Preferential feeding by the ciliates Chilodonella and Tetrahymena spp. and effects of these protozoa on bacterial biofilm structure and composition

Abstract

Protozoa are important components of microbial food webs, but protozoan feeding preferences and their effects in the context of bacterial biofilms are not well understood. The feeding interactions of two contrasting ciliates, the free-swimming filter feeder Tetrahymena sp. and the surface-associated predator Chilodonella sp., were investigated using biofilm-forming bacteria genetically modified to express fluorescent proteins. According to microscopy, both ciliates readily consumed cells from both Pseudomonas costantinii and Serratia plymuthica biofilms. When offered a choice between spatially separated biofilms, each ciliate showed a preference for P. costantinii biofilms. Experiments with bacterial cell extracts indicated that both ciliates used dissolved chemical cues to locate biofilms. Chilodonella sp. evidently used bacterial chemical cues as a basis for preferential feeding decisions, but it was unclear whether Tetrahymena sp. did also. Confocal microscopy of live biofilms revealed that Tetrahymena sp. had a major impact on biofilm morphology, forming holes and channels throughout S. plymuthica biofilms and reducing P. costantinii biofilms to isolated, grazing-resistant microcolonies. Grazing by Chilodonella sp. resulted in the development of less-defined trails through S. plymuthica biofilms and caused P. costantinii biofilms to become homogeneous scatterings of cells. It was not clear whether the observed feeding preferences for spatially separated P. costantinii biofilms over S. plymuthica biofilms resulted in selective targeting of P. costantinii cells in mixed biofilms. Grazing of mixed biofilms resulted in the depletion of both types of bacteria, with Tetrahymena sp. having a larger impact than Chilodonella sp., and effects similar to those seen in grazed single-species biofilms.

Fig. 1.

Top and side views of microcosms used in ciliate preferential-feeding experiments. In microcosm type 1 (A), biofilms on glass coverslips (16-mm diameter) or blocks of bacterial cell extracts were placed in the outermost chambers, separated by 35 mm, and ciliates were added to the middle chamber. In microcosm type 2 (B), biofilms on squares of acetate (16 mm²) or cell extracts were placed at each end, separated by 25 mm, and ciliates were added to the center.

Fig. 2.

Representative Chilodonella sp. (A and B) and Tetrahymena sp. (C and D) cells after feeding on GFP-expressing P. costantinii (A and C) or RFP-expressing S. plymuthica (B and D) biofilms. Ciliates were added to bacterial biofilms 45 min before being fixed with formalin. Composite images constructed from phase-contrast microscopy images overlaid with fluorescence microscopy images are shown.

Fig. 3.

Distribution of Tetrahymena sp. cells in microcosms (type 1) containing spatially separated P. costantinii and S. plymuthica biofilms (A) or bacterial cell extracts (B) and distribution of Chilodonella sp. cells in microcosms (type 2) containing biofilms (C) or bacterial cell extracts (D). Cell counts were made by direct microscopic observations and video recordings. Counts of Chilodonella sp. cells were not made until after sufficient time had elapsed to allow cells added in suspension to settle. Each data point is the mean number of counts at nine different locations on each biofilm (for Tetrahymena sp. experiments) or three separate replicates (for Chilodonella sp. experiments). Error bars represent ± one standard deviation.

Fig. 4.

Representative confocal microscopy images of live bacterial biofilms. Ungrazed biofilms of P. costantinii (i), S. plymuthica (ii), and both bacteria together (iii) after 24 and 72 h of growth. Chilodonella sp.-grazed biofilms of P. costantinii (iv), S. plymuthica (v), and both bacteria (vi) after 48 and 72 h (24 h of biofilm growth followed by 24 h and 48 h of grazing). Tetrahymena sp.-grazed biofilms of P. costantinii (vii), S. plymuthica (viii), and both bacteria (ix) after 48 and 72 h (24 h of biofilm growth followed by 24 and 48 h of grazing). The S. plymuthica-filled vacuoles of a feeding Tetrahymena sp. cell are indicated (arrow). Green fluorescence indicates P. costantinii cells. Red fluorescence indicates S. plymuthica cells. Yellow fluorescence in biofilms composed of both bacteria also represents S. plymuthica cells and is the result of superimposed red and green fluorescence due to the use of high-intensity blue laser illumination causing simultaneous fluorescence of both S. plymuthica and P. costantinii cells in combination with a dual-wavelength filter. The scale bar applies to all images.