Biphasic Metabolism and Host Interaction of a Chlamydial Symbiont

Abstract

Chlamydiae are obligate intracellular bacteria comprising well-known human pathogens and ubiquitous symbionts of protists, which are characterized by a unique developmental cycle. Here we comprehensively analyzed gene expression dynamics of Protochlamydia amoebophila during infection of its Acanthamoeba host by RNA sequencing. This revealed a highly dynamic transcriptional landscape, where major transcriptional shifts are conserved among chlamydial symbionts and pathogens. Our data served to propose a time-resolved model for type III protein secretion during the developmental cycle, and we provide evidence for a biphasic metabolism of P. amoebophila during infection, which involves energy parasitism and amino acids as the carbon source during initial stages and a postreplicative switch to endogenous glucose-based ATP production. This fits well with major transcriptional changes in the amoeba host, where upregulation of complex sugar breakdown precedes the P. amoebophila metabolic switch. The biphasic chlamydial metabolism represents a unique adaptation to exploit eukaryotic host cells, which likely contributed to the evolutionary success of this group of microbes. IMPORTANCE Chlamydiae are known as major bacterial pathogens of humans, causing the ancient disease trachoma, but they are also frequently found in the environment where they infect ubiquitous protists such as amoebae. All known chlamydiae require a eukaryotic host cell to thrive. Using the environmental chlamydia Protochlamydia amoebophila within its natural host, Acanthamoeba castellanii, we investigated gene expression dynamics in vivo and throughout the complete chlamydial developmental cycle for the first time. This allowed us to infer how a major virulence mechanism, the type III secretion system, is regulated and employed, and we show that the physiology of chlamydiae undergoes a complete shift regarding carbon metabolism and energy generation. This study provides comprehensive insights into the infection strategy of chlamydiae and reveals a unique adaptation to life within a eukaryotic host cell.

Keywords: Protochlamydia; RNA-seq; chlamydia; developmental cycle; gene expression; host-microbe interaction; metabolism; symbiont; type III secretion system.

FIG 1

Developmental cycle of Protochlamydia amoebophila. (A) Fluorescence in situ hybridization in combination with DAPI staining was used to differentiate between RBs (pink) and EBs (blue). At 2 h postinfection EBs clearly dominate, but a few cells have already started to convert to RBs (white arrowheads). Exclusively RBs were detected at 48 hpi. The first EBs (white arrows) were seen at 72 hpi. The inset for the 24 hpi image is an enlargement of dividing RBs (to enhance clarity, the DAPI signal is shown in white). Dotted white lines indicate the outlines of amoeba host cells. If not indicated otherwise, all bars represent 10 µm. (B) The course of EB production and release was quantified by collecting intra- and extracellular bacteria, respectively, at indicated time points, and subsequent reinfection of fresh amoebae. The first intracellular EBs were present at 72 hpi, the first release of EBs was observed at 96 hpi. Values that are significantly different (P < 0.05) at the various time points by one-way analysis of variance (ANOVA) and Tukey’s posttest are indicated by an asterisk (n = 3). Prop., proportion. (C) Transmission electron micrographs visualizing developmental events at the ultrastructural level. Black arrows point to bacterial cells in overview images. Black arrowheads indicate vesicles that were observed in the inclusion lumen from 24 hpi on. Bars = 1 µm.

FIG 2

Temporal classes of gene expression during the Protochlamydia developmental cycle. A total of 797 genes were detected as differentially expressed; tRNA genes (20 genes), rRNA genes (2 genes), and genes detected only in a single replicate (38 genes) were excluded from further analysis. Clustering identified three main temporal classes of gene sets (colored bars to the left of the heatmap) that could be further divided into five large subclasses (colored bars to the right of the heatmap). The largest group of genes was most highly expressed early (n = 304), whereas the expression of the smallest group of genes peaked at midcycle when only RBs were present (n = 161). The third main gene cluster generally showed highest expression at the end of the cycle and the extracellular stage (n = 273; see Data Set S1 in the supplemental material). Gene products detected in the EB proteome in a previous study (n = 231) (39) are indicated in the bar plot next to the heatmap. To illustrate the course of gene expression for each subcluster, the expression values (log₂ RPKM plotted on the y axis) were averaged per time point (x axis) and visualized as line plots (error bars indicate standard deviations). Selected gene names are shown for each of the temporal clusters. RPKM, reads per kilobase per million; hpi, h postinfection; extracell., extracellular; PG, peptidoglycan synthesis; T3SS, type III secretion system.

