A Survey of Two-Component Systems in Coxiella burnetii Reveals Redundant Regulatory Schemes and a Requirement for an Atypical PhoBR System in Mammalian Cell Infection

Abstract

Coxiella burnetii is an obligate intracellular bacterium and the etiological agent of Q fever in humans. C. burnetii transitions between a replicative, metabolically active large-cell variant (LCV) and a spore-like, quiescent small-cell variant (SCV) as a likely mechanism to ensure survival between host cells and mammalian hosts. C. burnetii encodes three canonical two-component systems, four orphan hybrid histidine kinases, five orphan response regulators, and a histidine phosphotransfer protein, which have been speculated to play roles in the signaling required for C. burnetii morphogenesis and virulence. However, very few of these systems have been characterized. By employing a CRISPR interference system for genetic manipulation of C. burnetii, we created single- and multigene transcriptional knockdown strains targeting most of these signaling genes. Through this, we revealed a role for the C. burnetii PhoBR canonical two-component system in virulence, regulation of [P_i] maintenance, and P_i transport. We also outline a novel mechanism by which PhoBR function may be regulated by an atypical PhoU-like protein. We also determined that the GacA.2/GacA.3/GacA.4/GacS orphan response regulators coordinately and disparately regulate expression of SCV-associated genes in C. burnetii LCVs. These foundational results will inform future studies on the role of C. burnetii two-component systems in virulence and morphogenesis. IMPORTANCE C. burnetii is an obligate intracellular bacterium with a spore-like stability allowing it to survive long periods of time in the environment. This stability is likely due to its biphasic developmental cycle, whereby it can transition from an environmentally stable small-cell variant (SCV) to a metabolically active large-cell variant (LCV). Here, we define the role of two-component phosphorelay systems (TCS) in C. burnetii's ability to survive within the harsh environment contained in the phagolysosome of host cells. We show that the canonical PhoBR TCS has an important role in C. burnetii virulence and phosphate sensing. Further examination of the regulons controlled by orphan regulators indicated a role in modulating gene expression of SCV-associated genes, including genes essential for cell wall remodeling.

Keywords: bacteria; cell membranes; cell signaling; pathogenesis; phosphate; two-component regulatory systems.

FIG 1

CRISPRi KD strains of C. burnetii PhoBR-associated genes have replication defects in Vero cells. (A) Volcano plot of the average microcolony area measurements of three independent Vero infections (n = 3). At 4 days postinfection, for each Vero infection, fluorescence microscopy was used to capture six images (magnification, ×20) each for uninduced and CRISPRi-induced infections of each strain outlined in Table 2. The average microcolony size measurements from these three replicates were then plotted. Unpaired t tests for each strain were used to determine significance, with a P value threshold of 0.05. Blue circles represent CRISPRi KD strains (see Table 2) displaying a significantly decreased microcolony size when chemically induced in comparison to uninduced samples. Black circles represent CRISPRi KD strains in which chemical induction led to insignificant changes in microcolony size. Black arrows indicate strains chosen for further study. (B) Representative images (magnification, ×40) of select strains indicated with a black arrow in panel A. Thick black lines are used to separate image series from each strain. Scale bars, 20 μm. (C) Genetic arrangement of Pho-related genes. Thick arrows represent genes, while thin arrows on sticks represent putative promoters.

FIG 2

Growth curve analysis confirms replication phenotypes for select C. burnetii CRISPRi KD strains. For each strain in Fig. 1B, growth curve analyses in auxotrophic axenic medium, Vero cells (Veros), and differentiated THP-1 cells (Thp-1s) were performed. Each growth curve represents the average of four independent experiments (n = 4). Quantitative PCR (qPCR) was used to determine genome equivalents (GEs). Uninduced cultures are represented by blue lines and circles, while CRISPRi-induced cultures are represented by red lines and squares. Statistical significance was determined by an unpaired t test between the uninduced (blue lines) and CRISPRi-induced (red lines) strains (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 3

