]
    28. 28
  28. Simek, K., J. Nedoma, J. Pernthaler, T. Posch, and J. R. Dolan. 2002. Altering the balance between bacterial production and protistan bacterivory triggers shifts in freshwater bacterial community composition. Antonie Leeuwenhoek 81:453-463.
    29. 29
  29. Simek, K., J. Pernthaler, M. G. Weinbauer, K. Hornak, J. R. Dolan, J. Nedoma, M. Masin, and R. Amann. 2001. Changes in bacterial community composition and dynamics and viral mortality rates associated with enhanced flagellate grazing in a mesoeutrophic reservoir. Appl. Environ. Microbiol. 67:2723-2733.[Abstract/Free Full Text]
    30. 30
  30. Sin, Y., K. L. Webb, and M. E. Sieracki. 1998. Carbon and nitrogen densities of the cultured marine heterotrophic flagellate Paraphysomonas sp. J. Microbiol. Methods 34:151-163.
    31. 31
  31. Tso, S. F., and G. L. Taghon. 1997. Enumeration of protozoa and bacteria in muddy sediment. Microb. Ecol. 33:144-148.[CrossRef][Medline]
    32. 32
  32. Verity, P. G. 1991. Feeding in planktonic protozoans: evidence for non-random acquisition of prey. J. Protozool. 38:69-76.
    33. 33
  33. Weisse, T. 2002. The significance of inter- and intraspecific variation in bacterivorous and herbivorous protists. Antonie Leeuwenhoek 81:327-341.

Applied and Environmental Microbiology, November 2003, p. 6848-6855, Vol. 69, No. 11
0099-2240/03/$08.00+0     DOI: 10.1128/AEM.69.11.6848-6855.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • de Werra, P., Baehler, E., Huser, A., Keel, C., Maurhofer, M. (2008). Detection of Plant-Modulated Alterations in Antifungal Gene Expression in Pseudomonas fluorescens CHA0 on Roots by Flow Cytometry. Appl. Environ. Microbiol. 74: 1339-1349 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.
Agricola
Right arrow Articles by Fu, Y.
Right arrow Articles by Distel, D. L.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»