FIG 3

Enrichment of functional categories by temporal class. The overrepresentation of functional categories among genes assigned to temporal classes provides evidence for stage-specific activities during the Protochlamydia developmental cycle. Only functional categories that were significantly enriched with a false-discovery rate (FDR) of ≤0.05 are shown here; the color code indicates the degree of significance (dark red indicates highly significant). The significant enrichment of putatively type III secreted gene products was tested using Fisher’s exact test, and only P values of ≤0.05 are shown. cat, category.

FIG 4

Course of Protochlamydia type III secretion system activity during the developmental cycle. This model is based on the observation that structural components of the type III secretion system and its (putative) effectors are expressed at different time points during the developmental cycle (Fig. S4). This suggests a scenario in which novel, fully assembled, and thus functional secretion systems occur only late in the developmental cycle, and type III secretion reaches its full capacity and highest activity during early stages of the infection. The indicated polarity of the active type III secretion system has been shown for C. trachomatis (54), but it is unclear whether this is also true for P. amoebophila, as the symbionts reside within single-cell inclusions. The color code for type III secretion components and effectors (nomenclature according to Hueck [105]) refers to the respective temporal gene expression classes (Fig. 2 and Fig. S4). Circles inside the cells represent chaperones. Differentially expressed components/effectors are labeled with asterisks; PEX1 and PEX2 refer to members of the expanded effector gene families in Protochlamydia (33); pc0309 is an ortholog of the putative chaperone encoded by CT274 (106). inc, inclusion membrane; hcm, host cell membrane; im, inner membrane; om, outer membrane.

FIG 5

Expression maps of selected metabolic pathways of Protochlamydia. A pronounced expression of the ATP/ADP translocase (ntt1) at midcycle and an early-to-mid activity of genes involved in amino acid breakdown to pyruvate is observed, whereas pathways involved in central carbon metabolism and energy generation were generally only upregulated at later stages (with the ATPases indicated by purple boxes being notable exceptions). This suggests that a major metabolic shift occurs during the developmental cycle and provides evidence for a stage-specific metabolism. All genes marked with an asterisk were detected to be significantly differentially expressed. RPKM, reads per kilobase per million; hpi, h postinfection; extracell., extracellular.

FIG 6

Biphasic metabolism of Protochlamydia during development in Acanthamoeba castellanii. This model is based on observed transcriptional patterns, enriched functional categories at different developmental stages (Fig. 2, 3, and 5 and Fig. S2A), and independent experimental evidence reported previously (see text for references). Activity of metabolic pathways as inferred from gene expression levels followed similar trends early and at midcycle, and at the two later stages, respectively. This suggests that ATP import and an amino acid-based anabolism prevails during the EB-to-RB transition and RB replication. Later stages are characterized by a glucose-based metabolism and a pronounced increase in the activity of the tricarboxylic acid (TCA) cycle and oxidative (ox.) phosphorylation pathway. Nucleotide transporters (Ntt’s) are shown in blue, and amino acid (AA) and oligopeptide transporters are shown in green. “Multi” indicates that multiple amino acid/peptide transporters with different substrate specificities are expressed. Question marks refer to hypothetical transporters not yet identified. Asterisks indicate an increased expression at the RB stage compared to the early time point. “RNA” denotes transcription, whereas “DNA” indicates DNA replication. Glc-6-P, glucose 6-phosphate; PPP, pentose phosphate pathway; glyconeog., gluconeogenesis.