C. burnetii is maladapted to high [P_i] in axenic medium. (A) Growth curves of C. burnetii phoB-sgRNA1, hptA-sgRNA1, and pitA-sgRNA1 strains under various [P_i]. Each [P_i] condition is represented by a different color line, while uninduced and CRISPRi-induced samples are represented by circles and squares, respectively. The results are an average of four independent experiments (n = 4). Statistical significance was determined by an unpaired t test between the uninduced and CRISPRi-induced strains (*, P < 0.05; **, P < 0.01; ***, P < 0.001). (B) An intracellular [P_i] assay was carried out for each of the strains shown in panel A, for both uninduced (blue bars) and CRISPRi-induced (red bars) cultures under high and normal [P_i]. Five-day-old cultures were harvested, washed, and lysed, and the P_i content was measured. The results are an average of three independent experiments (n = 3). Significance was determined by an unpaired t test between the uninduced (blue circles) and CRISPRi-induced (red circles) strains (*, P < 0.05; **, P < 0.01).

FIG 4

PhoB regulates genes with predicted PHO box-like sequences. (A) List of genes differentially regulated in 5-day-old axenic cultures of the phoB-sgRNA strain versus the empty-sgRNA1 strain. The values represent the log₂ fold change of the indicated strain versus that of empty-sgRNA1. Genes that are significantly differentially expressed have values that are boldface and slightly larger than the values of genes that were not deemed to be significantly differentially expressed by DESeq2. The cells are shaded based on how strongly the genes are differentially expressed, with genes upregulated and downregulated in the CRISPRi KD strains being shaded red and blue, respectively. (B) Illustration of the consensus E. coli PHO box sequence in comparison to putative PHO box-like sequences within 150 bp upstream of the start codon for several C. burnetii genes that are either differentially regulated in the phoB-sgRNA1 strain by RNA-Seq or are common genes regulated by PhoB in other bacteria. Matching nucleotides are boldface, red, and slightly larger than nucleotides that do not match. The W nucleotide code represents a nucleotide that can be either A or T, while Y represents a nucleotide that can be C or T.

FIG 5

PhoB and Pap regulate a similar set of genes in both LCVs and SCVs. (A to D) A series of transcriptional mScarlet-based reporter assay results for C. burnetii phoB-sgRNA1 LCV (A) and SCV (B) axenic cultures and for pap-sgRNA1 LCV (C) and SCV (D) axenic cultures are shown (n = 3 for each assay). Low [P_i], 0.046 mM; Norm (normal) [P_i], 4.6 mM; High [P_i], 46 mM. Significance was determined by an unpaired t test between the uninduced (blue circles) and IPTG-induced (red circles) strains (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 6

CRISPRi KD strains of C. burnetii orphan RRs have replication defects in Vero cells. (A) Volcano plot of the average microcolony area measurements of three independent Vero infections (n = 3). At 4 dpi, for each Vero infection, fluorescence microscopy was used to capture six images (magnification, ×20) each for uninduced and CRISPRi-induced infections of each strain (as outlined in Table 2). The average microcolony size measurements from these three replicates were then plotted. Unpaired t tests for each strain were used to determine significance, with a P value threshold of 0.05. Blue and red circles represent CRISPRi KD strains (see Table 2) displaying significantly decreased and increased microcolony size, respectively, when chemically induced in comparison to uninduced samples. Black circles represent CRISPRi KD strains in which chemical induction led to insignificant changes in microcolony size. Black arrows indicate strains chosen for further study. (B) Representative images (magnification, ×40) of select strains denoted with a black arrow in panel A. Thick black lines are used to separate image series from each strain. Scale bars, 20 μm. (C) Proposed phosphotransfer schematic relevant to C. burnetii orphan RRs and SKs. Upon stimulation by an unknown signal, the ATP-binding kinase domain is activated (green arrow) and the histidine kinase domain is autophosphorylated (red arrow). A phosphotransfer reaction occurs (blue arrow), resulting in phosphorylation of a conserved aspartate residue in the SK receiver domain. An additional phosphotransfer reaction occurs (gold line), phosphorylating a conserved histidine residue in a histidine phosphotransfer domain. A final phosphotransfer reaction occurs (black line), resulting in phosphorylation of a conserved aspartate residue in the receiver domain of an orphan RR.

