Comparative Transcriptomics Reveals Distinct Gene Expressions of a Model Ciliated Protozoa Feeding on Bacteria-Free Medium, Digestible, and Digestion-Resistant Bacteria

Abstract

Bacterivory is an important ecological function of protists in natural ecosystems. However, there are diverse bacterial species resistant to protistan digestion, which reduces the carbon flow to higher trophic levels. So far, a molecular biological view of metabolic processes in heterotrophic protists during predation of bacterial preys of different digestibility is still lacking. In this study, we investigated the growth performance a ciliated protozoan Tetrahymena thermophila cultivated in a bacteria-free Super Proteose Peptone (SPP) medium (control), and in the media mixed with either a digestion-resistant bacterial species (DRB) or a digestible strain of E. coli (ECO). We found the protist population grew fastest in the SPP and slowest in the DRB treatment. Fluorescence in situ hybridization confirmed that there were indeed non-digested, viable bacteria in the ciliate cells fed with DRB, but none in other treatments. Comparative analysis of RNA-seq data showed that, relative to the control, 637 and 511 genes in T. thermophila were significantly and differentially expressed in the DRB and ECO treatments, respectively. The protistan expression of lysosomal proteases (especially papain-like cysteine proteinases), GH18 chitinases, and an isocitrate lyase were upregulated in both bacterial treatments. The genes encoding protease, glycosidase and involving glycolysis, TCA and glyoxylate cycles of carbon metabolic processes were higher expressed in the DRB treatment when compared with the ECO. Nevertheless, the genes for glutathione metabolism were more upregulated in the control than those in both bacterial treatments, regardless of the digestibility of the bacteria. The results of this study indicate that not only bacterial food but also digestibility of bacterial taxa modulate multiple metabolic processes in heterotrophic protists, which contribute to a better understanding of protistan bacterivory and bacteria-protists interactions on a molecular basis.

Keywords: Protozoa; RNA-seq; bacterivory; feeding; gene expression; microbial loop.

Figure 1

(A) Growth curves of Tetrahymena thermophila fed with water suspensions containing the selected bacterial strains (YT1–YT13), E. coli, and in the Super Proteose Peptone (SPP) medium. (B) A comparison of the growth performance of T. thermophila in three treatments: the axenic SPP, the SPP mixed with E. coli (ECO), and the SPP mixed with Bacillus sp. YT1 (BAC). The arrow indicates the timing point that cells were harvested for molecular analysis. Error bars represent the mean and standard errors. (C–N) Microphotographs of fixed cells of T. thermophila in bright field (C,G,K), after 4′,6-diamidino-2-phenylindole (DAPI) staining (D,H,L), fluoresence in situ hybridization using Cy3-labeled eubacterial probes (E,I,M), and overlay of DAPI and fluorescence in situ hybridization (FISH) (F,J,N), showing that there are no bacterial signals in the cells of the protist in the SPP treatment (C–F), few E. coli cells remained in the ECO treatment (G–J), and abundance of Bacillus sp. YT1 in the BAC treatment. (I–L). Scale bar: 20 μm.

Figure 2

Overview of transcriptomes of T. thermophila cultured in the BAC, ECO, and SPP treatments, respectively. (A) Distribution of gene expression level in each treatment. (B) Density plot of the densities of the log₁₀FPKM values across all genes, showing that fragments per kilobase of transcript per million fragments mapped (FPKM) ranges of the SPP (control) and two bacterial treatments were very similar, indicating no bias in the sequencing coverage among the treatments. (C) Principle component analysis and (D) multidimensional scaling based on all transcriptomic data, demonstrating the distinct gene expression pattern of the ciliate cultivated in the BAC (in red), compared to the SPP (in yellow) and the ECO treatment (in blue).

Figure 3

Visualization of gene expression changes of T. thermophila between any two of the treatments. (A) MA plots for the average expression and fold changes in log scale to visualize the change in gene expression and distribution of comparisons among three treatments for all 19,831 transcripts. The colored data points indicate the upregulated (in red), downregulated (in blue) and unaltered transcripts (in grey), respectively; scatter plots for each comparison of FPKM for all transcripts (inset). (B) Three-way Venn diagrams display the number of up- and downregulated genes that are shared and unique between these treatments. (C) Hierarchical cluster analysis performed on the profiles of 1919 differentially expressed transcripts, and the heatmap of log₂(fold change) showed a distinct expression pattern in ECO vs. SPP relative to the other two comparisons.

Figure 4

Gene ontology (GO) enrichment analysis of differentially expressed genes in T. thermophila. (A) The significantly upregulated genes (FPKM ≥ 1, log₂ (fold change) ≥ −2, and p < 0.05); and (B) the downregulated (FPKM ≥ 1, log₂ (fold change) ≤ −2, and p < 0.05) genes associated with biological process (BP), cellular component (CC), and molecular function (MF). The numbers inside the dots indicate the number of the associated genes.

Figure 5

Relative expression levels of differentially expressed genes (DEGs, solid circles; |log₂ (fold change)| ≥ 2) and significantly differentiated transcripts with relatively lower fold changes (SDTs, open circles; 0 < |log₂ (fold change)| < 2) of T. thermophila associated with lysosomes in the pairwise comparisons between two bacterial treatment (BAC and ECO) and SPP (A), and BAC vs. ECO (B). Each dot represents an individual gene, and sizes of dots indicate expression abundance. Abbreviations: CTS, cathepsin; LGMN, legumain; Papain, papain family cysteine protease; GH, glycoside hydrolase.

Figure 6

Relative expression levels of differentially expressed genes (DEGs, solid circles; |log₂ (fold change)| ≥ 2) and significantly differentiated transcripts with relatively lower fold changes (SDTs, open circles; 0 < |log₂ (fold change)| < 2) of T. thermophila involved in three carbon metabolism (glycolysis, TCA, and glyoxylate) and glutathione metabolism in the pairwise comparisons between two bacterial treatment (BAC and ECO) and SPP (A), and BAC vs. ECO (B). Each circle represents an individual gene, and its size indicates relative expression abundance. Abbreviations: GK, glucokinase; GPI, glucose-6-phosphate isomerase; PFK, 6-phosphofructokinase; FBP, fructose-1,6-bisphosphatase; FBA, fructose-bisphosphate aldolase; PK, pyruvate kinase; PEPCK, phosphoenolpyruvate carboxykinase; CS, citrate synthase; MD, malate dehydrogenase; ICL, isocitrate lyase; MS, malate synthase; GCL, glutamate-cysteine ligase; GS, glutathione synthase; GST, glutathione S-transferase; GP, glutathione peroxidase.

Figure 7

Schematic model illustrating the phagocytic processes and genes expression changes in T. thermophila under three treatments. Abbreviations: WTA, wall teichoic acid; LTA, lipoteichoic acid; LPS, lipopolysaccharide; ROS, Reactive oxygen species; GCL, glutamate-cysteine ligase; GS, glutathione synthase; GST, glutathione S-transferase; GP, glutathione peroxidase.