The Anaplasma phagocytophilum PleC histidine kinase and PleD diguanylate cyclase two-component system and role of cyclic Di-GMP in host cell infection

Abstract

Anaplasma phagocytophilum, the etiologic agent of human granulocytic anaplasmosis (HGA), has genes predicted to encode three sensor kinases, one of which is annotated PleC, and three response regulators, one of which is PleD. Prior to this study, the roles of PleC and PleD in the obligatory intracellular parasitism of A. phagocytophilum and their biochemical activities were unknown. The present study illustrates the relevance of these factors by demonstrating that both pleC and pleD were expressed in an HGA patient. During A. phagocytophilum development in human promyelocytic HL-60 cells, PleC and PleD were synchronously upregulated at the exponential growth stage and downregulated prior to extracellular release. A recombinant PleC kinase domain (rPleCHKD) has histidine kinase activity; no activity was observed when the conserved site of phosphorylation was replaced with alanine. A recombinant PleD (rPleD) has autokinase activity using phosphorylated rPleCHKD as the phosphoryl donor but not with two other recombinant histidine kinases. rPleCHKD could not serve as the phosphoryl donor for a mutant rPleD (with a conserved aspartic acid, the site of phosphorylation, replaced by alanine) or two other A. phagocytophilum recombinant response regulators. rPleD had diguanylate cyclase activity to generate cyclic (c) di-GMP from GTP in vitro. UV cross-linking of A. phagocytophilum lysate with c-di-[(32)P]GMP detected an approximately 47-kDa endogenous protein, presumably c-di-GMP downstream receptor. A new hydrophobic c-di-GMP derivative, 2'-O-di(tert-butyldimethylsilyl)-c-di-GMP, inhibited A. phagocytophilum infection in HL-60 cells. Our results suggest that the two-component PleC-PleD system is a diguanylate cyclase and that a c-di-GMP-receptor complex regulates A. phagocytophilum intracellular infection.

FIG. 1.

Schematic representation of the domain structures of PleC and PleD in A. phagocytophilum and C. crescentus. The numbers represent amino acid residues. The horizontal bars above PleC and PleD indicate the regions cloned for the functional study. The percentages in parentheses between the two aligned proteins indicate sequence identity.

FIG. 2.

pleC and pleD are transcribed by A. phagocytophilum in the blood of an HGA patient and in human myelocytic leukemia HL-60 cells. M, molecular size marker; D, positive control (chromosomal DNA used as a template) for the PCR; + and − indicate the presence or absence of reverse transcriptase, respectively. Genes and base pair sizes of amplified products are indicated.

FIG. 3.

PleC and PleD are expressed by A. phagocytophilum in human myelocytic leukemia HL-60 cells. (A) Lanes 1 and 2, nickel affinity-purified rPleC; lanes 5 and 6, nickel affinity-purified rPleD; lanes 3 and 7, A. phagocytophilum-infected HL-60 cells; lanes 4 and 8, uninfected HL-60 cells. Lanes 1, 2, 5, and 6, 0.5 μg protein per lane; lanes 3, 4, 7, and 8, 20 μg protein per lane. Lanes 1 and 5, Coomassie brilliant blue stain; lanes 2 to 4, Western blotting using anti-rPleC; lanes 6 to 8, Western blotting using anti-rPleD. Note rPleCHKD (30 kDa) and A. phagocytophilum native PleC (50 kDa) recognition by anti-rPleC and rPleD (56 kDa) and A. phagocytophilum native PleD (52 kDa) recognition by anti-PleD. (B) Infected HL-60 cells at 2 days p.i. (60% infected cells) were double labeled for immunofluorescence. The following antisera were used: anti-A. phagocytophilum P44 (anti-mouse IgG; green; P44) and anti-rPleC or anti-rPleD (anti-rabbit IgG; red; PleC or PleD) MAbs. The images on the right are superimposed images viewed with green and red filters (Merge) and a phase-contrast image to show the relative intracellular location of morulae. As controls for immunofluorescence labeling, A. phagocytophilum-infected HL-60 cells were incubated with preimmune rabbit serum (Preimmune) and secondary conjugated anti-rabbit IgG or with mouse isotype IgG2b control and secondary conjugated anti-mouse IgG. Scale bar, 5 μm.

FIG. 4.