FIG 7

Growth curve analysis confirms replication phenotypes for select C. burnetii CRISPRi KD strains. For each strain in Fig. 6B, growth curve analyses in auxotrophic axenic medium, Vero cells, and differentiated THP-1 cells were performed. Each growth curve represents the average of four independent experiments (n = 4). Quantitative PCR (qPCR) was used to determine genome equivalents (GEs). Uninduced cultures are represented by blue lines and circles, while CRISPRi-induced cultures are represented by red lines and squares. Statistical significance was determined by an unpaired t test between the uninduced (blue lines) and CRISPRi-induced (red lines) strains (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 8

Multi-KD strains of C. burnetii orphan RRs/SKs produce replication defects in Vero cells. Various multi-KD strains (as outlined in Table 4) were used to infect Vero cells. Infected cells were harvested at both 0 dpi (D0) and 6 dpi (D6), and qPCR was used to determine GEs. The results are shown as an average of three independent experiments (n = 3). Statistical significance was determined by an unpaired t test between the uninduced (blue circles) and CRISPRi-induced (red circles) strains (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 9

RNA-Seq analysis identifies genes regulated by the C. burnetii orphan RRs/SKs in LCVs. (A) List of differentially expressed genes shared between C. burnetii strains with CRISPRi systems repressing gacA.2, gacA.3, gacA.4, and/or gacS. Five 5-day-old axenic cultures for each strain were harvested before RNA extraction, cDNA library preparation, and sequencing (n = 5). The values represent the log₂ fold change of the indicated strain versus that of empty-sgRNA1. Genes that are significantly differentially expressed have values that are boldface and slightly larger than the values of genes that were not deemed to be significantly differentially expressed by DESeq2. The cells are shaded based on how strongly the genes are differentially expressed, with genes upregulated and downregulated in the CRISPRi KD strains being shaded red and blue, respectively. There are also columns for whether the detected genes are SCV associated (SCV-assoc?) (^aSandoz et al., 2016 [36]) and if the differentially expressed genes are RpoS coregulated (RpoS Co-reg?) (^bMoormeier et al., 2019 [37]). (B) List of genes differentially expressed in strains where gacA.1 or the three orphan sensor kinase genes gacS.1/gacS.2/gacS.3 are repressed.

FIG 10

Repression of gacA.2 and gacA.3 causes increased outer membrane vesiculation in C. burnetii LCVs. (A) Representative SEM images (top panel) and negatively stained TEM images (bottom panel) for 5-day-old axenic cultures of the indicated strains, displaying the increased OM vesiculation phenotype. Black arrows indicate OM vesicles in the negatively stained TEM images. (B) Quantitative measure of the percentage of C. burnetii cells with OM vesicles for each strain examined. For each strain, ≥35 individual bacterial cells were examined. Blue bars represent uninduced cultures of the respective strains, while red bars represent the results of CRISPRi-induced cultures.

FIG 11

Model for regulation by C. burnetii orphan SKs/RRs and the PhoBR canonical TCS in LCVs. (Left panel) A C. burnetii SK, which may or may not be membrane bound, transduces an unknown signal to C. burnetii orphan RRs via the CBU0351 histidine phosphotransfer domain-containing protein. (Right panel) A potential model for C. burnetii PhoBR regulation by the sensing of low extracellular [Pi] by PitA and Pap, which regulate PhoR by an unknown mechanism (gold arrow). Red arrows in both panels represent phosphotransfer reactions, while black arrows represent transcriptional regulation events.