PleC and PleD expression in synchronously cultured A. phagocytophilum. (A) Synchronously cultured A. phagocytophilum. Ap, host cell-free A. phagocytophilum used as inoculum; Diff-Quik stain. L/D, Live/Dead BacLight bacterial-viability test of the host cell-free A. phagocytophilum; green, live bacteria; red, dead bacteria. (0 to 72) Synchronously cultured A. phagocytophilum in HL-60 cells using host cell-free A. phagocytophilum harvested from 0 to 72 h p.i.; Diff-Quik stain. The bacteria or morulae are indicated by arrowheads. Scale bar, 10 μm. (B) Synchronous growth of A. phagocytophilum as determined by quantitative PCR. Total DNA was extracted at 0 to 72 h p.i. The amount of DNA was determined by quantitative PCR and is presented relative to the DNA content measured at 0 h p.i., which was defined as 1. The values are the means ± standard deviations of triplicate samples. (C) Temporal expression of PleC and PleD during A. phagocytophilum intracellular development. PleC and PleD expression was determined by Western blot analysis using antisera against A. phagocytophilum rPleCHKD and rPleD. The numbers of bacterial-genome equivalents loaded per lane were normalized by real-time PCR using the single-copy genomic pleD as a template. Note the stronger band densities of both PleC and PleD at 60 h p.i. (times in hours are indicated below the blot).

FIG. 5.

A. phagocytophilum PleC has specific His-dependent autokinase activity, and phosphotransfer from A. phagocytophilum PleC to PleD is dependent on a specific Asp of PleD. (A) Autoradiogram showing the specific His-dependent autokinase activity of rPleCHKD. Ten micrograms each of the wild-type rPleCHKD (lane 1) and rPleCHKDH244A (lane 2) were incubated with [γ-³²P]ATP. Only the wild-type HKD was ³²P phosphorylated. Lanes 4 and 5, autoradiogram showing the specific Asp-dependent phosphotransfer of ³²P from rPleCHKD to rPleDs. rPleCHKD (2 μg) was incubated with [γ-³²P]ATP, followed by incubation with 20 μg of rPleD (lane 4) or rPleDD53A (lane 5). rPleD alone (lane 3) was incubated with [γ-³²P]ATP as a negative control. (B) Autoradiogram showing phosphotransfer of ³²P from rPleCHKD to three recombinant RRs. rPleCHKD (10 μg) was incubated with [γ-³²P]ATP (lane 1; PleC alone), followed by incubation with 20 μg wild-type rPleD (lane 2), rNtrX (lane 3), or rCtrA (lane 4). Phosphotransfer was evident only for rPleD (lane 2). (C) Autoradiogram showing phosphotransfer of ³²P from three SKs to rPleD. rPleCHKD, rCckAHKD, and rNtrYHKD (10 μg each) were incubated with [γ-³²P]ATP (lanes 1, 3, and 5; HKD alone), followed by incubation with rPleD (20 μg) (lanes 2, 4, and 6). Phosphotransfer occurred only from rPleC (lane 2). The numbers on the left indicate molecular masses in kDa.

FIG. 6.

A. phagocytophilum rPleD has diguanylate cyclase activity. rPleDCHis (20 μg) was incubated with 50 nmol GTP for 1 h at room temperature; then, the sample was boiled, centrifuged, filtered, loaded onto an HPLC reverse-phase RC₁₈ column, and eluted with a methanol gradient (dashed line). The solid line is a representative elution profile, and the retention times of GTP and a chemically synthesized c-di-GMP standard are indicated. OD₂₅₄, optical density at 254 nm.

FIG. 7.

A. phagocytophilum produces a c-di-GMP binding protein. Autoradiogram of host cell-free A. phagocytophilum lysate (AP) (70 μg) and nickel affinity-purified rPleD (PleD) (1 μg) incubated with c-di-[³²P]GMP, followed by UV cross-linking without (−) or with (+) 1,000-fold excess unlabeled c-di-GMP (2 μM). An ∼47-kDa protein band was detected (arrows), and the c-di-[³²P]GMP binding was blocked by unlabeled c-di-GMP.

FIG. 8.

2′-O-TBDMS-c-di-GMP inhibits A. phagocytophilum infection in human myelocytic leukemia HL-60 cells. A. phagocytophilum bacteria were pretreated with DMSO solvent control or 0.05 mM 2′-O-TBDMS-c-di-GMP and, after being washed, incubated with HL-60 cells in the absence of the compound. (A) Representative Diff-Quik-stained images of HL-60 cells infected with A. phagocytophilum pretreated with 0.05 mM 2′-O-TBDMS-c-di-GMP or DMSO at 0, 36, 84, or 132 h p.i. The arrowheads indicate bacterial inclusions or bacteria. Scale bar, 10 μm. (B) Relative numbers of organisms per HL-60 cell at 84 and 132 h p.i. determined by real-time PCR in triplicate wells. The percentages of infected HL-60 cells at 84 and 132 h p.i. were determined in 100 HL-60 cells in triplicate wells. *, significantly different from the DMSO solvent control by Student's t test (P < 0.05; n = 3).